| { | |
| "best_metric": 2.129138708114624, | |
| "best_model_checkpoint": "/home/sunggeunan/data/ICL/outputs/lora/SKIML-ICL_mrqa_nq_v3/Meta-Llama-3-8B-Instruct-unanswerable-2Q-1U-0C-qa_first/checkpoint-402", | |
| "epoch": 0.9996891513832763, | |
| "eval_steps": 500, | |
| "global_step": 402, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024867889337892445, | |
| "grad_norm": 0.4153629243373871, | |
| "learning_rate": 8.19672131147541e-09, | |
| "loss": 2.121, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004973577867578489, | |
| "grad_norm": 0.4008300006389618, | |
| "learning_rate": 1.639344262295082e-08, | |
| "loss": 2.1542, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007460366801367734, | |
| "grad_norm": 0.42680642008781433, | |
| "learning_rate": 2.459016393442623e-08, | |
| "loss": 2.1439, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009947155735156978, | |
| "grad_norm": 0.3831591308116913, | |
| "learning_rate": 3.278688524590164e-08, | |
| "loss": 2.1025, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012433944668946224, | |
| "grad_norm": 0.40637049078941345, | |
| "learning_rate": 4.0983606557377046e-08, | |
| "loss": 2.1374, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014920733602735468, | |
| "grad_norm": 0.3883218467235565, | |
| "learning_rate": 4.918032786885246e-08, | |
| "loss": 2.1412, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.017407522536524712, | |
| "grad_norm": 0.42491665482521057, | |
| "learning_rate": 5.7377049180327866e-08, | |
| "loss": 2.205, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.019894311470313956, | |
| "grad_norm": 0.40381714701652527, | |
| "learning_rate": 6.557377049180328e-08, | |
| "loss": 2.1575, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.022381100404103203, | |
| "grad_norm": 0.3807780146598816, | |
| "learning_rate": 7.377049180327868e-08, | |
| "loss": 2.2206, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.024867889337892447, | |
| "grad_norm": 0.3880959451198578, | |
| "learning_rate": 8.196721311475409e-08, | |
| "loss": 2.1772, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02735467827168169, | |
| "grad_norm": 0.36500561237335205, | |
| "learning_rate": 9.01639344262295e-08, | |
| "loss": 2.1232, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.029841467205470935, | |
| "grad_norm": 0.3805394172668457, | |
| "learning_rate": 9.836065573770492e-08, | |
| "loss": 2.1446, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03232825613926018, | |
| "grad_norm": 0.38014543056488037, | |
| "learning_rate": 1.0655737704918032e-07, | |
| "loss": 2.1277, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.034815045073049423, | |
| "grad_norm": 0.3875851333141327, | |
| "learning_rate": 1.1475409836065573e-07, | |
| "loss": 2.1064, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03730183400683867, | |
| "grad_norm": 0.39165419340133667, | |
| "learning_rate": 1.2295081967213113e-07, | |
| "loss": 2.1352, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03978862294062791, | |
| "grad_norm": 0.4067535102367401, | |
| "learning_rate": 1.3114754098360656e-07, | |
| "loss": 2.1601, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.042275411874417156, | |
| "grad_norm": 0.41718506813049316, | |
| "learning_rate": 1.3934426229508196e-07, | |
| "loss": 2.1471, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04476220080820641, | |
| "grad_norm": 0.4221360981464386, | |
| "learning_rate": 1.4754098360655736e-07, | |
| "loss": 2.1499, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04724898974199565, | |
| "grad_norm": 0.39923396706581116, | |
| "learning_rate": 1.5573770491803278e-07, | |
| "loss": 2.1003, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.049735778675784895, | |
| "grad_norm": 0.3728751242160797, | |
| "learning_rate": 1.6393442622950818e-07, | |
| "loss": 2.0842, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05222256760957414, | |
| "grad_norm": 0.3873041868209839, | |
| "learning_rate": 1.7213114754098358e-07, | |
| "loss": 2.1152, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05470935654336338, | |
| "grad_norm": 0.3714573383331299, | |
| "learning_rate": 1.80327868852459e-07, | |
| "loss": 2.1215, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05719614547715263, | |
| "grad_norm": 0.4204677939414978, | |
| "learning_rate": 1.885245901639344e-07, | |
| "loss": 2.1859, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05968293441094187, | |
| "grad_norm": 0.4137566089630127, | |
| "learning_rate": 1.9672131147540984e-07, | |
| "loss": 2.1317, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.062169723344731115, | |
| "grad_norm": 0.3629921078681946, | |
| "learning_rate": 2.0491803278688524e-07, | |
| "loss": 2.0563, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06465651227852036, | |
| "grad_norm": 0.39072492718696594, | |
| "learning_rate": 2.1311475409836064e-07, | |
| "loss": 2.164, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0671433012123096, | |
| "grad_norm": 0.37331125140190125, | |
| "learning_rate": 2.2131147540983606e-07, | |
| "loss": 2.2048, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06963009014609885, | |
| "grad_norm": 0.3944483697414398, | |
| "learning_rate": 2.2950819672131146e-07, | |
| "loss": 2.1835, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07211687907988809, | |
| "grad_norm": 0.39379164576530457, | |
| "learning_rate": 2.3770491803278686e-07, | |
| "loss": 2.1465, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07460366801367734, | |
| "grad_norm": 0.3914564549922943, | |
| "learning_rate": 2.4590163934426226e-07, | |
| "loss": 2.091, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07709045694746658, | |
| "grad_norm": 0.4301564395427704, | |
| "learning_rate": 2.540983606557377e-07, | |
| "loss": 2.1183, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07957724588125582, | |
| "grad_norm": 0.40827327966690063, | |
| "learning_rate": 2.622950819672131e-07, | |
| "loss": 2.1588, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08206403481504507, | |
| "grad_norm": 0.3868783116340637, | |
| "learning_rate": 2.704918032786885e-07, | |
| "loss": 2.1667, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08455082374883431, | |
| "grad_norm": 0.40489786863327026, | |
| "learning_rate": 2.786885245901639e-07, | |
| "loss": 2.1857, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08703761268262357, | |
| "grad_norm": 0.3836217224597931, | |
| "learning_rate": 2.868852459016393e-07, | |
| "loss": 2.1313, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08952440161641281, | |
| "grad_norm": 0.4212404787540436, | |
| "learning_rate": 2.950819672131147e-07, | |
| "loss": 2.2024, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09201119055020206, | |
| "grad_norm": 0.395867258310318, | |
| "learning_rate": 3.0327868852459017e-07, | |
| "loss": 2.1355, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0944979794839913, | |
| "grad_norm": 0.3836336135864258, | |
| "learning_rate": 3.1147540983606557e-07, | |
| "loss": 2.1254, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09698476841778055, | |
| "grad_norm": 0.42212599515914917, | |
| "learning_rate": 3.1967213114754097e-07, | |
| "loss": 2.1262, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09947155735156979, | |
| "grad_norm": 0.43291711807250977, | |
| "learning_rate": 3.2786885245901637e-07, | |
| "loss": 2.1664, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10195834628535903, | |
| "grad_norm": 0.3876365125179291, | |
| "learning_rate": 3.3606557377049177e-07, | |
| "loss": 2.147, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10444513521914828, | |
| "grad_norm": 0.39074528217315674, | |
| "learning_rate": 3.4426229508196717e-07, | |
| "loss": 2.1113, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10693192415293752, | |
| "grad_norm": 0.4135940670967102, | |
| "learning_rate": 3.524590163934426e-07, | |
| "loss": 2.1834, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10941871308672677, | |
| "grad_norm": 0.4124310612678528, | |
| "learning_rate": 3.60655737704918e-07, | |
| "loss": 2.1019, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11190550202051601, | |
| "grad_norm": 0.3812576234340668, | |
| "learning_rate": 3.6885245901639347e-07, | |
| "loss": 2.1368, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11439229095430525, | |
| "grad_norm": 0.3919021189212799, | |
| "learning_rate": 3.770491803278688e-07, | |
| "loss": 2.1226, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1168790798880945, | |
| "grad_norm": 0.37712955474853516, | |
| "learning_rate": 3.852459016393442e-07, | |
| "loss": 2.1723, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.11936586882188374, | |
| "grad_norm": 0.40433424711227417, | |
| "learning_rate": 3.9344262295081967e-07, | |
| "loss": 2.1469, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12185265775567299, | |
| "grad_norm": 0.4323996603488922, | |
| "learning_rate": 4.0163934426229507e-07, | |
| "loss": 2.1625, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12433944668946223, | |
| "grad_norm": 0.37467238306999207, | |
| "learning_rate": 4.0983606557377047e-07, | |
| "loss": 2.0877, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1268262356232515, | |
| "grad_norm": 0.3842613399028778, | |
| "learning_rate": 4.180327868852459e-07, | |
| "loss": 2.1367, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12931302455704072, | |
| "grad_norm": 0.41727927327156067, | |
| "learning_rate": 4.2622950819672127e-07, | |
| "loss": 2.158, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13179981349082998, | |
| "grad_norm": 0.427172988653183, | |
| "learning_rate": 4.3442622950819667e-07, | |
| "loss": 2.185, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1342866024246192, | |
| "grad_norm": 0.3944658041000366, | |
| "learning_rate": 4.426229508196721e-07, | |
| "loss": 2.1537, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.13677339135840846, | |
| "grad_norm": 0.3892759382724762, | |
| "learning_rate": 4.508196721311475e-07, | |
| "loss": 2.1187, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1392601802921977, | |
| "grad_norm": 0.40089288353919983, | |
| "learning_rate": 4.590163934426229e-07, | |
| "loss": 2.1299, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.14174696922598695, | |
| "grad_norm": 0.4039812982082367, | |
| "learning_rate": 4.672131147540984e-07, | |
| "loss": 2.1734, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14423375815977618, | |
| "grad_norm": 0.43650051951408386, | |
| "learning_rate": 4.754098360655737e-07, | |
| "loss": 2.1511, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.14672054709356544, | |
| "grad_norm": 0.40934914350509644, | |
| "learning_rate": 4.836065573770492e-07, | |
| "loss": 2.1244, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.14920733602735467, | |
| "grad_norm": 0.40043023228645325, | |
| "learning_rate": 4.918032786885245e-07, | |
| "loss": 2.1584, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15169412496114393, | |
| "grad_norm": 0.4129016697406769, | |
| "learning_rate": 5e-07, | |
| "loss": 2.158, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.15418091389493316, | |
| "grad_norm": 0.38239961862564087, | |
| "learning_rate": 4.995633187772926e-07, | |
| "loss": 2.1423, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.15666770282872242, | |
| "grad_norm": 0.4175527095794678, | |
| "learning_rate": 4.991266375545852e-07, | |
| "loss": 2.1224, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15915449176251165, | |
| "grad_norm": 0.4162661135196686, | |
| "learning_rate": 4.986899563318778e-07, | |
| "loss": 2.1816, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1616412806963009, | |
| "grad_norm": 0.4039028584957123, | |
| "learning_rate": 4.982532751091702e-07, | |
| "loss": 2.1734, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16412806963009013, | |
| "grad_norm": 0.4020048677921295, | |
| "learning_rate": 4.978165938864628e-07, | |
| "loss": 2.1546, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1666148585638794, | |
| "grad_norm": 0.4302126169204712, | |
| "learning_rate": 4.973799126637554e-07, | |
| "loss": 2.1906, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.16910164749766862, | |
| "grad_norm": 0.4021979570388794, | |
| "learning_rate": 4.96943231441048e-07, | |
| "loss": 2.1297, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.17158843643145788, | |
| "grad_norm": 0.3942105174064636, | |
| "learning_rate": 4.965065502183406e-07, | |
| "loss": 2.1292, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.17407522536524714, | |
| "grad_norm": 0.4106265902519226, | |
| "learning_rate": 4.960698689956332e-07, | |
| "loss": 2.1454, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17656201429903637, | |
| "grad_norm": 0.4014648199081421, | |
| "learning_rate": 4.956331877729257e-07, | |
| "loss": 2.0864, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17904880323282563, | |
| "grad_norm": 0.41134366393089294, | |
| "learning_rate": 4.951965065502184e-07, | |
| "loss": 2.1021, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.18153559216661486, | |
| "grad_norm": 0.40096017718315125, | |
| "learning_rate": 4.947598253275109e-07, | |
| "loss": 2.1465, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.18402238110040411, | |
| "grad_norm": 0.41377922892570496, | |
| "learning_rate": 4.943231441048035e-07, | |
| "loss": 2.1694, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.18650917003419334, | |
| "grad_norm": 0.39552953839302063, | |
| "learning_rate": 4.93886462882096e-07, | |
| "loss": 2.1748, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1889959589679826, | |
| "grad_norm": 0.44786471128463745, | |
| "learning_rate": 4.934497816593886e-07, | |
| "loss": 2.1785, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.19148274790177183, | |
| "grad_norm": 0.42525768280029297, | |
| "learning_rate": 4.930131004366812e-07, | |
| "loss": 2.1825, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1939695368355611, | |
| "grad_norm": 0.427071750164032, | |
| "learning_rate": 4.925764192139738e-07, | |
| "loss": 2.1463, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.19645632576935032, | |
| "grad_norm": 0.41076913475990295, | |
| "learning_rate": 4.921397379912663e-07, | |
| "loss": 2.124, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.19894311470313958, | |
| "grad_norm": 0.4056430160999298, | |
| "learning_rate": 4.917030567685589e-07, | |
| "loss": 2.143, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2014299036369288, | |
| "grad_norm": 0.4058414101600647, | |
| "learning_rate": 4.912663755458515e-07, | |
| "loss": 2.1337, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.20391669257071807, | |
| "grad_norm": 0.4427083730697632, | |
| "learning_rate": 4.908296943231441e-07, | |
| "loss": 2.1631, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2064034815045073, | |
| "grad_norm": 0.4002906084060669, | |
| "learning_rate": 4.903930131004367e-07, | |
| "loss": 2.0538, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.20889027043829655, | |
| "grad_norm": 0.40065857768058777, | |
| "learning_rate": 4.899563318777293e-07, | |
| "loss": 2.1129, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.21137705937208578, | |
| "grad_norm": 0.42688536643981934, | |
| "learning_rate": 4.895196506550219e-07, | |
| "loss": 2.139, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.21386384830587504, | |
| "grad_norm": 0.4278879165649414, | |
| "learning_rate": 4.890829694323143e-07, | |
| "loss": 2.1395, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.21635063723966427, | |
| "grad_norm": 0.43649378418922424, | |
| "learning_rate": 4.886462882096069e-07, | |
| "loss": 2.0751, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.21883742617345353, | |
| "grad_norm": 0.3865818977355957, | |
| "learning_rate": 4.882096069868995e-07, | |
| "loss": 2.0862, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.22132421510724276, | |
| "grad_norm": 0.42509347200393677, | |
| "learning_rate": 4.877729257641921e-07, | |
| "loss": 2.1478, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.22381100404103202, | |
| "grad_norm": 0.4220832884311676, | |
| "learning_rate": 4.873362445414847e-07, | |
| "loss": 2.1609, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22629779297482125, | |
| "grad_norm": 0.40812230110168457, | |
| "learning_rate": 4.868995633187773e-07, | |
| "loss": 2.1551, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2287845819086105, | |
| "grad_norm": 0.4381932020187378, | |
| "learning_rate": 4.864628820960698e-07, | |
| "loss": 2.1594, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.23127137084239976, | |
| "grad_norm": 0.4095819890499115, | |
| "learning_rate": 4.860262008733625e-07, | |
| "loss": 2.1396, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.233758159776189, | |
| "grad_norm": 0.42679563164711, | |
| "learning_rate": 4.85589519650655e-07, | |
| "loss": 2.1771, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.23624494870997825, | |
| "grad_norm": 0.44320186972618103, | |
| "learning_rate": 4.851528384279476e-07, | |
| "loss": 2.1842, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.23873173764376748, | |
| "grad_norm": 0.39184531569480896, | |
| "learning_rate": 4.847161572052402e-07, | |
| "loss": 2.1378, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.24121852657755674, | |
| "grad_norm": 0.4558006823062897, | |
| "learning_rate": 4.842794759825327e-07, | |
| "loss": 2.1706, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.24370531551134597, | |
| "grad_norm": 0.4423806667327881, | |
| "learning_rate": 4.838427947598253e-07, | |
| "loss": 2.1841, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.24619210444513523, | |
| "grad_norm": 0.4314688444137573, | |
| "learning_rate": 4.834061135371178e-07, | |
| "loss": 2.1154, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.24867889337892446, | |
| "grad_norm": 0.44223618507385254, | |
| "learning_rate": 4.829694323144104e-07, | |
| "loss": 2.1124, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2511656823127137, | |
| "grad_norm": 0.44006800651550293, | |
| "learning_rate": 4.82532751091703e-07, | |
| "loss": 2.1937, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.253652471246503, | |
| "grad_norm": 0.4089645445346832, | |
| "learning_rate": 4.820960698689956e-07, | |
| "loss": 2.1236, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2561392601802922, | |
| "grad_norm": 0.4407235383987427, | |
| "learning_rate": 4.816593886462882e-07, | |
| "loss": 2.1116, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.25862604911408144, | |
| "grad_norm": 0.4155865013599396, | |
| "learning_rate": 4.812227074235808e-07, | |
| "loss": 2.1259, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.26111283804787067, | |
| "grad_norm": 0.4286578893661499, | |
| "learning_rate": 4.807860262008734e-07, | |
| "loss": 2.1831, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.26359962698165995, | |
| "grad_norm": 0.44605061411857605, | |
| "learning_rate": 4.80349344978166e-07, | |
| "loss": 2.1373, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2660864159154492, | |
| "grad_norm": 0.4532274007797241, | |
| "learning_rate": 4.799126637554585e-07, | |
| "loss": 2.1249, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2685732048492384, | |
| "grad_norm": 0.4327315092086792, | |
| "learning_rate": 4.79475982532751e-07, | |
| "loss": 2.1414, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.27105999378302764, | |
| "grad_norm": 0.4438115358352661, | |
| "learning_rate": 4.790393013100436e-07, | |
| "loss": 2.0787, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2735467827168169, | |
| "grad_norm": 0.4239655137062073, | |
| "learning_rate": 4.786026200873362e-07, | |
| "loss": 2.1234, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.27603357165060616, | |
| "grad_norm": 0.4541226327419281, | |
| "learning_rate": 4.781659388646288e-07, | |
| "loss": 2.2023, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2785203605843954, | |
| "grad_norm": 0.4275488555431366, | |
| "learning_rate": 4.777292576419214e-07, | |
| "loss": 2.1872, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2810071495181846, | |
| "grad_norm": 0.4427001476287842, | |
| "learning_rate": 4.772925764192139e-07, | |
| "loss": 2.1646, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2834939384519739, | |
| "grad_norm": 0.43924546241760254, | |
| "learning_rate": 4.768558951965065e-07, | |
| "loss": 2.1175, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.28598072738576313, | |
| "grad_norm": 0.42863723635673523, | |
| "learning_rate": 4.764192139737991e-07, | |
| "loss": 2.1269, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.28846751631955236, | |
| "grad_norm": 0.40726035833358765, | |
| "learning_rate": 4.759825327510917e-07, | |
| "loss": 2.071, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.29095430525334165, | |
| "grad_norm": 0.4581323564052582, | |
| "learning_rate": 4.7554585152838427e-07, | |
| "loss": 2.1981, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2934410941871309, | |
| "grad_norm": 0.46849963068962097, | |
| "learning_rate": 4.751091703056768e-07, | |
| "loss": 2.1689, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2959278831209201, | |
| "grad_norm": 0.45309266448020935, | |
| "learning_rate": 4.746724890829694e-07, | |
| "loss": 2.1122, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.29841467205470934, | |
| "grad_norm": 0.4709586203098297, | |
| "learning_rate": 4.7423580786026193e-07, | |
| "loss": 2.1774, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3009014609884986, | |
| "grad_norm": 0.4502153992652893, | |
| "learning_rate": 4.737991266375546e-07, | |
| "loss": 2.1357, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.30338824992228786, | |
| "grad_norm": 0.4565674066543579, | |
| "learning_rate": 4.7336244541484717e-07, | |
| "loss": 2.1982, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3058750388560771, | |
| "grad_norm": 0.4817062020301819, | |
| "learning_rate": 4.729257641921397e-07, | |
| "loss": 2.1124, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3083618277898663, | |
| "grad_norm": 0.43636277318000793, | |
| "learning_rate": 4.724890829694323e-07, | |
| "loss": 2.1345, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3108486167236556, | |
| "grad_norm": 0.4348713159561157, | |
| "learning_rate": 4.720524017467249e-07, | |
| "loss": 2.1382, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.31333540565744483, | |
| "grad_norm": 0.43462586402893066, | |
| "learning_rate": 4.7161572052401743e-07, | |
| "loss": 2.1708, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.31582219459123406, | |
| "grad_norm": 0.4370459318161011, | |
| "learning_rate": 4.7117903930131e-07, | |
| "loss": 2.128, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3183089835250233, | |
| "grad_norm": 0.43912699818611145, | |
| "learning_rate": 4.7074235807860256e-07, | |
| "loss": 2.1016, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3207957724588126, | |
| "grad_norm": 0.4648686647415161, | |
| "learning_rate": 4.7030567685589515e-07, | |
| "loss": 2.073, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3232825613926018, | |
| "grad_norm": 0.45214056968688965, | |
| "learning_rate": 4.6986899563318775e-07, | |
| "loss": 2.1608, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.32576935032639104, | |
| "grad_norm": 0.4569113850593567, | |
| "learning_rate": 4.6943231441048034e-07, | |
| "loss": 2.1456, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.32825613926018027, | |
| "grad_norm": 0.4468495547771454, | |
| "learning_rate": 4.6899563318777293e-07, | |
| "loss": 2.104, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.33074292819396955, | |
| "grad_norm": 0.4579126834869385, | |
| "learning_rate": 4.685589519650655e-07, | |
| "loss": 2.1415, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3332297171277588, | |
| "grad_norm": 0.41133925318717957, | |
| "learning_rate": 4.6812227074235806e-07, | |
| "loss": 2.0556, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.335716506061548, | |
| "grad_norm": 0.45494189858436584, | |
| "learning_rate": 4.6768558951965065e-07, | |
| "loss": 2.1353, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.33820329499533724, | |
| "grad_norm": 0.45598360896110535, | |
| "learning_rate": 4.672489082969432e-07, | |
| "loss": 2.1608, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.34069008392912653, | |
| "grad_norm": 0.4804225564002991, | |
| "learning_rate": 4.668122270742358e-07, | |
| "loss": 2.125, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.34317687286291576, | |
| "grad_norm": 0.45764321088790894, | |
| "learning_rate": 4.6637554585152837e-07, | |
| "loss": 2.1547, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.345663661796705, | |
| "grad_norm": 0.469670832157135, | |
| "learning_rate": 4.659388646288209e-07, | |
| "loss": 2.1538, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3481504507304943, | |
| "grad_norm": 0.47154125571250916, | |
| "learning_rate": 4.655021834061135e-07, | |
| "loss": 2.1144, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3506372396642835, | |
| "grad_norm": 0.4631963074207306, | |
| "learning_rate": 4.6506550218340604e-07, | |
| "loss": 2.194, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.35312402859807274, | |
| "grad_norm": 0.4393676519393921, | |
| "learning_rate": 4.646288209606987e-07, | |
| "loss": 2.0945, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.35561081753186197, | |
| "grad_norm": 0.444672554731369, | |
| "learning_rate": 4.641921397379913e-07, | |
| "loss": 2.111, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.35809760646565125, | |
| "grad_norm": 0.46494928002357483, | |
| "learning_rate": 4.637554585152838e-07, | |
| "loss": 2.1777, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3605843953994405, | |
| "grad_norm": 0.4453777074813843, | |
| "learning_rate": 4.633187772925764e-07, | |
| "loss": 2.1631, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3630711843332297, | |
| "grad_norm": 0.4813487231731415, | |
| "learning_rate": 4.62882096069869e-07, | |
| "loss": 2.1595, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.36555797326701894, | |
| "grad_norm": 0.4742406904697418, | |
| "learning_rate": 4.6244541484716154e-07, | |
| "loss": 2.1576, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.36804476220080823, | |
| "grad_norm": 0.45844781398773193, | |
| "learning_rate": 4.6200873362445413e-07, | |
| "loss": 2.1756, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.37053155113459746, | |
| "grad_norm": 0.44323110580444336, | |
| "learning_rate": 4.6157205240174667e-07, | |
| "loss": 2.1483, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3730183400683867, | |
| "grad_norm": 0.49860548973083496, | |
| "learning_rate": 4.6113537117903926e-07, | |
| "loss": 2.2144, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3755051290021759, | |
| "grad_norm": 0.43856751918792725, | |
| "learning_rate": 4.6069868995633185e-07, | |
| "loss": 2.0581, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3779919179359652, | |
| "grad_norm": 0.4472333788871765, | |
| "learning_rate": 4.602620087336244e-07, | |
| "loss": 2.117, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.38047870686975443, | |
| "grad_norm": 0.4927634298801422, | |
| "learning_rate": 4.5982532751091704e-07, | |
| "loss": 2.153, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.38296549580354367, | |
| "grad_norm": 0.4599962532520294, | |
| "learning_rate": 4.5938864628820963e-07, | |
| "loss": 2.2226, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3854522847373329, | |
| "grad_norm": 0.45448487997055054, | |
| "learning_rate": 4.5895196506550217e-07, | |
| "loss": 2.1189, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3879390736711222, | |
| "grad_norm": 0.4686853587627411, | |
| "learning_rate": 4.5851528384279476e-07, | |
| "loss": 2.1257, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3904258626049114, | |
| "grad_norm": 0.5059552192687988, | |
| "learning_rate": 4.580786026200873e-07, | |
| "loss": 2.1586, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.39291265153870064, | |
| "grad_norm": 0.4529350996017456, | |
| "learning_rate": 4.576419213973799e-07, | |
| "loss": 2.1436, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.39539944047248987, | |
| "grad_norm": 0.44359931349754333, | |
| "learning_rate": 4.572052401746725e-07, | |
| "loss": 2.1086, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.39788622940627916, | |
| "grad_norm": 0.4643580913543701, | |
| "learning_rate": 4.56768558951965e-07, | |
| "loss": 2.1566, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4003730183400684, | |
| "grad_norm": 0.47713202238082886, | |
| "learning_rate": 4.563318777292576e-07, | |
| "loss": 2.1494, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4028598072738576, | |
| "grad_norm": 0.4480564296245575, | |
| "learning_rate": 4.5589519650655015e-07, | |
| "loss": 2.1188, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4053465962076469, | |
| "grad_norm": 0.4450179636478424, | |
| "learning_rate": 4.554585152838428e-07, | |
| "loss": 2.1035, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.40783338514143613, | |
| "grad_norm": 0.4772661328315735, | |
| "learning_rate": 4.550218340611354e-07, | |
| "loss": 2.1598, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.41032017407522536, | |
| "grad_norm": 0.47646352648735046, | |
| "learning_rate": 4.545851528384279e-07, | |
| "loss": 2.1995, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4128069630090146, | |
| "grad_norm": 0.4821939468383789, | |
| "learning_rate": 4.541484716157205e-07, | |
| "loss": 2.1441, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4152937519428039, | |
| "grad_norm": 0.4545115828514099, | |
| "learning_rate": 4.537117903930131e-07, | |
| "loss": 2.1598, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4177805408765931, | |
| "grad_norm": 0.45019111037254333, | |
| "learning_rate": 4.5327510917030565e-07, | |
| "loss": 2.1148, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.42026732981038234, | |
| "grad_norm": 0.4689873456954956, | |
| "learning_rate": 4.5283842794759824e-07, | |
| "loss": 2.1509, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.42275411874417157, | |
| "grad_norm": 0.47073739767074585, | |
| "learning_rate": 4.524017467248908e-07, | |
| "loss": 2.1425, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.42524090767796086, | |
| "grad_norm": 0.4607613980770111, | |
| "learning_rate": 4.5196506550218337e-07, | |
| "loss": 2.1226, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4277276966117501, | |
| "grad_norm": 0.46717172861099243, | |
| "learning_rate": 4.5152838427947596e-07, | |
| "loss": 2.1222, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4302144855455393, | |
| "grad_norm": 0.5047352313995361, | |
| "learning_rate": 4.510917030567685e-07, | |
| "loss": 2.2121, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.43270127447932855, | |
| "grad_norm": 0.49366188049316406, | |
| "learning_rate": 4.5065502183406115e-07, | |
| "loss": 2.1806, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.43518806341311783, | |
| "grad_norm": 0.4741223454475403, | |
| "learning_rate": 4.502183406113537e-07, | |
| "loss": 2.0808, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.43767485234690706, | |
| "grad_norm": 0.4672994911670685, | |
| "learning_rate": 4.497816593886463e-07, | |
| "loss": 2.1021, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4401616412806963, | |
| "grad_norm": 0.4813832640647888, | |
| "learning_rate": 4.4934497816593887e-07, | |
| "loss": 2.139, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4426484302144855, | |
| "grad_norm": 0.4757406413555145, | |
| "learning_rate": 4.489082969432314e-07, | |
| "loss": 2.1245, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4451352191482748, | |
| "grad_norm": 0.4674074351787567, | |
| "learning_rate": 4.48471615720524e-07, | |
| "loss": 2.1619, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.44762200808206404, | |
| "grad_norm": 0.4354044795036316, | |
| "learning_rate": 4.480349344978166e-07, | |
| "loss": 2.0624, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.45010879701585327, | |
| "grad_norm": 0.4484567940235138, | |
| "learning_rate": 4.4759825327510913e-07, | |
| "loss": 2.1071, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4525955859496425, | |
| "grad_norm": 0.4580535292625427, | |
| "learning_rate": 4.471615720524017e-07, | |
| "loss": 2.1039, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4550823748834318, | |
| "grad_norm": 0.504393994808197, | |
| "learning_rate": 4.4672489082969426e-07, | |
| "loss": 2.1172, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.457569163817221, | |
| "grad_norm": 0.4727741777896881, | |
| "learning_rate": 4.4628820960698685e-07, | |
| "loss": 2.1343, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.46005595275101024, | |
| "grad_norm": 0.4549051523208618, | |
| "learning_rate": 4.458515283842795e-07, | |
| "loss": 2.1586, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.46254274168479953, | |
| "grad_norm": 0.4877924919128418, | |
| "learning_rate": 4.4541484716157203e-07, | |
| "loss": 2.2136, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.46502953061858876, | |
| "grad_norm": 0.47328630089759827, | |
| "learning_rate": 4.449781659388646e-07, | |
| "loss": 2.1065, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.467516319552378, | |
| "grad_norm": 0.46814873814582825, | |
| "learning_rate": 4.445414847161572e-07, | |
| "loss": 2.0879, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4700031084861672, | |
| "grad_norm": 0.46883970499038696, | |
| "learning_rate": 4.4410480349344976e-07, | |
| "loss": 2.1495, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4724898974199565, | |
| "grad_norm": 0.5020297169685364, | |
| "learning_rate": 4.4366812227074235e-07, | |
| "loss": 2.161, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.47497668635374574, | |
| "grad_norm": 0.47197675704956055, | |
| "learning_rate": 4.432314410480349e-07, | |
| "loss": 2.1354, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.47746347528753497, | |
| "grad_norm": 0.47488582134246826, | |
| "learning_rate": 4.427947598253275e-07, | |
| "loss": 2.0913, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4799502642213242, | |
| "grad_norm": 0.49505242705345154, | |
| "learning_rate": 4.4235807860262007e-07, | |
| "loss": 2.1499, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4824370531551135, | |
| "grad_norm": 0.48239609599113464, | |
| "learning_rate": 4.419213973799126e-07, | |
| "loss": 2.1432, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4849238420889027, | |
| "grad_norm": 0.46357694268226624, | |
| "learning_rate": 4.4148471615720525e-07, | |
| "loss": 2.1352, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.48741063102269194, | |
| "grad_norm": 0.4855436086654663, | |
| "learning_rate": 4.410480349344978e-07, | |
| "loss": 2.0876, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4898974199564812, | |
| "grad_norm": 0.5182051658630371, | |
| "learning_rate": 4.406113537117904e-07, | |
| "loss": 2.1872, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.49238420889027046, | |
| "grad_norm": 0.4700855016708374, | |
| "learning_rate": 4.40174672489083e-07, | |
| "loss": 2.0742, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4948709978240597, | |
| "grad_norm": 0.4741506278514862, | |
| "learning_rate": 4.397379912663755e-07, | |
| "loss": 2.0981, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.4973577867578489, | |
| "grad_norm": 0.4737417697906494, | |
| "learning_rate": 4.393013100436681e-07, | |
| "loss": 2.1206, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.49984457569163815, | |
| "grad_norm": 0.4860036075115204, | |
| "learning_rate": 4.388646288209607e-07, | |
| "loss": 2.0539, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5023313646254274, | |
| "grad_norm": 0.4895828068256378, | |
| "learning_rate": 4.3842794759825324e-07, | |
| "loss": 2.2017, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5048181535592167, | |
| "grad_norm": 0.5107592940330505, | |
| "learning_rate": 4.3799126637554583e-07, | |
| "loss": 2.2033, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.507304942493006, | |
| "grad_norm": 0.49359220266342163, | |
| "learning_rate": 4.3755458515283837e-07, | |
| "loss": 2.1245, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5097917314267951, | |
| "grad_norm": 0.43286237120628357, | |
| "learning_rate": 4.3711790393013096e-07, | |
| "loss": 2.1112, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5122785203605844, | |
| "grad_norm": 0.4412092864513397, | |
| "learning_rate": 4.366812227074236e-07, | |
| "loss": 2.032, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5147653092943736, | |
| "grad_norm": 0.5042114853858948, | |
| "learning_rate": 4.3624454148471614e-07, | |
| "loss": 2.1303, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5172520982281629, | |
| "grad_norm": 0.4746697247028351, | |
| "learning_rate": 4.3580786026200873e-07, | |
| "loss": 2.1337, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5197388871619522, | |
| "grad_norm": 0.4542432427406311, | |
| "learning_rate": 4.353711790393013e-07, | |
| "loss": 2.1181, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5222256760957413, | |
| "grad_norm": 0.5013236403465271, | |
| "learning_rate": 4.3493449781659386e-07, | |
| "loss": 2.1283, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5247124650295306, | |
| "grad_norm": 0.44694000482559204, | |
| "learning_rate": 4.3449781659388646e-07, | |
| "loss": 2.1317, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5271992539633199, | |
| "grad_norm": 0.5043014287948608, | |
| "learning_rate": 4.34061135371179e-07, | |
| "loss": 2.1751, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5296860428971091, | |
| "grad_norm": 0.45841050148010254, | |
| "learning_rate": 4.336244541484716e-07, | |
| "loss": 2.1164, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5321728318308984, | |
| "grad_norm": 0.4801214635372162, | |
| "learning_rate": 4.331877729257642e-07, | |
| "loss": 2.0855, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5346596207646876, | |
| "grad_norm": 0.5102494955062866, | |
| "learning_rate": 4.327510917030567e-07, | |
| "loss": 2.1819, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5371464096984768, | |
| "grad_norm": 0.5200817584991455, | |
| "learning_rate": 4.323144104803493e-07, | |
| "loss": 2.175, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5396331986322661, | |
| "grad_norm": 0.49951592087745667, | |
| "learning_rate": 4.318777292576419e-07, | |
| "loss": 2.1599, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5421199875660553, | |
| "grad_norm": 0.5104175209999084, | |
| "learning_rate": 4.314410480349345e-07, | |
| "loss": 2.1793, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5446067764998446, | |
| "grad_norm": 0.49184200167655945, | |
| "learning_rate": 4.310043668122271e-07, | |
| "loss": 2.1548, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5470935654336339, | |
| "grad_norm": 0.48827120661735535, | |
| "learning_rate": 4.305676855895196e-07, | |
| "loss": 2.1406, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.549580354367423, | |
| "grad_norm": 0.47386690974235535, | |
| "learning_rate": 4.301310043668122e-07, | |
| "loss": 2.1014, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5520671433012123, | |
| "grad_norm": 0.47124195098876953, | |
| "learning_rate": 4.296943231441048e-07, | |
| "loss": 2.086, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5545539322350016, | |
| "grad_norm": 0.5211581587791443, | |
| "learning_rate": 4.2925764192139734e-07, | |
| "loss": 2.0998, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5570407211687908, | |
| "grad_norm": 0.4680314064025879, | |
| "learning_rate": 4.2882096069868994e-07, | |
| "loss": 2.0983, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5595275101025801, | |
| "grad_norm": 0.4839833080768585, | |
| "learning_rate": 4.283842794759825e-07, | |
| "loss": 2.1144, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5620142990363692, | |
| "grad_norm": 0.4539274275302887, | |
| "learning_rate": 4.2794759825327507e-07, | |
| "loss": 2.0693, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5645010879701585, | |
| "grad_norm": 0.4671003818511963, | |
| "learning_rate": 4.275109170305677e-07, | |
| "loss": 2.0779, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5669878769039478, | |
| "grad_norm": 0.5179879069328308, | |
| "learning_rate": 4.2707423580786025e-07, | |
| "loss": 2.1674, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.569474665837737, | |
| "grad_norm": 0.4587318003177643, | |
| "learning_rate": 4.2663755458515284e-07, | |
| "loss": 2.1075, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5719614547715263, | |
| "grad_norm": 0.5107843279838562, | |
| "learning_rate": 4.262008733624454e-07, | |
| "loss": 2.1966, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5744482437053156, | |
| "grad_norm": 0.48125070333480835, | |
| "learning_rate": 4.2576419213973797e-07, | |
| "loss": 2.111, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5769350326391047, | |
| "grad_norm": 0.5291087031364441, | |
| "learning_rate": 4.2532751091703056e-07, | |
| "loss": 2.1401, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.579421821572894, | |
| "grad_norm": 0.5241518020629883, | |
| "learning_rate": 4.248908296943231e-07, | |
| "loss": 2.1662, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5819086105066833, | |
| "grad_norm": 0.5210862755775452, | |
| "learning_rate": 4.244541484716157e-07, | |
| "loss": 2.2149, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5843953994404725, | |
| "grad_norm": 0.5254886150360107, | |
| "learning_rate": 4.240174672489083e-07, | |
| "loss": 2.2019, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5868821883742618, | |
| "grad_norm": 0.49172264337539673, | |
| "learning_rate": 4.235807860262008e-07, | |
| "loss": 2.1671, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5893689773080509, | |
| "grad_norm": 0.47738420963287354, | |
| "learning_rate": 4.231441048034934e-07, | |
| "loss": 2.0794, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5918557662418402, | |
| "grad_norm": 0.5100018978118896, | |
| "learning_rate": 4.22707423580786e-07, | |
| "loss": 2.1181, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5943425551756295, | |
| "grad_norm": 0.5403950810432434, | |
| "learning_rate": 4.222707423580786e-07, | |
| "loss": 2.1559, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5968293441094187, | |
| "grad_norm": 0.5101498961448669, | |
| "learning_rate": 4.218340611353712e-07, | |
| "loss": 2.2091, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.599316133043208, | |
| "grad_norm": 0.5162122845649719, | |
| "learning_rate": 4.2139737991266373e-07, | |
| "loss": 2.1471, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6018029219769973, | |
| "grad_norm": 0.5074677467346191, | |
| "learning_rate": 4.209606986899563e-07, | |
| "loss": 2.0849, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6042897109107864, | |
| "grad_norm": 0.5043840408325195, | |
| "learning_rate": 4.205240174672489e-07, | |
| "loss": 2.109, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6067764998445757, | |
| "grad_norm": 0.49023503065109253, | |
| "learning_rate": 4.2008733624454145e-07, | |
| "loss": 2.157, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6092632887783649, | |
| "grad_norm": 0.5031821131706238, | |
| "learning_rate": 4.1965065502183404e-07, | |
| "loss": 2.1597, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6117500777121542, | |
| "grad_norm": 0.5147417783737183, | |
| "learning_rate": 4.192139737991266e-07, | |
| "loss": 2.1502, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6142368666459435, | |
| "grad_norm": 0.5135524272918701, | |
| "learning_rate": 4.187772925764192e-07, | |
| "loss": 2.1204, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6167236555797326, | |
| "grad_norm": 0.516242265701294, | |
| "learning_rate": 4.1834061135371177e-07, | |
| "loss": 2.1121, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6192104445135219, | |
| "grad_norm": 0.5270472764968872, | |
| "learning_rate": 4.1790393013100436e-07, | |
| "loss": 2.1296, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6216972334473112, | |
| "grad_norm": 0.5023481249809265, | |
| "learning_rate": 4.1746724890829695e-07, | |
| "loss": 2.1138, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6241840223811004, | |
| "grad_norm": 0.5072234869003296, | |
| "learning_rate": 4.170305676855895e-07, | |
| "loss": 2.1852, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6266708113148897, | |
| "grad_norm": 0.5261276364326477, | |
| "learning_rate": 4.165938864628821e-07, | |
| "loss": 2.1707, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6291576002486788, | |
| "grad_norm": 0.524861216545105, | |
| "learning_rate": 4.1615720524017467e-07, | |
| "loss": 2.1885, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6316443891824681, | |
| "grad_norm": 0.5122174620628357, | |
| "learning_rate": 4.157205240174672e-07, | |
| "loss": 2.1095, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6341311781162574, | |
| "grad_norm": 0.4976103603839874, | |
| "learning_rate": 4.152838427947598e-07, | |
| "loss": 2.1686, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6366179670500466, | |
| "grad_norm": 0.48984527587890625, | |
| "learning_rate": 4.148471615720524e-07, | |
| "loss": 2.0661, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6391047559838359, | |
| "grad_norm": 0.5184794664382935, | |
| "learning_rate": 4.1441048034934493e-07, | |
| "loss": 2.1323, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6415915449176252, | |
| "grad_norm": 0.4991200566291809, | |
| "learning_rate": 4.139737991266375e-07, | |
| "loss": 2.1294, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6440783338514143, | |
| "grad_norm": 0.5302152037620544, | |
| "learning_rate": 4.1353711790393006e-07, | |
| "loss": 2.0922, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6465651227852036, | |
| "grad_norm": 0.5143322348594666, | |
| "learning_rate": 4.131004366812227e-07, | |
| "loss": 2.1694, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6490519117189929, | |
| "grad_norm": 0.5043548941612244, | |
| "learning_rate": 4.126637554585153e-07, | |
| "loss": 2.0899, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6515387006527821, | |
| "grad_norm": 0.5160046815872192, | |
| "learning_rate": 4.1222707423580784e-07, | |
| "loss": 2.1185, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6540254895865714, | |
| "grad_norm": 0.5054792761802673, | |
| "learning_rate": 4.1179039301310043e-07, | |
| "loss": 2.1503, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6565122785203605, | |
| "grad_norm": 0.5056222677230835, | |
| "learning_rate": 4.11353711790393e-07, | |
| "loss": 2.1089, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6589990674541498, | |
| "grad_norm": 0.5285047292709351, | |
| "learning_rate": 4.1091703056768556e-07, | |
| "loss": 2.1556, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6614858563879391, | |
| "grad_norm": 0.48898041248321533, | |
| "learning_rate": 4.1048034934497815e-07, | |
| "loss": 2.1107, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6639726453217283, | |
| "grad_norm": 0.525590717792511, | |
| "learning_rate": 4.100436681222707e-07, | |
| "loss": 2.1374, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6664594342555176, | |
| "grad_norm": 0.5363737344741821, | |
| "learning_rate": 4.096069868995633e-07, | |
| "loss": 2.1198, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6689462231893069, | |
| "grad_norm": 0.5182633399963379, | |
| "learning_rate": 4.091703056768559e-07, | |
| "loss": 2.126, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.671433012123096, | |
| "grad_norm": 0.4978923201560974, | |
| "learning_rate": 4.0873362445414847e-07, | |
| "loss": 2.0764, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6739198010568853, | |
| "grad_norm": 0.5094720125198364, | |
| "learning_rate": 4.0829694323144106e-07, | |
| "loss": 2.145, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6764065899906745, | |
| "grad_norm": 0.540023148059845, | |
| "learning_rate": 4.078602620087336e-07, | |
| "loss": 2.1327, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6788933789244638, | |
| "grad_norm": 0.5420276522636414, | |
| "learning_rate": 4.074235807860262e-07, | |
| "loss": 2.1707, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6813801678582531, | |
| "grad_norm": 0.5282043218612671, | |
| "learning_rate": 4.069868995633188e-07, | |
| "loss": 2.1473, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6838669567920422, | |
| "grad_norm": 0.5049037933349609, | |
| "learning_rate": 4.065502183406113e-07, | |
| "loss": 2.1335, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6863537457258315, | |
| "grad_norm": 0.5107303261756897, | |
| "learning_rate": 4.061135371179039e-07, | |
| "loss": 2.1349, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6888405346596208, | |
| "grad_norm": 0.4959608018398285, | |
| "learning_rate": 4.056768558951965e-07, | |
| "loss": 2.1044, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.69132732359341, | |
| "grad_norm": 0.5125852227210999, | |
| "learning_rate": 4.0524017467248904e-07, | |
| "loss": 2.1428, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6938141125271993, | |
| "grad_norm": 0.511873185634613, | |
| "learning_rate": 4.0480349344978163e-07, | |
| "loss": 2.0763, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6963009014609886, | |
| "grad_norm": 0.5032888054847717, | |
| "learning_rate": 4.0436681222707417e-07, | |
| "loss": 2.122, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6987876903947777, | |
| "grad_norm": 0.5102598667144775, | |
| "learning_rate": 4.039301310043668e-07, | |
| "loss": 2.0693, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.701274479328567, | |
| "grad_norm": 0.5118304491043091, | |
| "learning_rate": 4.034934497816594e-07, | |
| "loss": 2.11, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7037612682623562, | |
| "grad_norm": 0.5202342867851257, | |
| "learning_rate": 4.0305676855895195e-07, | |
| "loss": 2.1582, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7062480571961455, | |
| "grad_norm": 0.48433917760849, | |
| "learning_rate": 4.0262008733624454e-07, | |
| "loss": 2.103, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7087348461299348, | |
| "grad_norm": 0.4986036717891693, | |
| "learning_rate": 4.0218340611353713e-07, | |
| "loss": 2.1203, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7112216350637239, | |
| "grad_norm": 0.5467602014541626, | |
| "learning_rate": 4.0174672489082967e-07, | |
| "loss": 2.1204, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7137084239975132, | |
| "grad_norm": 0.5108657479286194, | |
| "learning_rate": 4.0131004366812226e-07, | |
| "loss": 2.1478, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7161952129313025, | |
| "grad_norm": 0.5145993232727051, | |
| "learning_rate": 4.008733624454148e-07, | |
| "loss": 2.1563, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7186820018650917, | |
| "grad_norm": 0.5134692788124084, | |
| "learning_rate": 4.004366812227074e-07, | |
| "loss": 2.1067, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.721168790798881, | |
| "grad_norm": 0.5436774492263794, | |
| "learning_rate": 4e-07, | |
| "loss": 2.1369, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7236555797326701, | |
| "grad_norm": 0.5296205282211304, | |
| "learning_rate": 3.995633187772925e-07, | |
| "loss": 2.1452, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7261423686664594, | |
| "grad_norm": 0.4911108911037445, | |
| "learning_rate": 3.9912663755458517e-07, | |
| "loss": 2.1279, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7286291576002487, | |
| "grad_norm": 0.5625902414321899, | |
| "learning_rate": 3.986899563318777e-07, | |
| "loss": 2.169, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7311159465340379, | |
| "grad_norm": 0.5042857527732849, | |
| "learning_rate": 3.982532751091703e-07, | |
| "loss": 2.0692, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7336027354678272, | |
| "grad_norm": 0.5251498222351074, | |
| "learning_rate": 3.978165938864629e-07, | |
| "loss": 2.1261, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7360895244016165, | |
| "grad_norm": 0.5093502402305603, | |
| "learning_rate": 3.973799126637554e-07, | |
| "loss": 2.1175, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7385763133354056, | |
| "grad_norm": 0.49675241112709045, | |
| "learning_rate": 3.96943231441048e-07, | |
| "loss": 2.1679, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7410631022691949, | |
| "grad_norm": 0.523313045501709, | |
| "learning_rate": 3.965065502183406e-07, | |
| "loss": 2.1195, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7435498912029841, | |
| "grad_norm": 0.5194100737571716, | |
| "learning_rate": 3.9606986899563315e-07, | |
| "loss": 2.1431, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7460366801367734, | |
| "grad_norm": 0.5145063996315002, | |
| "learning_rate": 3.9563318777292574e-07, | |
| "loss": 2.0968, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7485234690705627, | |
| "grad_norm": 0.5165944695472717, | |
| "learning_rate": 3.951965065502183e-07, | |
| "loss": 2.1316, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7510102580043518, | |
| "grad_norm": 0.5502745509147644, | |
| "learning_rate": 3.947598253275109e-07, | |
| "loss": 2.1215, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7534970469381411, | |
| "grad_norm": 0.5752532482147217, | |
| "learning_rate": 3.943231441048035e-07, | |
| "loss": 2.165, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7559838358719304, | |
| "grad_norm": 0.5388760566711426, | |
| "learning_rate": 3.9388646288209605e-07, | |
| "loss": 2.136, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.7584706248057196, | |
| "grad_norm": 0.5256951451301575, | |
| "learning_rate": 3.9344978165938865e-07, | |
| "loss": 2.0978, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7609574137395089, | |
| "grad_norm": 0.5247829556465149, | |
| "learning_rate": 3.930131004366812e-07, | |
| "loss": 2.1591, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7634442026732982, | |
| "grad_norm": 0.508574366569519, | |
| "learning_rate": 3.925764192139738e-07, | |
| "loss": 2.1542, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7659309916070873, | |
| "grad_norm": 0.49671751260757446, | |
| "learning_rate": 3.9213973799126637e-07, | |
| "loss": 2.1469, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7684177805408766, | |
| "grad_norm": 0.5673956274986267, | |
| "learning_rate": 3.917030567685589e-07, | |
| "loss": 2.2209, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7709045694746658, | |
| "grad_norm": 0.4881182610988617, | |
| "learning_rate": 3.912663755458515e-07, | |
| "loss": 2.1353, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7733913584084551, | |
| "grad_norm": 0.5374391078948975, | |
| "learning_rate": 3.908296943231441e-07, | |
| "loss": 2.2102, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7758781473422444, | |
| "grad_norm": 0.520723283290863, | |
| "learning_rate": 3.9039301310043663e-07, | |
| "loss": 2.1716, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7783649362760335, | |
| "grad_norm": 0.5542478561401367, | |
| "learning_rate": 3.8995633187772927e-07, | |
| "loss": 2.2128, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7808517252098228, | |
| "grad_norm": 0.5180374979972839, | |
| "learning_rate": 3.895196506550218e-07, | |
| "loss": 2.1245, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7833385141436121, | |
| "grad_norm": 0.5454829931259155, | |
| "learning_rate": 3.890829694323144e-07, | |
| "loss": 2.1609, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7858253030774013, | |
| "grad_norm": 0.568573534488678, | |
| "learning_rate": 3.88646288209607e-07, | |
| "loss": 2.175, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7883120920111906, | |
| "grad_norm": 0.5162298679351807, | |
| "learning_rate": 3.8820960698689953e-07, | |
| "loss": 2.1474, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7907988809449797, | |
| "grad_norm": 0.5148350596427917, | |
| "learning_rate": 3.877729257641921e-07, | |
| "loss": 2.0697, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.793285669878769, | |
| "grad_norm": 0.5210283994674683, | |
| "learning_rate": 3.873362445414847e-07, | |
| "loss": 2.1587, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7957724588125583, | |
| "grad_norm": 0.4845898151397705, | |
| "learning_rate": 3.8689956331877726e-07, | |
| "loss": 2.0581, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7982592477463475, | |
| "grad_norm": 0.5222198367118835, | |
| "learning_rate": 3.8646288209606985e-07, | |
| "loss": 2.0504, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8007460366801368, | |
| "grad_norm": 0.5040515065193176, | |
| "learning_rate": 3.860262008733624e-07, | |
| "loss": 2.0898, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8032328256139261, | |
| "grad_norm": 0.501930296421051, | |
| "learning_rate": 3.85589519650655e-07, | |
| "loss": 2.1335, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.8057196145477152, | |
| "grad_norm": 0.5378695726394653, | |
| "learning_rate": 3.851528384279476e-07, | |
| "loss": 2.0953, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8082064034815045, | |
| "grad_norm": 0.49689194560050964, | |
| "learning_rate": 3.8471615720524016e-07, | |
| "loss": 2.0837, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8106931924152938, | |
| "grad_norm": 0.5431040525436401, | |
| "learning_rate": 3.8427947598253275e-07, | |
| "loss": 2.1151, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.813179981349083, | |
| "grad_norm": 0.510339081287384, | |
| "learning_rate": 3.838427947598253e-07, | |
| "loss": 2.1399, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8156667702828723, | |
| "grad_norm": 0.5451592206954956, | |
| "learning_rate": 3.834061135371179e-07, | |
| "loss": 2.1182, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8181535592166614, | |
| "grad_norm": 0.5272311568260193, | |
| "learning_rate": 3.829694323144105e-07, | |
| "loss": 2.139, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8206403481504507, | |
| "grad_norm": 0.5389718413352966, | |
| "learning_rate": 3.82532751091703e-07, | |
| "loss": 2.1086, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.82312713708424, | |
| "grad_norm": 0.5228806138038635, | |
| "learning_rate": 3.820960698689956e-07, | |
| "loss": 2.1435, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8256139260180292, | |
| "grad_norm": 0.5186501741409302, | |
| "learning_rate": 3.816593886462882e-07, | |
| "loss": 2.095, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8281007149518185, | |
| "grad_norm": 0.5484049320220947, | |
| "learning_rate": 3.8122270742358074e-07, | |
| "loss": 2.1666, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8305875038856078, | |
| "grad_norm": 0.5375939607620239, | |
| "learning_rate": 3.807860262008734e-07, | |
| "loss": 2.125, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8330742928193969, | |
| "grad_norm": 0.5139255523681641, | |
| "learning_rate": 3.803493449781659e-07, | |
| "loss": 2.075, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8355610817531862, | |
| "grad_norm": 0.5235688090324402, | |
| "learning_rate": 3.799126637554585e-07, | |
| "loss": 2.0731, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8380478706869754, | |
| "grad_norm": 0.5630027651786804, | |
| "learning_rate": 3.794759825327511e-07, | |
| "loss": 2.1261, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8405346596207647, | |
| "grad_norm": 0.547572910785675, | |
| "learning_rate": 3.7903930131004364e-07, | |
| "loss": 2.1641, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.843021448554554, | |
| "grad_norm": 0.5533425211906433, | |
| "learning_rate": 3.7860262008733623e-07, | |
| "loss": 2.1293, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8455082374883431, | |
| "grad_norm": 0.5256425738334656, | |
| "learning_rate": 3.781659388646288e-07, | |
| "loss": 2.1397, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8479950264221324, | |
| "grad_norm": 0.5411325693130493, | |
| "learning_rate": 3.7772925764192136e-07, | |
| "loss": 2.1631, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8504818153559217, | |
| "grad_norm": 0.5244682431221008, | |
| "learning_rate": 3.7729257641921396e-07, | |
| "loss": 2.1655, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.8529686042897109, | |
| "grad_norm": 0.5120859742164612, | |
| "learning_rate": 3.768558951965065e-07, | |
| "loss": 2.0929, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8554553932235002, | |
| "grad_norm": 0.5486117601394653, | |
| "learning_rate": 3.764192139737991e-07, | |
| "loss": 2.1333, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8579421821572893, | |
| "grad_norm": 0.5485012531280518, | |
| "learning_rate": 3.7598253275109173e-07, | |
| "loss": 2.1497, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8604289710910786, | |
| "grad_norm": 0.5423093438148499, | |
| "learning_rate": 3.7554585152838427e-07, | |
| "loss": 2.2169, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.8629157600248679, | |
| "grad_norm": 0.5003622770309448, | |
| "learning_rate": 3.7510917030567686e-07, | |
| "loss": 2.1818, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.8654025489586571, | |
| "grad_norm": 0.5931081771850586, | |
| "learning_rate": 3.746724890829694e-07, | |
| "loss": 2.1631, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8678893378924464, | |
| "grad_norm": 0.5221492052078247, | |
| "learning_rate": 3.74235807860262e-07, | |
| "loss": 2.1087, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.8703761268262357, | |
| "grad_norm": 0.5065641403198242, | |
| "learning_rate": 3.737991266375546e-07, | |
| "loss": 2.0622, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8728629157600248, | |
| "grad_norm": 0.5329532027244568, | |
| "learning_rate": 3.733624454148471e-07, | |
| "loss": 2.1275, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8753497046938141, | |
| "grad_norm": 0.5383079648017883, | |
| "learning_rate": 3.729257641921397e-07, | |
| "loss": 2.1342, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8778364936276034, | |
| "grad_norm": 0.49477216601371765, | |
| "learning_rate": 3.724890829694323e-07, | |
| "loss": 2.0404, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8803232825613926, | |
| "grad_norm": 0.5197799205780029, | |
| "learning_rate": 3.7205240174672484e-07, | |
| "loss": 2.1228, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8828100714951819, | |
| "grad_norm": 0.5122123956680298, | |
| "learning_rate": 3.7161572052401744e-07, | |
| "loss": 2.1329, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.885296860428971, | |
| "grad_norm": 0.5379232168197632, | |
| "learning_rate": 3.7117903930131003e-07, | |
| "loss": 2.0743, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8877836493627603, | |
| "grad_norm": 0.5164668560028076, | |
| "learning_rate": 3.707423580786026e-07, | |
| "loss": 2.1474, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8902704382965496, | |
| "grad_norm": 0.518368661403656, | |
| "learning_rate": 3.703056768558952e-07, | |
| "loss": 2.1987, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8927572272303388, | |
| "grad_norm": 0.5662968754768372, | |
| "learning_rate": 3.6986899563318775e-07, | |
| "loss": 2.1301, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8952440161641281, | |
| "grad_norm": 0.5161558389663696, | |
| "learning_rate": 3.6943231441048034e-07, | |
| "loss": 2.1033, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8977308050979174, | |
| "grad_norm": 0.5516855120658875, | |
| "learning_rate": 3.6899563318777293e-07, | |
| "loss": 2.1003, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9002175940317065, | |
| "grad_norm": 0.5291304588317871, | |
| "learning_rate": 3.6855895196506547e-07, | |
| "loss": 2.0533, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9027043829654958, | |
| "grad_norm": 0.5586827397346497, | |
| "learning_rate": 3.6812227074235806e-07, | |
| "loss": 2.1052, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.905191171899285, | |
| "grad_norm": 0.5328514575958252, | |
| "learning_rate": 3.676855895196506e-07, | |
| "loss": 2.1548, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9076779608330743, | |
| "grad_norm": 0.5259972810745239, | |
| "learning_rate": 3.672489082969432e-07, | |
| "loss": 2.1101, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9101647497668636, | |
| "grad_norm": 0.5482295751571655, | |
| "learning_rate": 3.6681222707423584e-07, | |
| "loss": 2.1678, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9126515387006527, | |
| "grad_norm": 0.5381218194961548, | |
| "learning_rate": 3.663755458515284e-07, | |
| "loss": 2.2098, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.915138327634442, | |
| "grad_norm": 0.5494764447212219, | |
| "learning_rate": 3.6593886462882097e-07, | |
| "loss": 2.1338, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9176251165682313, | |
| "grad_norm": 0.5393621921539307, | |
| "learning_rate": 3.655021834061135e-07, | |
| "loss": 2.0952, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9201119055020205, | |
| "grad_norm": 0.5395556092262268, | |
| "learning_rate": 3.650655021834061e-07, | |
| "loss": 2.1402, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9225986944358098, | |
| "grad_norm": 0.5069707632064819, | |
| "learning_rate": 3.646288209606987e-07, | |
| "loss": 2.0925, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9250854833695991, | |
| "grad_norm": 0.5580669641494751, | |
| "learning_rate": 3.6419213973799123e-07, | |
| "loss": 2.1585, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9275722723033882, | |
| "grad_norm": 0.5407446026802063, | |
| "learning_rate": 3.637554585152838e-07, | |
| "loss": 2.1448, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9300590612371775, | |
| "grad_norm": 0.5261268019676208, | |
| "learning_rate": 3.633187772925764e-07, | |
| "loss": 2.1687, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9325458501709667, | |
| "grad_norm": 0.5728645920753479, | |
| "learning_rate": 3.6288209606986895e-07, | |
| "loss": 2.0929, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.935032639104756, | |
| "grad_norm": 0.536983072757721, | |
| "learning_rate": 3.6244541484716154e-07, | |
| "loss": 2.1669, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9375194280385453, | |
| "grad_norm": 0.5492017269134521, | |
| "learning_rate": 3.6200873362445414e-07, | |
| "loss": 2.1449, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9400062169723344, | |
| "grad_norm": 0.5745022296905518, | |
| "learning_rate": 3.6157205240174673e-07, | |
| "loss": 2.1315, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9424930059061237, | |
| "grad_norm": 0.5852669477462769, | |
| "learning_rate": 3.611353711790393e-07, | |
| "loss": 2.1405, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.944979794839913, | |
| "grad_norm": 0.5169341564178467, | |
| "learning_rate": 3.6069868995633186e-07, | |
| "loss": 2.0788, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9474665837737022, | |
| "grad_norm": 0.5499164462089539, | |
| "learning_rate": 3.6026200873362445e-07, | |
| "loss": 2.1843, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.9499533727074915, | |
| "grad_norm": 0.5195809006690979, | |
| "learning_rate": 3.59825327510917e-07, | |
| "loss": 2.1045, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9524401616412806, | |
| "grad_norm": 0.5368107557296753, | |
| "learning_rate": 3.593886462882096e-07, | |
| "loss": 2.1261, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9549269505750699, | |
| "grad_norm": 0.5721762776374817, | |
| "learning_rate": 3.5895196506550217e-07, | |
| "loss": 2.1323, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.9574137395088592, | |
| "grad_norm": 0.5255040526390076, | |
| "learning_rate": 3.585152838427947e-07, | |
| "loss": 2.13, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9599005284426484, | |
| "grad_norm": 0.5373786687850952, | |
| "learning_rate": 3.580786026200873e-07, | |
| "loss": 2.0763, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9623873173764377, | |
| "grad_norm": 0.5432249307632446, | |
| "learning_rate": 3.576419213973799e-07, | |
| "loss": 2.1305, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.964874106310227, | |
| "grad_norm": 0.5505443811416626, | |
| "learning_rate": 3.572052401746725e-07, | |
| "loss": 2.1311, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.9673608952440161, | |
| "grad_norm": 0.5119839906692505, | |
| "learning_rate": 3.567685589519651e-07, | |
| "loss": 2.1121, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9698476841778054, | |
| "grad_norm": 0.5414577126502991, | |
| "learning_rate": 3.563318777292576e-07, | |
| "loss": 2.1076, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9723344731115946, | |
| "grad_norm": 0.5283794403076172, | |
| "learning_rate": 3.558951965065502e-07, | |
| "loss": 2.1293, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.9748212620453839, | |
| "grad_norm": 0.5475645065307617, | |
| "learning_rate": 3.554585152838428e-07, | |
| "loss": 2.1373, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.9773080509791732, | |
| "grad_norm": 0.5172975063323975, | |
| "learning_rate": 3.5502183406113534e-07, | |
| "loss": 2.1378, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.9797948399129623, | |
| "grad_norm": 0.5674493312835693, | |
| "learning_rate": 3.5458515283842793e-07, | |
| "loss": 2.0797, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.9822816288467516, | |
| "grad_norm": 0.510979950428009, | |
| "learning_rate": 3.541484716157205e-07, | |
| "loss": 2.091, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9847684177805409, | |
| "grad_norm": 0.5517850518226624, | |
| "learning_rate": 3.5371179039301306e-07, | |
| "loss": 2.1703, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9872552067143301, | |
| "grad_norm": 0.5487313270568848, | |
| "learning_rate": 3.5327510917030565e-07, | |
| "loss": 2.1213, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9897419956481194, | |
| "grad_norm": 0.5256079435348511, | |
| "learning_rate": 3.528384279475982e-07, | |
| "loss": 2.1052, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9922287845819087, | |
| "grad_norm": 0.5553068518638611, | |
| "learning_rate": 3.5240174672489084e-07, | |
| "loss": 2.0818, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9947155735156978, | |
| "grad_norm": 0.5434982180595398, | |
| "learning_rate": 3.5196506550218343e-07, | |
| "loss": 2.0612, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9972023624494871, | |
| "grad_norm": 0.5237376689910889, | |
| "learning_rate": 3.5152838427947597e-07, | |
| "loss": 2.1489, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.9996891513832763, | |
| "grad_norm": 0.5455615520477295, | |
| "learning_rate": 3.5109170305676856e-07, | |
| "loss": 2.0467, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9996891513832763, | |
| "eval_loss": 2.129138708114624, | |
| "eval_runtime": 458.2461, | |
| "eval_samples_per_second": 1.065, | |
| "eval_steps_per_second": 0.266, | |
| "step": 402 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1206, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1872324383890473e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |