{ "best_metric": 0.007892865687608719, "best_model_checkpoint": "/home/paperspace/Data/models/spinny/llm3br256/checkpoint-250", "epoch": 4.971098265895954, "eval_steps": 5, "global_step": 430, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.011560693641618497, "grad_norm": 0.15129172801971436, "learning_rate": 2.325581395348837e-06, "loss": 0.068, "step": 1 }, { "epoch": 0.023121387283236993, "grad_norm": 0.14174357056617737, "learning_rate": 4.651162790697674e-06, "loss": 0.0635, "step": 2 }, { "epoch": 0.03468208092485549, "grad_norm": 0.14430448412895203, "learning_rate": 6.976744186046512e-06, "loss": 0.0649, "step": 3 }, { "epoch": 0.046242774566473986, "grad_norm": 0.1410493552684784, "learning_rate": 9.302325581395349e-06, "loss": 0.063, "step": 4 }, { "epoch": 0.057803468208092484, "grad_norm": 0.13995003700256348, "learning_rate": 1.1627906976744187e-05, "loss": 0.0598, "step": 5 }, { "epoch": 0.057803468208092484, "eval_loss": 0.05565303936600685, "eval_runtime": 11.8576, "eval_samples_per_second": 4.217, "eval_steps_per_second": 1.096, "step": 5 }, { "epoch": 0.06936416184971098, "grad_norm": 0.12226737290620804, "learning_rate": 1.3953488372093024e-05, "loss": 0.0533, "step": 6 }, { "epoch": 0.08092485549132948, "grad_norm": 0.10596886277198792, "learning_rate": 1.6279069767441862e-05, "loss": 0.0544, "step": 7 }, { "epoch": 0.09248554913294797, "grad_norm": 0.07781537622213364, "learning_rate": 1.8604651162790697e-05, "loss": 0.0448, "step": 8 }, { "epoch": 0.10404624277456648, "grad_norm": 0.06554879248142242, "learning_rate": 2.0930232558139536e-05, "loss": 0.0373, "step": 9 }, { "epoch": 0.11560693641618497, "grad_norm": 0.08980869501829147, "learning_rate": 2.3255813953488374e-05, "loss": 0.0388, "step": 10 }, { "epoch": 0.11560693641618497, "eval_loss": 0.034586433321237564, "eval_runtime": 10.0247, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 10 }, { "epoch": 0.12716763005780346, "grad_norm": 0.08015415817499161, "learning_rate": 2.5581395348837212e-05, "loss": 0.0393, "step": 11 }, { "epoch": 0.13872832369942195, "grad_norm": 0.06148159131407738, "learning_rate": 2.7906976744186048e-05, "loss": 0.0377, "step": 12 }, { "epoch": 0.15028901734104047, "grad_norm": 0.04883963614702225, "learning_rate": 3.0232558139534883e-05, "loss": 0.0356, "step": 13 }, { "epoch": 0.16184971098265896, "grad_norm": 0.04954546317458153, "learning_rate": 3.2558139534883724e-05, "loss": 0.0363, "step": 14 }, { "epoch": 0.17341040462427745, "grad_norm": 0.041509971022605896, "learning_rate": 3.488372093023256e-05, "loss": 0.0275, "step": 15 }, { "epoch": 0.17341040462427745, "eval_loss": 0.027632026001811028, "eval_runtime": 10.0235, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 15 }, { "epoch": 0.18497109826589594, "grad_norm": 0.041887976229190826, "learning_rate": 3.7209302325581394e-05, "loss": 0.0303, "step": 16 }, { "epoch": 0.19653179190751446, "grad_norm": 0.03965295851230621, "learning_rate": 3.953488372093023e-05, "loss": 0.0284, "step": 17 }, { "epoch": 0.20809248554913296, "grad_norm": 0.03458382189273834, "learning_rate": 4.186046511627907e-05, "loss": 0.03, "step": 18 }, { "epoch": 0.21965317919075145, "grad_norm": 0.03138720989227295, "learning_rate": 4.418604651162791e-05, "loss": 0.0242, "step": 19 }, { "epoch": 0.23121387283236994, "grad_norm": 0.02681548520922661, "learning_rate": 4.651162790697675e-05, "loss": 0.0218, "step": 20 }, { "epoch": 0.23121387283236994, "eval_loss": 0.02280299924314022, "eval_runtime": 10.0675, "eval_samples_per_second": 4.967, "eval_steps_per_second": 1.291, "step": 20 }, { "epoch": 0.24277456647398843, "grad_norm": 0.03174450621008873, "learning_rate": 4.883720930232558e-05, "loss": 0.0235, "step": 21 }, { "epoch": 0.2543352601156069, "grad_norm": 0.033451005816459656, "learning_rate": 5.1162790697674425e-05, "loss": 0.0291, "step": 22 }, { "epoch": 0.2658959537572254, "grad_norm": 0.03227928280830383, "learning_rate": 5.348837209302326e-05, "loss": 0.0254, "step": 23 }, { "epoch": 0.2774566473988439, "grad_norm": 0.03783709928393364, "learning_rate": 5.5813953488372095e-05, "loss": 0.0256, "step": 24 }, { "epoch": 0.28901734104046245, "grad_norm": 0.03228365629911423, "learning_rate": 5.8139534883720933e-05, "loss": 0.0236, "step": 25 }, { "epoch": 0.28901734104046245, "eval_loss": 0.020257851108908653, "eval_runtime": 10.0419, "eval_samples_per_second": 4.979, "eval_steps_per_second": 1.295, "step": 25 }, { "epoch": 0.30057803468208094, "grad_norm": 0.02823326177895069, "learning_rate": 6.0465116279069765e-05, "loss": 0.0229, "step": 26 }, { "epoch": 0.31213872832369943, "grad_norm": 0.02882234752178192, "learning_rate": 6.27906976744186e-05, "loss": 0.0226, "step": 27 }, { "epoch": 0.3236994219653179, "grad_norm": 0.02527730166912079, "learning_rate": 6.511627906976745e-05, "loss": 0.0196, "step": 28 }, { "epoch": 0.3352601156069364, "grad_norm": 0.026926985010504723, "learning_rate": 6.744186046511628e-05, "loss": 0.0182, "step": 29 }, { "epoch": 0.3468208092485549, "grad_norm": 0.02383883111178875, "learning_rate": 6.976744186046513e-05, "loss": 0.0182, "step": 30 }, { "epoch": 0.3468208092485549, "eval_loss": 0.01791333220899105, "eval_runtime": 10.1038, "eval_samples_per_second": 4.949, "eval_steps_per_second": 1.287, "step": 30 }, { "epoch": 0.3583815028901734, "grad_norm": 0.020905988290905952, "learning_rate": 7.209302325581396e-05, "loss": 0.0173, "step": 31 }, { "epoch": 0.3699421965317919, "grad_norm": 0.021837348118424416, "learning_rate": 7.441860465116279e-05, "loss": 0.0186, "step": 32 }, { "epoch": 0.3815028901734104, "grad_norm": 0.022534186020493507, "learning_rate": 7.674418604651163e-05, "loss": 0.0192, "step": 33 }, { "epoch": 0.3930635838150289, "grad_norm": 0.021404536440968513, "learning_rate": 7.906976744186047e-05, "loss": 0.0163, "step": 34 }, { "epoch": 0.4046242774566474, "grad_norm": 0.02087543159723282, "learning_rate": 8.139534883720931e-05, "loss": 0.019, "step": 35 }, { "epoch": 0.4046242774566474, "eval_loss": 0.016155000776052475, "eval_runtime": 10.0056, "eval_samples_per_second": 4.997, "eval_steps_per_second": 1.299, "step": 35 }, { "epoch": 0.4161849710982659, "grad_norm": 0.023041941225528717, "learning_rate": 8.372093023255814e-05, "loss": 0.0157, "step": 36 }, { "epoch": 0.4277456647398844, "grad_norm": 0.020684899762272835, "learning_rate": 8.604651162790697e-05, "loss": 0.0154, "step": 37 }, { "epoch": 0.4393063583815029, "grad_norm": 0.019204530864953995, "learning_rate": 8.837209302325582e-05, "loss": 0.0142, "step": 38 }, { "epoch": 0.4508670520231214, "grad_norm": 0.019893651828169823, "learning_rate": 9.069767441860465e-05, "loss": 0.0155, "step": 39 }, { "epoch": 0.4624277456647399, "grad_norm": 0.022773414850234985, "learning_rate": 9.30232558139535e-05, "loss": 0.017, "step": 40 }, { "epoch": 0.4624277456647399, "eval_loss": 0.01472202967852354, "eval_runtime": 10.018, "eval_samples_per_second": 4.991, "eval_steps_per_second": 1.298, "step": 40 }, { "epoch": 0.47398843930635837, "grad_norm": 0.021773461252450943, "learning_rate": 9.534883720930233e-05, "loss": 0.0161, "step": 41 }, { "epoch": 0.48554913294797686, "grad_norm": 0.023095447570085526, "learning_rate": 9.767441860465116e-05, "loss": 0.0146, "step": 42 }, { "epoch": 0.49710982658959535, "grad_norm": 0.0202298853546381, "learning_rate": 0.0001, "loss": 0.015, "step": 43 }, { "epoch": 0.5086705202312138, "grad_norm": 0.01809552311897278, "learning_rate": 9.999835253787473e-05, "loss": 0.0149, "step": 44 }, { "epoch": 0.5202312138728323, "grad_norm": 0.019870778545737267, "learning_rate": 9.999341026006419e-05, "loss": 0.0147, "step": 45 }, { "epoch": 0.5202312138728323, "eval_loss": 0.013716931454837322, "eval_runtime": 10.0201, "eval_samples_per_second": 4.99, "eval_steps_per_second": 1.297, "step": 45 }, { "epoch": 0.5317919075144508, "grad_norm": 0.017238808795809746, "learning_rate": 9.998517349225698e-05, "loss": 0.0107, "step": 46 }, { "epoch": 0.5433526011560693, "grad_norm": 0.022697314620018005, "learning_rate": 9.997364277724361e-05, "loss": 0.0149, "step": 47 }, { "epoch": 0.5549132947976878, "grad_norm": 0.023472866043448448, "learning_rate": 9.99588188748808e-05, "loss": 0.0144, "step": 48 }, { "epoch": 0.5664739884393064, "grad_norm": 0.017129473388195038, "learning_rate": 9.994070276204116e-05, "loss": 0.0126, "step": 49 }, { "epoch": 0.5780346820809249, "grad_norm": 0.01756882295012474, "learning_rate": 9.991929563254914e-05, "loss": 0.0118, "step": 50 }, { "epoch": 0.5780346820809249, "eval_loss": 0.013151152059435844, "eval_runtime": 10.0009, "eval_samples_per_second": 5.0, "eval_steps_per_second": 1.3, "step": 50 }, { "epoch": 0.5895953757225434, "grad_norm": 0.019612474367022514, "learning_rate": 9.989459889710213e-05, "loss": 0.0144, "step": 51 }, { "epoch": 0.6011560693641619, "grad_norm": 0.016709089279174805, "learning_rate": 9.986661418317759e-05, "loss": 0.0119, "step": 52 }, { "epoch": 0.6127167630057804, "grad_norm": 0.023054232820868492, "learning_rate": 9.983534333492575e-05, "loss": 0.0137, "step": 53 }, { "epoch": 0.6242774566473989, "grad_norm": 0.02029217965900898, "learning_rate": 9.980078841304816e-05, "loss": 0.014, "step": 54 }, { "epoch": 0.6358381502890174, "grad_norm": 0.018532264977693558, "learning_rate": 9.976295169466178e-05, "loss": 0.0107, "step": 55 }, { "epoch": 0.6358381502890174, "eval_loss": 0.012713148258626461, "eval_runtime": 10.0087, "eval_samples_per_second": 4.996, "eval_steps_per_second": 1.299, "step": 55 }, { "epoch": 0.6473988439306358, "grad_norm": 0.020468739792704582, "learning_rate": 9.97218356731491e-05, "loss": 0.0146, "step": 56 }, { "epoch": 0.6589595375722543, "grad_norm": 0.01887972466647625, "learning_rate": 9.967744305799357e-05, "loss": 0.0126, "step": 57 }, { "epoch": 0.6705202312138728, "grad_norm": 0.02040836587548256, "learning_rate": 9.962977677460132e-05, "loss": 0.0126, "step": 58 }, { "epoch": 0.6820809248554913, "grad_norm": 0.02685651369392872, "learning_rate": 9.957883996410821e-05, "loss": 0.0122, "step": 59 }, { "epoch": 0.6936416184971098, "grad_norm": 0.02112666517496109, "learning_rate": 9.952463598317285e-05, "loss": 0.016, "step": 60 }, { "epoch": 0.6936416184971098, "eval_loss": 0.012294010259211063, "eval_runtime": 10.0073, "eval_samples_per_second": 4.996, "eval_steps_per_second": 1.299, "step": 60 }, { "epoch": 0.7052023121387283, "grad_norm": 0.020788514986634254, "learning_rate": 9.946716840375551e-05, "loss": 0.0106, "step": 61 }, { "epoch": 0.7167630057803468, "grad_norm": 0.01940411888062954, "learning_rate": 9.940644101288259e-05, "loss": 0.0098, "step": 62 }, { "epoch": 0.7283236994219653, "grad_norm": 0.02072029560804367, "learning_rate": 9.934245781239714e-05, "loss": 0.0142, "step": 63 }, { "epoch": 0.7398843930635838, "grad_norm": 0.017221063375473022, "learning_rate": 9.927522301869515e-05, "loss": 0.0114, "step": 64 }, { "epoch": 0.7514450867052023, "grad_norm": 0.026088256388902664, "learning_rate": 9.920474106244763e-05, "loss": 0.0144, "step": 65 }, { "epoch": 0.7514450867052023, "eval_loss": 0.011602817103266716, "eval_runtime": 10.0191, "eval_samples_per_second": 4.99, "eval_steps_per_second": 1.298, "step": 65 }, { "epoch": 0.7630057803468208, "grad_norm": 0.016981270164251328, "learning_rate": 9.91310165883088e-05, "loss": 0.0136, "step": 66 }, { "epoch": 0.7745664739884393, "grad_norm": 0.017894290387630463, "learning_rate": 9.905405445460972e-05, "loss": 0.0113, "step": 67 }, { "epoch": 0.7861271676300579, "grad_norm": 0.017901500687003136, "learning_rate": 9.897385973303845e-05, "loss": 0.0133, "step": 68 }, { "epoch": 0.7976878612716763, "grad_norm": 0.018053261563181877, "learning_rate": 9.889043770830566e-05, "loss": 0.0108, "step": 69 }, { "epoch": 0.8092485549132948, "grad_norm": 0.023738127201795578, "learning_rate": 9.880379387779637e-05, "loss": 0.0119, "step": 70 }, { "epoch": 0.8092485549132948, "eval_loss": 0.011262123472988605, "eval_runtime": 10.0281, "eval_samples_per_second": 4.986, "eval_steps_per_second": 1.296, "step": 70 }, { "epoch": 0.8208092485549133, "grad_norm": 0.01628166250884533, "learning_rate": 9.871393395120774e-05, "loss": 0.014, "step": 71 }, { "epoch": 0.8323699421965318, "grad_norm": 0.014171602204442024, "learning_rate": 9.862086385017283e-05, "loss": 0.0099, "step": 72 }, { "epoch": 0.8439306358381503, "grad_norm": 0.01973879709839821, "learning_rate": 9.852458970787026e-05, "loss": 0.011, "step": 73 }, { "epoch": 0.8554913294797688, "grad_norm": 0.015728101134300232, "learning_rate": 9.842511786862019e-05, "loss": 0.01, "step": 74 }, { "epoch": 0.8670520231213873, "grad_norm": 0.015097449533641338, "learning_rate": 9.832245488746611e-05, "loss": 0.0111, "step": 75 }, { "epoch": 0.8670520231213873, "eval_loss": 0.010938560590147972, "eval_runtime": 10.0124, "eval_samples_per_second": 4.994, "eval_steps_per_second": 1.298, "step": 75 }, { "epoch": 0.8786127167630058, "grad_norm": 0.02237934060394764, "learning_rate": 9.821660752974293e-05, "loss": 0.0142, "step": 76 }, { "epoch": 0.8901734104046243, "grad_norm": 0.01948779821395874, "learning_rate": 9.810758277063119e-05, "loss": 0.0138, "step": 77 }, { "epoch": 0.9017341040462428, "grad_norm": 0.022487910464406013, "learning_rate": 9.799538779469734e-05, "loss": 0.0151, "step": 78 }, { "epoch": 0.9132947976878613, "grad_norm": 0.017546942457556725, "learning_rate": 9.78800299954203e-05, "loss": 0.0098, "step": 79 }, { "epoch": 0.9248554913294798, "grad_norm": 0.021010980010032654, "learning_rate": 9.77615169747043e-05, "loss": 0.012, "step": 80 }, { "epoch": 0.9248554913294798, "eval_loss": 0.010658971033990383, "eval_runtime": 10.0307, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 80 }, { "epoch": 0.9364161849710982, "grad_norm": 0.018796751275658607, "learning_rate": 9.763985654237786e-05, "loss": 0.0126, "step": 81 }, { "epoch": 0.9479768786127167, "grad_norm": 0.020107241347432137, "learning_rate": 9.751505671567913e-05, "loss": 0.0129, "step": 82 }, { "epoch": 0.9595375722543352, "grad_norm": 0.015596112236380577, "learning_rate": 9.738712571872763e-05, "loss": 0.0093, "step": 83 }, { "epoch": 0.9710982658959537, "grad_norm": 0.018946580588817596, "learning_rate": 9.725607198198227e-05, "loss": 0.0139, "step": 84 }, { "epoch": 0.9826589595375722, "grad_norm": 0.01942116767168045, "learning_rate": 9.712190414168572e-05, "loss": 0.0139, "step": 85 }, { "epoch": 0.9826589595375722, "eval_loss": 0.010201876051723957, "eval_runtime": 10.0325, "eval_samples_per_second": 4.984, "eval_steps_per_second": 1.296, "step": 85 }, { "epoch": 0.9942196531791907, "grad_norm": 0.01808648556470871, "learning_rate": 9.698463103929542e-05, "loss": 0.0096, "step": 86 }, { "epoch": 1.0057803468208093, "grad_norm": 0.033104732632637024, "learning_rate": 9.684426172090085e-05, "loss": 0.0208, "step": 87 }, { "epoch": 1.0173410404624277, "grad_norm": 0.011730443686246872, "learning_rate": 9.67008054366274e-05, "loss": 0.0079, "step": 88 }, { "epoch": 1.0289017341040463, "grad_norm": 0.017196012660861015, "learning_rate": 9.65542716400269e-05, "loss": 0.0103, "step": 89 }, { "epoch": 1.0404624277456647, "grad_norm": 0.015379775315523148, "learning_rate": 9.640466998745456e-05, "loss": 0.0085, "step": 90 }, { "epoch": 1.0404624277456647, "eval_loss": 0.01041076984256506, "eval_runtime": 10.0354, "eval_samples_per_second": 4.982, "eval_steps_per_second": 1.295, "step": 90 }, { "epoch": 1.0520231213872833, "grad_norm": 0.01682409830391407, "learning_rate": 9.625201033743261e-05, "loss": 0.0102, "step": 91 }, { "epoch": 1.0635838150289016, "grad_norm": 0.017098382115364075, "learning_rate": 9.609630275000072e-05, "loss": 0.0081, "step": 92 }, { "epoch": 1.0751445086705202, "grad_norm": 0.017288707196712494, "learning_rate": 9.5937557486053e-05, "loss": 0.0094, "step": 93 }, { "epoch": 1.0867052023121386, "grad_norm": 0.018024342134594917, "learning_rate": 9.577578500666187e-05, "loss": 0.0116, "step": 94 }, { "epoch": 1.0982658959537572, "grad_norm": 0.017229463905096054, "learning_rate": 9.56109959723886e-05, "loss": 0.01, "step": 95 }, { "epoch": 1.0982658959537572, "eval_loss": 0.010240535251796246, "eval_runtime": 10.0835, "eval_samples_per_second": 4.959, "eval_steps_per_second": 1.289, "step": 95 }, { "epoch": 1.1098265895953756, "grad_norm": 0.015899403020739555, "learning_rate": 9.544320124258092e-05, "loss": 0.0099, "step": 96 }, { "epoch": 1.1213872832369942, "grad_norm": 0.019960559904575348, "learning_rate": 9.527241187465734e-05, "loss": 0.0098, "step": 97 }, { "epoch": 1.1329479768786128, "grad_norm": 0.01425469946116209, "learning_rate": 9.509863912337842e-05, "loss": 0.0084, "step": 98 }, { "epoch": 1.1445086705202312, "grad_norm": 0.015252132900059223, "learning_rate": 9.492189444010521e-05, "loss": 0.01, "step": 99 }, { "epoch": 1.1560693641618498, "grad_norm": 0.014408037066459656, "learning_rate": 9.474218947204459e-05, "loss": 0.009, "step": 100 }, { "epoch": 1.1560693641618498, "eval_loss": 0.009868398308753967, "eval_runtime": 10.0137, "eval_samples_per_second": 4.993, "eval_steps_per_second": 1.298, "step": 100 }, { "epoch": 1.1676300578034682, "grad_norm": 0.01568921096622944, "learning_rate": 9.455953606148172e-05, "loss": 0.0125, "step": 101 }, { "epoch": 1.1791907514450868, "grad_norm": 0.01693912222981453, "learning_rate": 9.437394624499958e-05, "loss": 0.0116, "step": 102 }, { "epoch": 1.1907514450867052, "grad_norm": 0.018201276659965515, "learning_rate": 9.418543225268596e-05, "loss": 0.0117, "step": 103 }, { "epoch": 1.2023121387283238, "grad_norm": 0.015434252098202705, "learning_rate": 9.399400650732735e-05, "loss": 0.011, "step": 104 }, { "epoch": 1.2138728323699421, "grad_norm": 0.01606677658855915, "learning_rate": 9.379968162359034e-05, "loss": 0.0094, "step": 105 }, { "epoch": 1.2138728323699421, "eval_loss": 0.009825030341744423, "eval_runtime": 10.0305, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 105 }, { "epoch": 1.2254335260115607, "grad_norm": 0.015970559790730476, "learning_rate": 9.360247040719039e-05, "loss": 0.0079, "step": 106 }, { "epoch": 1.2369942196531791, "grad_norm": 0.01553067285567522, "learning_rate": 9.340238585404788e-05, "loss": 0.0077, "step": 107 }, { "epoch": 1.2485549132947977, "grad_norm": 0.017365023493766785, "learning_rate": 9.319944114943171e-05, "loss": 0.0089, "step": 108 }, { "epoch": 1.260115606936416, "grad_norm": 0.01562884822487831, "learning_rate": 9.29936496670905e-05, "loss": 0.0076, "step": 109 }, { "epoch": 1.2716763005780347, "grad_norm": 0.016722485423088074, "learning_rate": 9.278502496837116e-05, "loss": 0.0069, "step": 110 }, { "epoch": 1.2716763005780347, "eval_loss": 0.009917480871081352, "eval_runtime": 9.9972, "eval_samples_per_second": 5.001, "eval_steps_per_second": 1.3, "step": 110 }, { "epoch": 1.2832369942196533, "grad_norm": 0.019191304221749306, "learning_rate": 9.257358080132523e-05, "loss": 0.0108, "step": 111 }, { "epoch": 1.2947976878612717, "grad_norm": 0.02037121169269085, "learning_rate": 9.235933109980301e-05, "loss": 0.011, "step": 112 }, { "epoch": 1.30635838150289, "grad_norm": 0.016349676996469498, "learning_rate": 9.214228998253527e-05, "loss": 0.0084, "step": 113 }, { "epoch": 1.3179190751445087, "grad_norm": 0.01671903021633625, "learning_rate": 9.192247175220276e-05, "loss": 0.0087, "step": 114 }, { "epoch": 1.3294797687861273, "grad_norm": 0.01710616424679756, "learning_rate": 9.16998908944939e-05, "loss": 0.0108, "step": 115 }, { "epoch": 1.3294797687861273, "eval_loss": 0.009632885456085205, "eval_runtime": 10.0191, "eval_samples_per_second": 4.99, "eval_steps_per_second": 1.298, "step": 115 }, { "epoch": 1.3410404624277457, "grad_norm": 0.01672513224184513, "learning_rate": 9.147456207714997e-05, "loss": 0.0077, "step": 116 }, { "epoch": 1.352601156069364, "grad_norm": 0.017153792083263397, "learning_rate": 9.124650014899867e-05, "loss": 0.0101, "step": 117 }, { "epoch": 1.3641618497109826, "grad_norm": 0.018158160150051117, "learning_rate": 9.101572013897555e-05, "loss": 0.0103, "step": 118 }, { "epoch": 1.3757225433526012, "grad_norm": 0.017670975998044014, "learning_rate": 9.078223725513366e-05, "loss": 0.0112, "step": 119 }, { "epoch": 1.3872832369942196, "grad_norm": 0.013258930295705795, "learning_rate": 9.05460668836413e-05, "loss": 0.0066, "step": 120 }, { "epoch": 1.3872832369942196, "eval_loss": 0.009479235857725143, "eval_runtime": 10.0092, "eval_samples_per_second": 4.995, "eval_steps_per_second": 1.299, "step": 120 }, { "epoch": 1.3988439306358382, "grad_norm": 0.016367750242352486, "learning_rate": 9.030722458776814e-05, "loss": 0.0083, "step": 121 }, { "epoch": 1.4104046242774566, "grad_norm": 0.01966327428817749, "learning_rate": 9.006572610685968e-05, "loss": 0.013, "step": 122 }, { "epoch": 1.4219653179190752, "grad_norm": 0.014227735809981823, "learning_rate": 8.98215873552999e-05, "loss": 0.0081, "step": 123 }, { "epoch": 1.4335260115606936, "grad_norm": 0.0126969488337636, "learning_rate": 8.957482442146272e-05, "loss": 0.0069, "step": 124 }, { "epoch": 1.4450867052023122, "grad_norm": 0.0182335264980793, "learning_rate": 8.932545356665157e-05, "loss": 0.0089, "step": 125 }, { "epoch": 1.4450867052023122, "eval_loss": 0.009439178742468357, "eval_runtime": 10.0231, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 125 }, { "epoch": 1.4566473988439306, "grad_norm": 0.013425913639366627, "learning_rate": 8.907349122402801e-05, "loss": 0.0064, "step": 126 }, { "epoch": 1.4682080924855492, "grad_norm": 0.017706867307424545, "learning_rate": 8.881895399752874e-05, "loss": 0.0085, "step": 127 }, { "epoch": 1.4797687861271676, "grad_norm": 0.015954311937093735, "learning_rate": 8.856185866077129e-05, "loss": 0.0091, "step": 128 }, { "epoch": 1.4913294797687862, "grad_norm": 0.01800154522061348, "learning_rate": 8.83022221559489e-05, "loss": 0.0118, "step": 129 }, { "epoch": 1.5028901734104045, "grad_norm": 0.022386785596609116, "learning_rate": 8.80400615927139e-05, "loss": 0.0084, "step": 130 }, { "epoch": 1.5028901734104045, "eval_loss": 0.009285805746912956, "eval_runtime": 10.0195, "eval_samples_per_second": 4.99, "eval_steps_per_second": 1.297, "step": 130 }, { "epoch": 1.5144508670520231, "grad_norm": 0.015237067826092243, "learning_rate": 8.777539424705023e-05, "loss": 0.0067, "step": 131 }, { "epoch": 1.5260115606936417, "grad_norm": 0.01740083284676075, "learning_rate": 8.750823756013498e-05, "loss": 0.0111, "step": 132 }, { "epoch": 1.5375722543352601, "grad_norm": 0.01614670269191265, "learning_rate": 8.72386091371891e-05, "loss": 0.0066, "step": 133 }, { "epoch": 1.5491329479768785, "grad_norm": 0.018181053921580315, "learning_rate": 8.696652674631717e-05, "loss": 0.0089, "step": 134 }, { "epoch": 1.560693641618497, "grad_norm": 0.01898285746574402, "learning_rate": 8.669200831733655e-05, "loss": 0.0102, "step": 135 }, { "epoch": 1.560693641618497, "eval_loss": 0.00925290398299694, "eval_runtime": 10.0096, "eval_samples_per_second": 4.995, "eval_steps_per_second": 1.299, "step": 135 }, { "epoch": 1.5722543352601157, "grad_norm": 0.01583796739578247, "learning_rate": 8.641507194059579e-05, "loss": 0.0084, "step": 136 }, { "epoch": 1.583815028901734, "grad_norm": 0.020259637385606766, "learning_rate": 8.613573586578262e-05, "loss": 0.0097, "step": 137 }, { "epoch": 1.5953757225433525, "grad_norm": 0.0170609038323164, "learning_rate": 8.585401850072113e-05, "loss": 0.0102, "step": 138 }, { "epoch": 1.606936416184971, "grad_norm": 0.013348150067031384, "learning_rate": 8.55699384101589e-05, "loss": 0.0076, "step": 139 }, { "epoch": 1.6184971098265897, "grad_norm": 0.0179379191249609, "learning_rate": 8.528351431454351e-05, "loss": 0.01, "step": 140 }, { "epoch": 1.6184971098265897, "eval_loss": 0.009090203791856766, "eval_runtime": 10.0116, "eval_samples_per_second": 4.994, "eval_steps_per_second": 1.298, "step": 140 }, { "epoch": 1.630057803468208, "grad_norm": 0.014754540286958218, "learning_rate": 8.499476508878893e-05, "loss": 0.0091, "step": 141 }, { "epoch": 1.6416184971098264, "grad_norm": 0.014680921100080013, "learning_rate": 8.47037097610317e-05, "loss": 0.008, "step": 142 }, { "epoch": 1.653179190751445, "grad_norm": 0.01806177943944931, "learning_rate": 8.441036751137696e-05, "loss": 0.0122, "step": 143 }, { "epoch": 1.6647398843930636, "grad_norm": 0.014901218004524708, "learning_rate": 8.411475767063455e-05, "loss": 0.0075, "step": 144 }, { "epoch": 1.6763005780346822, "grad_norm": 0.01642463356256485, "learning_rate": 8.381689971904514e-05, "loss": 0.0098, "step": 145 }, { "epoch": 1.6763005780346822, "eval_loss": 0.008807710371911526, "eval_runtime": 10.0289, "eval_samples_per_second": 4.986, "eval_steps_per_second": 1.296, "step": 145 }, { "epoch": 1.6878612716763006, "grad_norm": 0.0145337525755167, "learning_rate": 8.35168132849965e-05, "loss": 0.0074, "step": 146 }, { "epoch": 1.699421965317919, "grad_norm": 0.019870057702064514, "learning_rate": 8.321451814372997e-05, "loss": 0.0111, "step": 147 }, { "epoch": 1.7109826589595376, "grad_norm": 0.015163514763116837, "learning_rate": 8.29100342160374e-05, "loss": 0.0097, "step": 148 }, { "epoch": 1.7225433526011562, "grad_norm": 0.01458723470568657, "learning_rate": 8.260338156694836e-05, "loss": 0.007, "step": 149 }, { "epoch": 1.7341040462427746, "grad_norm": 0.015452525578439236, "learning_rate": 8.229458040440783e-05, "loss": 0.0071, "step": 150 }, { "epoch": 1.7341040462427746, "eval_loss": 0.008663667365908623, "eval_runtime": 10.0348, "eval_samples_per_second": 4.983, "eval_steps_per_second": 1.295, "step": 150 }, { "epoch": 1.745664739884393, "grad_norm": 0.017789160832762718, "learning_rate": 8.198365107794457e-05, "loss": 0.0098, "step": 151 }, { "epoch": 1.7572254335260116, "grad_norm": 0.019878627732396126, "learning_rate": 8.167061407733016e-05, "loss": 0.0077, "step": 152 }, { "epoch": 1.7687861271676302, "grad_norm": 0.018271176144480705, "learning_rate": 8.135549003122871e-05, "loss": 0.0096, "step": 153 }, { "epoch": 1.7803468208092486, "grad_norm": 0.017175493761897087, "learning_rate": 8.103829970583742e-05, "loss": 0.0111, "step": 154 }, { "epoch": 1.791907514450867, "grad_norm": 0.014682111330330372, "learning_rate": 8.071906400351822e-05, "loss": 0.0094, "step": 155 }, { "epoch": 1.791907514450867, "eval_loss": 0.008556585758924484, "eval_runtime": 10.0386, "eval_samples_per_second": 4.981, "eval_steps_per_second": 1.295, "step": 155 }, { "epoch": 1.8034682080924855, "grad_norm": 0.016677409410476685, "learning_rate": 8.039780396142022e-05, "loss": 0.0079, "step": 156 }, { "epoch": 1.8150289017341041, "grad_norm": 0.01885165646672249, "learning_rate": 8.007454075009351e-05, "loss": 0.0115, "step": 157 }, { "epoch": 1.8265895953757225, "grad_norm": 0.01522767636924982, "learning_rate": 7.9749295672094e-05, "loss": 0.0076, "step": 158 }, { "epoch": 1.838150289017341, "grad_norm": 0.01811189390718937, "learning_rate": 7.942209016057954e-05, "loss": 0.01, "step": 159 }, { "epoch": 1.8497109826589595, "grad_norm": 0.01717502437531948, "learning_rate": 7.909294577789766e-05, "loss": 0.008, "step": 160 }, { "epoch": 1.8497109826589595, "eval_loss": 0.008559009060263634, "eval_runtime": 10.0652, "eval_samples_per_second": 4.968, "eval_steps_per_second": 1.292, "step": 160 }, { "epoch": 1.861271676300578, "grad_norm": 0.015464117750525475, "learning_rate": 7.876188421416449e-05, "loss": 0.0068, "step": 161 }, { "epoch": 1.8728323699421965, "grad_norm": 0.015217979438602924, "learning_rate": 7.842892728583558e-05, "loss": 0.0074, "step": 162 }, { "epoch": 1.8843930635838149, "grad_norm": 0.01554564293473959, "learning_rate": 7.809409693426803e-05, "loss": 0.0071, "step": 163 }, { "epoch": 1.8959537572254335, "grad_norm": 0.01914595626294613, "learning_rate": 7.775741522427477e-05, "loss": 0.009, "step": 164 }, { "epoch": 1.907514450867052, "grad_norm": 0.01934712752699852, "learning_rate": 7.741890434267043e-05, "loss": 0.01, "step": 165 }, { "epoch": 1.907514450867052, "eval_loss": 0.00854802131652832, "eval_runtime": 10.0226, "eval_samples_per_second": 4.989, "eval_steps_per_second": 1.297, "step": 165 }, { "epoch": 1.9190751445086707, "grad_norm": 0.01625540666282177, "learning_rate": 7.707858659680924e-05, "loss": 0.0071, "step": 166 }, { "epoch": 1.930635838150289, "grad_norm": 0.020419662818312645, "learning_rate": 7.673648441311508e-05, "loss": 0.0076, "step": 167 }, { "epoch": 1.9421965317919074, "grad_norm": 0.012187760323286057, "learning_rate": 7.639262033560359e-05, "loss": 0.0064, "step": 168 }, { "epoch": 1.953757225433526, "grad_norm": 0.015659835189580917, "learning_rate": 7.604701702439651e-05, "loss": 0.0074, "step": 169 }, { "epoch": 1.9653179190751446, "grad_norm": 0.017916489392518997, "learning_rate": 7.56996972542285e-05, "loss": 0.0084, "step": 170 }, { "epoch": 1.9653179190751446, "eval_loss": 0.008561979979276657, "eval_runtime": 10.0323, "eval_samples_per_second": 4.984, "eval_steps_per_second": 1.296, "step": 170 }, { "epoch": 1.976878612716763, "grad_norm": 0.018545281141996384, "learning_rate": 7.535068391294617e-05, "loss": 0.0102, "step": 171 }, { "epoch": 1.9884393063583814, "grad_norm": 0.01769687980413437, "learning_rate": 7.500000000000001e-05, "loss": 0.0086, "step": 172 }, { "epoch": 2.0, "grad_norm": 0.021417992189526558, "learning_rate": 7.464766862492855e-05, "loss": 0.0092, "step": 173 }, { "epoch": 2.0115606936416186, "grad_norm": 0.01275323610752821, "learning_rate": 7.42937130058357e-05, "loss": 0.0069, "step": 174 }, { "epoch": 2.023121387283237, "grad_norm": 0.011270470917224884, "learning_rate": 7.393815646786046e-05, "loss": 0.0058, "step": 175 }, { "epoch": 2.023121387283237, "eval_loss": 0.008691162802278996, "eval_runtime": 10.0831, "eval_samples_per_second": 4.959, "eval_steps_per_second": 1.289, "step": 175 }, { "epoch": 2.0346820809248554, "grad_norm": 0.014580226503312588, "learning_rate": 7.358102244164003e-05, "loss": 0.0077, "step": 176 }, { "epoch": 2.046242774566474, "grad_norm": 0.014212528243660927, "learning_rate": 7.322233446176571e-05, "loss": 0.0074, "step": 177 }, { "epoch": 2.0578034682080926, "grad_norm": 0.01766706444323063, "learning_rate": 7.286211616523193e-05, "loss": 0.007, "step": 178 }, { "epoch": 2.069364161849711, "grad_norm": 0.015258056111633778, "learning_rate": 7.250039128987873e-05, "loss": 0.0073, "step": 179 }, { "epoch": 2.0809248554913293, "grad_norm": 0.01634743995964527, "learning_rate": 7.213718367282737e-05, "loss": 0.0056, "step": 180 }, { "epoch": 2.0809248554913293, "eval_loss": 0.008958614431321621, "eval_runtime": 10.0215, "eval_samples_per_second": 4.989, "eval_steps_per_second": 1.297, "step": 180 }, { "epoch": 2.092485549132948, "grad_norm": 0.014233705587685108, "learning_rate": 7.177251724890956e-05, "loss": 0.0056, "step": 181 }, { "epoch": 2.1040462427745665, "grad_norm": 0.013312169350683689, "learning_rate": 7.14064160490902e-05, "loss": 0.0055, "step": 182 }, { "epoch": 2.115606936416185, "grad_norm": 0.012955575250089169, "learning_rate": 7.103890419888367e-05, "loss": 0.0053, "step": 183 }, { "epoch": 2.1271676300578033, "grad_norm": 0.018879901617765427, "learning_rate": 7.067000591676416e-05, "loss": 0.0065, "step": 184 }, { "epoch": 2.138728323699422, "grad_norm": 0.01857951283454895, "learning_rate": 7.029974551256956e-05, "loss": 0.0077, "step": 185 }, { "epoch": 2.138728323699422, "eval_loss": 0.008626853115856647, "eval_runtime": 10.0278, "eval_samples_per_second": 4.986, "eval_steps_per_second": 1.296, "step": 185 }, { "epoch": 2.1502890173410405, "grad_norm": 0.015720047056674957, "learning_rate": 6.992814738589957e-05, "loss": 0.0057, "step": 186 }, { "epoch": 2.161849710982659, "grad_norm": 0.016932690516114235, "learning_rate": 6.95552360245078e-05, "loss": 0.0074, "step": 187 }, { "epoch": 2.1734104046242773, "grad_norm": 0.013904884457588196, "learning_rate": 6.918103600268799e-05, "loss": 0.0054, "step": 188 }, { "epoch": 2.184971098265896, "grad_norm": 0.015230257995426655, "learning_rate": 6.880557197965464e-05, "loss": 0.0053, "step": 189 }, { "epoch": 2.1965317919075145, "grad_norm": 0.017044221982359886, "learning_rate": 6.842886869791809e-05, "loss": 0.0061, "step": 190 }, { "epoch": 2.1965317919075145, "eval_loss": 0.00858032051473856, "eval_runtime": 10.0219, "eval_samples_per_second": 4.989, "eval_steps_per_second": 1.297, "step": 190 }, { "epoch": 2.208092485549133, "grad_norm": 0.01593642681837082, "learning_rate": 6.805095098165389e-05, "loss": 0.0056, "step": 191 }, { "epoch": 2.2196531791907512, "grad_norm": 0.018296780064702034, "learning_rate": 6.767184373506697e-05, "loss": 0.008, "step": 192 }, { "epoch": 2.23121387283237, "grad_norm": 0.02369534969329834, "learning_rate": 6.729157194075057e-05, "loss": 0.0116, "step": 193 }, { "epoch": 2.2427745664739884, "grad_norm": 0.020556915551424026, "learning_rate": 6.691016065803983e-05, "loss": 0.0078, "step": 194 }, { "epoch": 2.254335260115607, "grad_norm": 0.01714545115828514, "learning_rate": 6.652763502136043e-05, "loss": 0.008, "step": 195 }, { "epoch": 2.254335260115607, "eval_loss": 0.008329370059072971, "eval_runtime": 10.0304, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 195 }, { "epoch": 2.2658959537572256, "grad_norm": 0.013920610770583153, "learning_rate": 6.614402023857232e-05, "loss": 0.0057, "step": 196 }, { "epoch": 2.277456647398844, "grad_norm": 0.014315689913928509, "learning_rate": 6.57593415893085e-05, "loss": 0.0062, "step": 197 }, { "epoch": 2.2890173410404624, "grad_norm": 0.01451371144503355, "learning_rate": 6.537362442330916e-05, "loss": 0.0059, "step": 198 }, { "epoch": 2.300578034682081, "grad_norm": 0.013728981837630272, "learning_rate": 6.498689415875121e-05, "loss": 0.0054, "step": 199 }, { "epoch": 2.3121387283236996, "grad_norm": 0.01417378056794405, "learning_rate": 6.45991762805732e-05, "loss": 0.0058, "step": 200 }, { "epoch": 2.3121387283236996, "eval_loss": 0.00831348542124033, "eval_runtime": 10.0504, "eval_samples_per_second": 4.975, "eval_steps_per_second": 1.293, "step": 200 }, { "epoch": 2.3236994219653178, "grad_norm": 0.013795462436974049, "learning_rate": 6.421049633879588e-05, "loss": 0.005, "step": 201 }, { "epoch": 2.3352601156069364, "grad_norm": 0.01878678984940052, "learning_rate": 6.382087994683859e-05, "loss": 0.0075, "step": 202 }, { "epoch": 2.346820809248555, "grad_norm": 0.01882004365324974, "learning_rate": 6.343035277983127e-05, "loss": 0.0068, "step": 203 }, { "epoch": 2.3583815028901736, "grad_norm": 0.015659485012292862, "learning_rate": 6.303894057292261e-05, "loss": 0.006, "step": 204 }, { "epoch": 2.3699421965317917, "grad_norm": 0.015219546854496002, "learning_rate": 6.264666911958404e-05, "loss": 0.0047, "step": 205 }, { "epoch": 2.3699421965317917, "eval_loss": 0.00836299266666174, "eval_runtime": 10.0293, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 205 }, { "epoch": 2.3815028901734103, "grad_norm": 0.017362525686621666, "learning_rate": 6.225356426991007e-05, "loss": 0.0069, "step": 206 }, { "epoch": 2.393063583815029, "grad_norm": 0.014902186580002308, "learning_rate": 6.185965192891472e-05, "loss": 0.0055, "step": 207 }, { "epoch": 2.4046242774566475, "grad_norm": 0.016910729929804802, "learning_rate": 6.146495805482451e-05, "loss": 0.0062, "step": 208 }, { "epoch": 2.416184971098266, "grad_norm": 0.014512901194393635, "learning_rate": 6.106950865736777e-05, "loss": 0.0052, "step": 209 }, { "epoch": 2.4277456647398843, "grad_norm": 0.017721183598041534, "learning_rate": 6.0673329796060686e-05, "loss": 0.0066, "step": 210 }, { "epoch": 2.4277456647398843, "eval_loss": 0.008446984924376011, "eval_runtime": 10.0239, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 210 }, { "epoch": 2.439306358381503, "grad_norm": 0.01806679181754589, "learning_rate": 6.0276447578490035e-05, "loss": 0.0063, "step": 211 }, { "epoch": 2.4508670520231215, "grad_norm": 0.01761380210518837, "learning_rate": 5.987888815859266e-05, "loss": 0.0058, "step": 212 }, { "epoch": 2.4624277456647397, "grad_norm": 0.015691956505179405, "learning_rate": 5.9480677734932045e-05, "loss": 0.0057, "step": 213 }, { "epoch": 2.4739884393063583, "grad_norm": 0.016050070524215698, "learning_rate": 5.908184254897182e-05, "loss": 0.0073, "step": 214 }, { "epoch": 2.485549132947977, "grad_norm": 0.013452823273837566, "learning_rate": 5.868240888334653e-05, "loss": 0.0055, "step": 215 }, { "epoch": 2.485549132947977, "eval_loss": 0.008248448371887207, "eval_runtime": 10.0544, "eval_samples_per_second": 4.973, "eval_steps_per_second": 1.293, "step": 215 }, { "epoch": 2.4971098265895955, "grad_norm": 0.013310995884239674, "learning_rate": 5.8282403060129566e-05, "loss": 0.0064, "step": 216 }, { "epoch": 2.508670520231214, "grad_norm": 0.016307106241583824, "learning_rate": 5.788185143909868e-05, "loss": 0.006, "step": 217 }, { "epoch": 2.520231213872832, "grad_norm": 0.014403578825294971, "learning_rate": 5.7480780415998906e-05, "loss": 0.0053, "step": 218 }, { "epoch": 2.531791907514451, "grad_norm": 0.015278173610568047, "learning_rate": 5.707921642080299e-05, "loss": 0.0067, "step": 219 }, { "epoch": 2.5433526011560694, "grad_norm": 0.01573382504284382, "learning_rate": 5.66771859159699e-05, "loss": 0.0056, "step": 220 }, { "epoch": 2.5433526011560694, "eval_loss": 0.008289137855172157, "eval_runtime": 10.0106, "eval_samples_per_second": 4.995, "eval_steps_per_second": 1.299, "step": 220 }, { "epoch": 2.5549132947976876, "grad_norm": 0.014673051424324512, "learning_rate": 5.6274715394700805e-05, "loss": 0.0057, "step": 221 }, { "epoch": 2.5664739884393066, "grad_norm": 0.017399994656443596, "learning_rate": 5.587183137919332e-05, "loss": 0.0066, "step": 222 }, { "epoch": 2.578034682080925, "grad_norm": 0.015570064075291157, "learning_rate": 5.546856041889373e-05, "loss": 0.0062, "step": 223 }, { "epoch": 2.5895953757225434, "grad_norm": 0.012873631902039051, "learning_rate": 5.5064929088747316e-05, "loss": 0.0048, "step": 224 }, { "epoch": 2.601156069364162, "grad_norm": 0.01452395599335432, "learning_rate": 5.46609639874473e-05, "loss": 0.005, "step": 225 }, { "epoch": 2.601156069364162, "eval_loss": 0.008248119615018368, "eval_runtime": 10.0518, "eval_samples_per_second": 4.974, "eval_steps_per_second": 1.293, "step": 225 }, { "epoch": 2.61271676300578, "grad_norm": 0.015882886946201324, "learning_rate": 5.4256691735681786e-05, "loss": 0.006, "step": 226 }, { "epoch": 2.6242774566473988, "grad_norm": 0.01853213645517826, "learning_rate": 5.385213897437975e-05, "loss": 0.0072, "step": 227 }, { "epoch": 2.6358381502890174, "grad_norm": 0.01727055385708809, "learning_rate": 5.344733236295525e-05, "loss": 0.0056, "step": 228 }, { "epoch": 2.647398843930636, "grad_norm": 0.01715169847011566, "learning_rate": 5.3042298577550696e-05, "loss": 0.007, "step": 229 }, { "epoch": 2.6589595375722546, "grad_norm": 0.01657768152654171, "learning_rate": 5.263706430927895e-05, "loss": 0.0065, "step": 230 }, { "epoch": 2.6589595375722546, "eval_loss": 0.00817356538027525, "eval_runtime": 10.0412, "eval_samples_per_second": 4.979, "eval_steps_per_second": 1.295, "step": 230 }, { "epoch": 2.6705202312138727, "grad_norm": 0.016935985535383224, "learning_rate": 5.223165626246432e-05, "loss": 0.0056, "step": 231 }, { "epoch": 2.6820809248554913, "grad_norm": 0.016983183100819588, "learning_rate": 5.182610115288295e-05, "loss": 0.0054, "step": 232 }, { "epoch": 2.69364161849711, "grad_norm": 0.014110002666711807, "learning_rate": 5.142042570600212e-05, "loss": 0.0048, "step": 233 }, { "epoch": 2.705202312138728, "grad_norm": 0.015584226697683334, "learning_rate": 5.1014656655219197e-05, "loss": 0.0074, "step": 234 }, { "epoch": 2.7167630057803467, "grad_norm": 0.015623894520103931, "learning_rate": 5.060882074009988e-05, "loss": 0.0061, "step": 235 }, { "epoch": 2.7167630057803467, "eval_loss": 0.00805756263434887, "eval_runtime": 10.0337, "eval_samples_per_second": 4.983, "eval_steps_per_second": 1.296, "step": 235 }, { "epoch": 2.7283236994219653, "grad_norm": 0.014514243230223656, "learning_rate": 5.020294470461615e-05, "loss": 0.005, "step": 236 }, { "epoch": 2.739884393063584, "grad_norm": 0.016889233142137527, "learning_rate": 4.979705529538385e-05, "loss": 0.0065, "step": 237 }, { "epoch": 2.7514450867052025, "grad_norm": 0.014339439570903778, "learning_rate": 4.9391179259900125e-05, "loss": 0.0047, "step": 238 }, { "epoch": 2.7630057803468207, "grad_norm": 0.019386202096939087, "learning_rate": 4.8985343344780815e-05, "loss": 0.0055, "step": 239 }, { "epoch": 2.7745664739884393, "grad_norm": 0.016825564205646515, "learning_rate": 4.857957429399788e-05, "loss": 0.0052, "step": 240 }, { "epoch": 2.7745664739884393, "eval_loss": 0.008152703754603863, "eval_runtime": 10.0237, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 240 }, { "epoch": 2.786127167630058, "grad_norm": 0.018642200157046318, "learning_rate": 4.817389884711705e-05, "loss": 0.0054, "step": 241 }, { "epoch": 2.7976878612716765, "grad_norm": 0.017443155869841576, "learning_rate": 4.776834373753569e-05, "loss": 0.0066, "step": 242 }, { "epoch": 2.809248554913295, "grad_norm": 0.017855728045105934, "learning_rate": 4.7362935690721076e-05, "loss": 0.0064, "step": 243 }, { "epoch": 2.820809248554913, "grad_norm": 0.013038886711001396, "learning_rate": 4.695770142244931e-05, "loss": 0.0054, "step": 244 }, { "epoch": 2.832369942196532, "grad_norm": 0.0145301828160882, "learning_rate": 4.655266763704476e-05, "loss": 0.0053, "step": 245 }, { "epoch": 2.832369942196532, "eval_loss": 0.008140629157423973, "eval_runtime": 10.0264, "eval_samples_per_second": 4.987, "eval_steps_per_second": 1.297, "step": 245 }, { "epoch": 2.8439306358381504, "grad_norm": 0.015260045416653156, "learning_rate": 4.614786102562026e-05, "loss": 0.0048, "step": 246 }, { "epoch": 2.8554913294797686, "grad_norm": 0.014731310307979584, "learning_rate": 4.574330826431821e-05, "loss": 0.005, "step": 247 }, { "epoch": 2.867052023121387, "grad_norm": 0.017430312931537628, "learning_rate": 4.5339036012552716e-05, "loss": 0.0062, "step": 248 }, { "epoch": 2.878612716763006, "grad_norm": 0.01655244641005993, "learning_rate": 4.493507091125269e-05, "loss": 0.0056, "step": 249 }, { "epoch": 2.8901734104046244, "grad_norm": 0.016315851360559464, "learning_rate": 4.4531439581106295e-05, "loss": 0.0058, "step": 250 }, { "epoch": 2.8901734104046244, "eval_loss": 0.007892865687608719, "eval_runtime": 10.0251, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 250 }, { "epoch": 2.901734104046243, "grad_norm": 0.014720520935952663, "learning_rate": 4.412816862080668e-05, "loss": 0.0064, "step": 251 }, { "epoch": 2.913294797687861, "grad_norm": 0.015001147985458374, "learning_rate": 4.37252846052992e-05, "loss": 0.0057, "step": 252 }, { "epoch": 2.9248554913294798, "grad_norm": 0.015956144779920578, "learning_rate": 4.332281408403011e-05, "loss": 0.0056, "step": 253 }, { "epoch": 2.9364161849710984, "grad_norm": 0.016037022694945335, "learning_rate": 4.292078357919701e-05, "loss": 0.0059, "step": 254 }, { "epoch": 2.9479768786127165, "grad_norm": 0.014494068920612335, "learning_rate": 4.25192195840011e-05, "loss": 0.0052, "step": 255 }, { "epoch": 2.9479768786127165, "eval_loss": 0.007827498018741608, "eval_runtime": 10.0269, "eval_samples_per_second": 4.987, "eval_steps_per_second": 1.297, "step": 255 }, { "epoch": 2.959537572254335, "grad_norm": 0.015506073832511902, "learning_rate": 4.2118148560901325e-05, "loss": 0.0066, "step": 256 }, { "epoch": 2.9710982658959537, "grad_norm": 0.01482071541249752, "learning_rate": 4.171759693987046e-05, "loss": 0.0062, "step": 257 }, { "epoch": 2.9826589595375723, "grad_norm": 0.0164498183876276, "learning_rate": 4.131759111665349e-05, "loss": 0.0056, "step": 258 }, { "epoch": 2.994219653179191, "grad_norm": 0.014011871069669724, "learning_rate": 4.0918157451028185e-05, "loss": 0.0053, "step": 259 }, { "epoch": 3.005780346820809, "grad_norm": 0.024348195642232895, "learning_rate": 4.051932226506797e-05, "loss": 0.0071, "step": 260 }, { "epoch": 3.005780346820809, "eval_loss": 0.007992882281541824, "eval_runtime": 10.0301, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 260 }, { "epoch": 3.0173410404624277, "grad_norm": 0.018887817859649658, "learning_rate": 4.012111184140734e-05, "loss": 0.0051, "step": 261 }, { "epoch": 3.0289017341040463, "grad_norm": 0.017569325864315033, "learning_rate": 3.972355242150998e-05, "loss": 0.0051, "step": 262 }, { "epoch": 3.040462427745665, "grad_norm": 0.013190064579248428, "learning_rate": 3.932667020393933e-05, "loss": 0.0039, "step": 263 }, { "epoch": 3.052023121387283, "grad_norm": 0.015570548363029957, "learning_rate": 3.893049134263224e-05, "loss": 0.0035, "step": 264 }, { "epoch": 3.0635838150289016, "grad_norm": 0.018783550709486008, "learning_rate": 3.8535041945175506e-05, "loss": 0.0051, "step": 265 }, { "epoch": 3.0635838150289016, "eval_loss": 0.008158711716532707, "eval_runtime": 10.0165, "eval_samples_per_second": 4.992, "eval_steps_per_second": 1.298, "step": 265 }, { "epoch": 3.0751445086705202, "grad_norm": 0.01301144901663065, "learning_rate": 3.814034807108529e-05, "loss": 0.0034, "step": 266 }, { "epoch": 3.086705202312139, "grad_norm": 0.019954444840550423, "learning_rate": 3.774643573008995e-05, "loss": 0.0067, "step": 267 }, { "epoch": 3.098265895953757, "grad_norm": 0.01796858012676239, "learning_rate": 3.735333088041596e-05, "loss": 0.0037, "step": 268 }, { "epoch": 3.1098265895953756, "grad_norm": 0.016962885856628418, "learning_rate": 3.69610594270774e-05, "loss": 0.0039, "step": 269 }, { "epoch": 3.121387283236994, "grad_norm": 0.017634421586990356, "learning_rate": 3.656964722016875e-05, "loss": 0.0033, "step": 270 }, { "epoch": 3.121387283236994, "eval_loss": 0.00860854797065258, "eval_runtime": 10.0238, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 270 }, { "epoch": 3.132947976878613, "grad_norm": 0.019069235771894455, "learning_rate": 3.6179120053161416e-05, "loss": 0.0044, "step": 271 }, { "epoch": 3.1445086705202314, "grad_norm": 0.01804421842098236, "learning_rate": 3.578950366120414e-05, "loss": 0.004, "step": 272 }, { "epoch": 3.1560693641618496, "grad_norm": 0.016116416081786156, "learning_rate": 3.5400823719426815e-05, "loss": 0.0038, "step": 273 }, { "epoch": 3.167630057803468, "grad_norm": 0.019759787246584892, "learning_rate": 3.5013105841248795e-05, "loss": 0.0036, "step": 274 }, { "epoch": 3.179190751445087, "grad_norm": 0.014977723360061646, "learning_rate": 3.462637557669084e-05, "loss": 0.004, "step": 275 }, { "epoch": 3.179190751445087, "eval_loss": 0.008404241874814034, "eval_runtime": 10.0295, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 275 }, { "epoch": 3.1907514450867054, "grad_norm": 0.016791895031929016, "learning_rate": 3.424065841069152e-05, "loss": 0.0049, "step": 276 }, { "epoch": 3.2023121387283235, "grad_norm": 0.013895588926970959, "learning_rate": 3.38559797614277e-05, "loss": 0.0039, "step": 277 }, { "epoch": 3.213872832369942, "grad_norm": 0.017442645505070686, "learning_rate": 3.347236497863957e-05, "loss": 0.0042, "step": 278 }, { "epoch": 3.2254335260115607, "grad_norm": 0.018163053318858147, "learning_rate": 3.308983934196018e-05, "loss": 0.0044, "step": 279 }, { "epoch": 3.2369942196531793, "grad_norm": 0.017282122746109962, "learning_rate": 3.2708428059249436e-05, "loss": 0.0032, "step": 280 }, { "epoch": 3.2369942196531793, "eval_loss": 0.008181248791515827, "eval_runtime": 10.0205, "eval_samples_per_second": 4.99, "eval_steps_per_second": 1.297, "step": 280 }, { "epoch": 3.2485549132947975, "grad_norm": 0.017854265868663788, "learning_rate": 3.232815626493304e-05, "loss": 0.0054, "step": 281 }, { "epoch": 3.260115606936416, "grad_norm": 0.0203102994710207, "learning_rate": 3.1949049018346126e-05, "loss": 0.0052, "step": 282 }, { "epoch": 3.2716763005780347, "grad_norm": 0.015286475419998169, "learning_rate": 3.157113130208191e-05, "loss": 0.0049, "step": 283 }, { "epoch": 3.2832369942196533, "grad_norm": 0.014563402161002159, "learning_rate": 3.1194428020345376e-05, "loss": 0.0039, "step": 284 }, { "epoch": 3.294797687861272, "grad_norm": 0.016170360147953033, "learning_rate": 3.081896399731202e-05, "loss": 0.0042, "step": 285 }, { "epoch": 3.294797687861272, "eval_loss": 0.008185806684195995, "eval_runtime": 10.0071, "eval_samples_per_second": 4.996, "eval_steps_per_second": 1.299, "step": 285 }, { "epoch": 3.30635838150289, "grad_norm": 0.017802009359002113, "learning_rate": 3.0444763975492208e-05, "loss": 0.0045, "step": 286 }, { "epoch": 3.3179190751445087, "grad_norm": 0.013784883543848991, "learning_rate": 3.0071852614100426e-05, "loss": 0.0036, "step": 287 }, { "epoch": 3.3294797687861273, "grad_norm": 0.01664113625884056, "learning_rate": 2.9700254487430444e-05, "loss": 0.0044, "step": 288 }, { "epoch": 3.3410404624277454, "grad_norm": 0.01927308365702629, "learning_rate": 2.9329994083235857e-05, "loss": 0.0055, "step": 289 }, { "epoch": 3.352601156069364, "grad_norm": 0.01657339371740818, "learning_rate": 2.896109580111634e-05, "loss": 0.0035, "step": 290 }, { "epoch": 3.352601156069364, "eval_loss": 0.008166169747710228, "eval_runtime": 10.0143, "eval_samples_per_second": 4.993, "eval_steps_per_second": 1.298, "step": 290 }, { "epoch": 3.3641618497109826, "grad_norm": 0.017788324505090714, "learning_rate": 2.859358395090983e-05, "loss": 0.0042, "step": 291 }, { "epoch": 3.3757225433526012, "grad_norm": 0.016261830925941467, "learning_rate": 2.8227482751090445e-05, "loss": 0.0043, "step": 292 }, { "epoch": 3.38728323699422, "grad_norm": 0.01472635380923748, "learning_rate": 2.7862816327172636e-05, "loss": 0.0037, "step": 293 }, { "epoch": 3.398843930635838, "grad_norm": 0.016420332714915276, "learning_rate": 2.7499608710121288e-05, "loss": 0.0042, "step": 294 }, { "epoch": 3.4104046242774566, "grad_norm": 0.016155356541275978, "learning_rate": 2.7137883834768073e-05, "loss": 0.0041, "step": 295 }, { "epoch": 3.4104046242774566, "eval_loss": 0.008127645589411259, "eval_runtime": 10.0238, "eval_samples_per_second": 4.988, "eval_steps_per_second": 1.297, "step": 295 }, { "epoch": 3.421965317919075, "grad_norm": 0.016076598316431046, "learning_rate": 2.6777665538234293e-05, "loss": 0.004, "step": 296 }, { "epoch": 3.433526011560694, "grad_norm": 0.017685944214463234, "learning_rate": 2.641897755835997e-05, "loss": 0.0046, "step": 297 }, { "epoch": 3.445086705202312, "grad_norm": 0.03867075592279434, "learning_rate": 2.606184353213956e-05, "loss": 0.0039, "step": 298 }, { "epoch": 3.4566473988439306, "grad_norm": 0.014819984324276447, "learning_rate": 2.5706286994164315e-05, "loss": 0.0037, "step": 299 }, { "epoch": 3.468208092485549, "grad_norm": 0.016593076288700104, "learning_rate": 2.5352331375071437e-05, "loss": 0.0048, "step": 300 }, { "epoch": 3.468208092485549, "eval_loss": 0.007986278273165226, "eval_runtime": 10.0089, "eval_samples_per_second": 4.996, "eval_steps_per_second": 1.299, "step": 300 }, { "epoch": 3.479768786127168, "grad_norm": 0.015115424059331417, "learning_rate": 2.500000000000001e-05, "loss": 0.0034, "step": 301 }, { "epoch": 3.491329479768786, "grad_norm": 0.016738681122660637, "learning_rate": 2.4649316087053837e-05, "loss": 0.0041, "step": 302 }, { "epoch": 3.5028901734104045, "grad_norm": 0.014471864327788353, "learning_rate": 2.4300302745771507e-05, "loss": 0.0034, "step": 303 }, { "epoch": 3.514450867052023, "grad_norm": 0.015291919000446796, "learning_rate": 2.3952982975603496e-05, "loss": 0.0043, "step": 304 }, { "epoch": 3.5260115606936417, "grad_norm": 0.017901018261909485, "learning_rate": 2.360737966439641e-05, "loss": 0.0046, "step": 305 }, { "epoch": 3.5260115606936417, "eval_loss": 0.00799076072871685, "eval_runtime": 10.1032, "eval_samples_per_second": 4.949, "eval_steps_per_second": 1.287, "step": 305 }, { "epoch": 3.5375722543352603, "grad_norm": 0.017450664192438126, "learning_rate": 2.326351558688493e-05, "loss": 0.0047, "step": 306 }, { "epoch": 3.5491329479768785, "grad_norm": 0.017974600195884705, "learning_rate": 2.2921413403190772e-05, "loss": 0.0046, "step": 307 }, { "epoch": 3.560693641618497, "grad_norm": 0.014545281417667866, "learning_rate": 2.2581095657329598e-05, "loss": 0.0033, "step": 308 }, { "epoch": 3.5722543352601157, "grad_norm": 0.015802579000592232, "learning_rate": 2.224258477572524e-05, "loss": 0.0037, "step": 309 }, { "epoch": 3.583815028901734, "grad_norm": 0.017255526036024094, "learning_rate": 2.1905903065731974e-05, "loss": 0.004, "step": 310 }, { "epoch": 3.583815028901734, "eval_loss": 0.007966941222548485, "eval_runtime": 10.0846, "eval_samples_per_second": 4.958, "eval_steps_per_second": 1.289, "step": 310 }, { "epoch": 3.5953757225433525, "grad_norm": 0.014843069948256016, "learning_rate": 2.1571072714164443e-05, "loss": 0.0035, "step": 311 }, { "epoch": 3.606936416184971, "grad_norm": 0.01944170519709587, "learning_rate": 2.123811578583551e-05, "loss": 0.0056, "step": 312 }, { "epoch": 3.6184971098265897, "grad_norm": 0.017738202586770058, "learning_rate": 2.090705422210237e-05, "loss": 0.0043, "step": 313 }, { "epoch": 3.6300578034682083, "grad_norm": 0.018833834677934647, "learning_rate": 2.057790983942047e-05, "loss": 0.0049, "step": 314 }, { "epoch": 3.6416184971098264, "grad_norm": 0.013943798840045929, "learning_rate": 2.0250704327906024e-05, "loss": 0.0032, "step": 315 }, { "epoch": 3.6416184971098264, "eval_loss": 0.008140211924910545, "eval_runtime": 10.0297, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 315 }, { "epoch": 3.653179190751445, "grad_norm": 0.015995629131793976, "learning_rate": 1.9925459249906485e-05, "loss": 0.0037, "step": 316 }, { "epoch": 3.6647398843930636, "grad_norm": 0.017212923616170883, "learning_rate": 1.960219603857977e-05, "loss": 0.0048, "step": 317 }, { "epoch": 3.6763005780346822, "grad_norm": 0.019134720787405968, "learning_rate": 1.928093599648179e-05, "loss": 0.0047, "step": 318 }, { "epoch": 3.687861271676301, "grad_norm": 0.016614330932497978, "learning_rate": 1.8961700294162577e-05, "loss": 0.0034, "step": 319 }, { "epoch": 3.699421965317919, "grad_norm": 0.017316371202468872, "learning_rate": 1.86445099687713e-05, "loss": 0.0039, "step": 320 }, { "epoch": 3.699421965317919, "eval_loss": 0.008361711166799068, "eval_runtime": 10.0095, "eval_samples_per_second": 4.995, "eval_steps_per_second": 1.299, "step": 320 }, { "epoch": 3.7109826589595376, "grad_norm": 0.015041043981909752, "learning_rate": 1.832938592266984e-05, "loss": 0.0037, "step": 321 }, { "epoch": 3.722543352601156, "grad_norm": 0.0185141172260046, "learning_rate": 1.801634892205545e-05, "loss": 0.0041, "step": 322 }, { "epoch": 3.7341040462427744, "grad_norm": 0.017875155434012413, "learning_rate": 1.7705419595592193e-05, "loss": 0.0037, "step": 323 }, { "epoch": 3.745664739884393, "grad_norm": 0.015563211403787136, "learning_rate": 1.7396618433051647e-05, "loss": 0.0043, "step": 324 }, { "epoch": 3.7572254335260116, "grad_norm": 0.019554760307073593, "learning_rate": 1.7089965783962608e-05, "loss": 0.0042, "step": 325 }, { "epoch": 3.7572254335260116, "eval_loss": 0.008260131813585758, "eval_runtime": 10.0115, "eval_samples_per_second": 4.994, "eval_steps_per_second": 1.299, "step": 325 }, { "epoch": 3.76878612716763, "grad_norm": 0.016527488827705383, "learning_rate": 1.678548185627004e-05, "loss": 0.0043, "step": 326 }, { "epoch": 3.7803468208092488, "grad_norm": 0.015184947289526463, "learning_rate": 1.6483186715003523e-05, "loss": 0.0035, "step": 327 }, { "epoch": 3.791907514450867, "grad_norm": 0.020948296412825584, "learning_rate": 1.618310028095486e-05, "loss": 0.0039, "step": 328 }, { "epoch": 3.8034682080924855, "grad_norm": 0.014377256855368614, "learning_rate": 1.5885242329365448e-05, "loss": 0.0033, "step": 329 }, { "epoch": 3.815028901734104, "grad_norm": 0.017349926754832268, "learning_rate": 1.5589632488623053e-05, "loss": 0.0046, "step": 330 }, { "epoch": 3.815028901734104, "eval_loss": 0.008042494766414165, "eval_runtime": 10.1315, "eval_samples_per_second": 4.935, "eval_steps_per_second": 1.283, "step": 330 }, { "epoch": 3.8265895953757223, "grad_norm": 0.014084051363170147, "learning_rate": 1.5296290238968303e-05, "loss": 0.0032, "step": 331 }, { "epoch": 3.838150289017341, "grad_norm": 0.016601048409938812, "learning_rate": 1.500523491121108e-05, "loss": 0.0035, "step": 332 }, { "epoch": 3.8497109826589595, "grad_norm": 0.015788624063134193, "learning_rate": 1.47164856854565e-05, "loss": 0.0038, "step": 333 }, { "epoch": 3.861271676300578, "grad_norm": 0.014880317263305187, "learning_rate": 1.4430061589841121e-05, "loss": 0.0034, "step": 334 }, { "epoch": 3.8728323699421967, "grad_norm": 0.01920904405415058, "learning_rate": 1.4145981499278876e-05, "loss": 0.0035, "step": 335 }, { "epoch": 3.8728323699421967, "eval_loss": 0.008081013336777687, "eval_runtime": 10.0106, "eval_samples_per_second": 4.995, "eval_steps_per_second": 1.299, "step": 335 }, { "epoch": 3.884393063583815, "grad_norm": 0.016501102596521378, "learning_rate": 1.386426413421738e-05, "loss": 0.004, "step": 336 }, { "epoch": 3.8959537572254335, "grad_norm": 0.016265930607914925, "learning_rate": 1.3584928059404205e-05, "loss": 0.0034, "step": 337 }, { "epoch": 3.907514450867052, "grad_norm": 0.019726134836673737, "learning_rate": 1.3307991682663462e-05, "loss": 0.0031, "step": 338 }, { "epoch": 3.9190751445086707, "grad_norm": 0.016232285648584366, "learning_rate": 1.3033473253682848e-05, "loss": 0.0034, "step": 339 }, { "epoch": 3.9306358381502893, "grad_norm": 0.02013998292386532, "learning_rate": 1.2761390862810907e-05, "loss": 0.0048, "step": 340 }, { "epoch": 3.9306358381502893, "eval_loss": 0.008075674995779991, "eval_runtime": 10.0299, "eval_samples_per_second": 4.985, "eval_steps_per_second": 1.296, "step": 340 }, { "epoch": 3.9421965317919074, "grad_norm": 0.013452830724418163, "learning_rate": 1.2491762439865035e-05, "loss": 0.003, "step": 341 }, { "epoch": 3.953757225433526, "grad_norm": 0.012992053292691708, "learning_rate": 1.2224605752949786e-05, "loss": 0.0028, "step": 342 }, { "epoch": 3.9653179190751446, "grad_norm": 0.015416258946061134, "learning_rate": 1.1959938407286097e-05, "loss": 0.0037, "step": 343 }, { "epoch": 3.976878612716763, "grad_norm": 0.015255059115588665, "learning_rate": 1.1697777844051105e-05, "loss": 0.0036, "step": 344 }, { "epoch": 3.9884393063583814, "grad_norm": 0.018186306580901146, "learning_rate": 1.143814133922872e-05, "loss": 0.0056, "step": 345 }, { "epoch": 3.9884393063583814, "eval_loss": 0.008018155582249165, "eval_runtime": 10.115, "eval_samples_per_second": 4.943, "eval_steps_per_second": 1.285, "step": 345 }, { "epoch": 4.0, "grad_norm": 0.026972023770213127, "learning_rate": 1.118104600247129e-05, "loss": 0.0065, "step": 346 }, { "epoch": 4.011560693641618, "grad_norm": 0.01358290296047926, "learning_rate": 1.0926508775971994e-05, "loss": 0.0029, "step": 347 }, { "epoch": 4.023121387283237, "grad_norm": 0.011624607257544994, "learning_rate": 1.0674546433348454e-05, "loss": 0.0028, "step": 348 }, { "epoch": 4.034682080924855, "grad_norm": 0.012059401720762253, "learning_rate": 1.0425175578537299e-05, "loss": 0.0033, "step": 349 }, { "epoch": 4.046242774566474, "grad_norm": 0.011458152905106544, "learning_rate": 1.0178412644700092e-05, "loss": 0.0025, "step": 350 }, { "epoch": 4.046242774566474, "eval_loss": 0.008002113550901413, "eval_runtime": 10.1777, "eval_samples_per_second": 4.913, "eval_steps_per_second": 1.277, "step": 350 }, { "epoch": 4.057803468208093, "grad_norm": 0.012259789742529392, "learning_rate": 9.934273893140334e-06, "loss": 0.0027, "step": 351 }, { "epoch": 4.069364161849711, "grad_norm": 0.013933391310274601, "learning_rate": 9.692775412231863e-06, "loss": 0.0027, "step": 352 }, { "epoch": 4.08092485549133, "grad_norm": 0.012962523847818375, "learning_rate": 9.453933116358715e-06, "loss": 0.0024, "step": 353 }, { "epoch": 4.092485549132948, "grad_norm": 0.01309504546225071, "learning_rate": 9.21776274486636e-06, "loss": 0.0028, "step": 354 }, { "epoch": 4.104046242774566, "grad_norm": 0.015043212100863457, "learning_rate": 8.984279861024453e-06, "loss": 0.0035, "step": 355 }, { "epoch": 4.104046242774566, "eval_loss": 0.008168390020728111, "eval_runtime": 10.0257, "eval_samples_per_second": 4.987, "eval_steps_per_second": 1.297, "step": 355 }, { "epoch": 4.115606936416185, "grad_norm": 0.014944328926503658, "learning_rate": 8.75349985100134e-06, "loss": 0.0034, "step": 356 }, { "epoch": 4.127167630057803, "grad_norm": 0.016272595152258873, "learning_rate": 8.525437922850032e-06, "loss": 0.0025, "step": 357 }, { "epoch": 4.138728323699422, "grad_norm": 0.013790381141006947, "learning_rate": 8.30010910550611e-06, "loss": 0.0025, "step": 358 }, { "epoch": 4.1502890173410405, "grad_norm": 0.013851807452738285, "learning_rate": 8.077528247797234e-06, "loss": 0.0029, "step": 359 }, { "epoch": 4.161849710982659, "grad_norm": 0.013270116411149502, "learning_rate": 7.857710017464737e-06, "loss": 0.0028, "step": 360 }, { "epoch": 4.161849710982659, "eval_loss": 0.00829649344086647, "eval_runtime": 10.0, "eval_samples_per_second": 5.0, "eval_steps_per_second": 1.3, "step": 360 }, { "epoch": 4.173410404624278, "grad_norm": 0.017382489517331123, "learning_rate": 7.640668900196984e-06, "loss": 0.0041, "step": 361 }, { "epoch": 4.184971098265896, "grad_norm": 0.014215102419257164, "learning_rate": 7.426419198674772e-06, "loss": 0.0025, "step": 362 }, { "epoch": 4.196531791907514, "grad_norm": 0.01837097853422165, "learning_rate": 7.214975031628857e-06, "loss": 0.0036, "step": 363 }, { "epoch": 4.208092485549133, "grad_norm": 0.01614932157099247, "learning_rate": 7.006350332909495e-06, "loss": 0.0027, "step": 364 }, { "epoch": 4.219653179190751, "grad_norm": 0.016307765617966652, "learning_rate": 6.800558850568295e-06, "loss": 0.0028, "step": 365 }, { "epoch": 4.219653179190751, "eval_loss": 0.008427992463111877, "eval_runtime": 10.0341, "eval_samples_per_second": 4.983, "eval_steps_per_second": 1.296, "step": 365 }, { "epoch": 4.23121387283237, "grad_norm": 0.024037158116698265, "learning_rate": 6.5976141459521355e-06, "loss": 0.0026, "step": 366 }, { "epoch": 4.242774566473988, "grad_norm": 0.018489433452486992, "learning_rate": 6.397529592809614e-06, "loss": 0.0034, "step": 367 }, { "epoch": 4.254335260115607, "grad_norm": 0.013452818617224693, "learning_rate": 6.2003183764096695e-06, "loss": 0.0025, "step": 368 }, { "epoch": 4.265895953757226, "grad_norm": 0.017129331827163696, "learning_rate": 6.005993492672657e-06, "loss": 0.0037, "step": 369 }, { "epoch": 4.277456647398844, "grad_norm": 0.01633543334901333, "learning_rate": 5.814567747314048e-06, "loss": 0.003, "step": 370 }, { "epoch": 4.277456647398844, "eval_loss": 0.008522397838532925, "eval_runtime": 10.0223, "eval_samples_per_second": 4.989, "eval_steps_per_second": 1.297, "step": 370 }, { "epoch": 4.289017341040463, "grad_norm": 0.013097944669425488, "learning_rate": 5.626053755000421e-06, "loss": 0.0021, "step": 371 }, { "epoch": 4.300578034682081, "grad_norm": 0.016527358442544937, "learning_rate": 5.440463938518303e-06, "loss": 0.003, "step": 372 }, { "epoch": 4.312138728323699, "grad_norm": 0.015918167307972908, "learning_rate": 5.257810527955409e-06, "loss": 0.0027, "step": 373 }, { "epoch": 4.323699421965318, "grad_norm": 0.01487621571868658, "learning_rate": 5.078105559894791e-06, "loss": 0.0025, "step": 374 }, { "epoch": 4.335260115606936, "grad_norm": 0.01791987195611, "learning_rate": 4.901360876621597e-06, "loss": 0.0033, "step": 375 }, { "epoch": 4.335260115606936, "eval_loss": 0.008536826819181442, "eval_runtime": 10.0084, "eval_samples_per_second": 4.996, "eval_steps_per_second": 1.299, "step": 375 }, { "epoch": 4.3468208092485545, "grad_norm": 0.013367927633225918, "learning_rate": 4.727588125342669e-06, "loss": 0.0022, "step": 376 }, { "epoch": 4.358381502890174, "grad_norm": 0.013060873374342918, "learning_rate": 4.556798757419068e-06, "loss": 0.0025, "step": 377 }, { "epoch": 4.369942196531792, "grad_norm": 0.014499946497380733, "learning_rate": 4.389004027611404e-06, "loss": 0.0027, "step": 378 }, { "epoch": 4.381502890173411, "grad_norm": 0.01943526417016983, "learning_rate": 4.224214993338149e-06, "loss": 0.004, "step": 379 }, { "epoch": 4.393063583815029, "grad_norm": 0.019275827333331108, "learning_rate": 4.062442513947007e-06, "loss": 0.003, "step": 380 }, { "epoch": 4.393063583815029, "eval_loss": 0.008570068515837193, "eval_runtime": 10.0111, "eval_samples_per_second": 4.994, "eval_steps_per_second": 1.299, "step": 380 }, { "epoch": 4.404624277456647, "grad_norm": 0.01620578207075596, "learning_rate": 3.903697249999289e-06, "loss": 0.0031, "step": 381 }, { "epoch": 4.416184971098266, "grad_norm": 0.01429419033229351, "learning_rate": 3.7479896625674027e-06, "loss": 0.0026, "step": 382 }, { "epoch": 4.427745664739884, "grad_norm": 0.015918832272291183, "learning_rate": 3.595330012545445e-06, "loss": 0.0026, "step": 383 }, { "epoch": 4.4393063583815024, "grad_norm": 0.019388854503631592, "learning_rate": 3.445728359973094e-06, "loss": 0.0029, "step": 384 }, { "epoch": 4.4508670520231215, "grad_norm": 0.012549116276204586, "learning_rate": 3.299194563372604e-06, "loss": 0.0022, "step": 385 }, { "epoch": 4.4508670520231215, "eval_loss": 0.008573681116104126, "eval_runtime": 10.0229, "eval_samples_per_second": 4.989, "eval_steps_per_second": 1.297, "step": 385 }, { "epoch": 4.46242774566474, "grad_norm": 0.01509170513600111, "learning_rate": 3.1557382790991687e-06, "loss": 0.0026, "step": 386 }, { "epoch": 4.473988439306359, "grad_norm": 0.014067728072404861, "learning_rate": 3.0153689607045845e-06, "loss": 0.0025, "step": 387 }, { "epoch": 4.485549132947977, "grad_norm": 0.028946029022336006, "learning_rate": 2.878095858314278e-06, "loss": 0.0027, "step": 388 }, { "epoch": 4.497109826589595, "grad_norm": 0.01480345893651247, "learning_rate": 2.743928018017744e-06, "loss": 0.0025, "step": 389 }, { "epoch": 4.508670520231214, "grad_norm": 0.02504677325487137, "learning_rate": 2.6128742812723704e-06, "loss": 0.0028, "step": 390 }, { "epoch": 4.508670520231214, "eval_loss": 0.008553516119718552, "eval_runtime": 10.0039, "eval_samples_per_second": 4.998, "eval_steps_per_second": 1.299, "step": 390 }, { "epoch": 4.520231213872832, "grad_norm": 0.016089266166090965, "learning_rate": 2.4849432843208785e-06, "loss": 0.0033, "step": 391 }, { "epoch": 4.531791907514451, "grad_norm": 0.015426091849803925, "learning_rate": 2.3601434576221546e-06, "loss": 0.0029, "step": 392 }, { "epoch": 4.543352601156069, "grad_norm": 0.017408043146133423, "learning_rate": 2.238483025295707e-06, "loss": 0.0031, "step": 393 }, { "epoch": 4.554913294797688, "grad_norm": 0.017079446464776993, "learning_rate": 2.1199700045797077e-06, "loss": 0.0027, "step": 394 }, { "epoch": 4.566473988439307, "grad_norm": 0.016879020258784294, "learning_rate": 2.0046122053026694e-06, "loss": 0.0028, "step": 395 }, { "epoch": 4.566473988439307, "eval_loss": 0.008501513861119747, "eval_runtime": 10.036, "eval_samples_per_second": 4.982, "eval_steps_per_second": 1.295, "step": 395 }, { "epoch": 4.578034682080925, "grad_norm": 0.030611367896199226, "learning_rate": 1.8924172293688147e-06, "loss": 0.0032, "step": 396 }, { "epoch": 4.589595375722544, "grad_norm": 0.017865322530269623, "learning_rate": 1.7833924702570725e-06, "loss": 0.0035, "step": 397 }, { "epoch": 4.601156069364162, "grad_norm": 0.014931568875908852, "learning_rate": 1.6775451125338959e-06, "loss": 0.0025, "step": 398 }, { "epoch": 4.61271676300578, "grad_norm": 0.015031742863357067, "learning_rate": 1.5748821313798124e-06, "loss": 0.0024, "step": 399 }, { "epoch": 4.624277456647399, "grad_norm": 0.017130030319094658, "learning_rate": 1.4754102921297364e-06, "loss": 0.0031, "step": 400 }, { "epoch": 4.624277456647399, "eval_loss": 0.008478561416268349, "eval_runtime": 10.0323, "eval_samples_per_second": 4.984, "eval_steps_per_second": 1.296, "step": 400 }, { "epoch": 4.635838150289017, "grad_norm": 0.014694861136376858, "learning_rate": 1.3791361498271705e-06, "loss": 0.0024, "step": 401 }, { "epoch": 4.6473988439306355, "grad_norm": 0.013526085764169693, "learning_rate": 1.2860660487922616e-06, "loss": 0.0023, "step": 402 }, { "epoch": 4.658959537572255, "grad_norm": 0.04836466535925865, "learning_rate": 1.196206122203647e-06, "loss": 0.0024, "step": 403 }, { "epoch": 4.670520231213873, "grad_norm": 0.016095180064439774, "learning_rate": 1.1095622916943494e-06, "loss": 0.0033, "step": 404 }, { "epoch": 4.682080924855491, "grad_norm": 0.019678572192788124, "learning_rate": 1.0261402669615505e-06, "loss": 0.0038, "step": 405 }, { "epoch": 4.682080924855491, "eval_loss": 0.00843281950801611, "eval_runtime": 10.0323, "eval_samples_per_second": 4.984, "eval_steps_per_second": 1.296, "step": 405 }, { "epoch": 4.69364161849711, "grad_norm": 0.013817240484058857, "learning_rate": 9.459455453902866e-07, "loss": 0.0024, "step": 406 }, { "epoch": 4.705202312138728, "grad_norm": 0.0165092833340168, "learning_rate": 8.68983411691221e-07, "loss": 0.0029, "step": 407 }, { "epoch": 4.716763005780347, "grad_norm": 0.016268854960799217, "learning_rate": 7.952589375523567e-07, "loss": 0.0028, "step": 408 }, { "epoch": 4.728323699421965, "grad_norm": 0.013906327076256275, "learning_rate": 7.247769813048644e-07, "loss": 0.0025, "step": 409 }, { "epoch": 4.7398843930635834, "grad_norm": 0.0136026069521904, "learning_rate": 6.57542187602872e-07, "loss": 0.0024, "step": 410 }, { "epoch": 4.7398843930635834, "eval_loss": 0.00842086412012577, "eval_runtime": 10.0656, "eval_samples_per_second": 4.967, "eval_steps_per_second": 1.292, "step": 410 }, { "epoch": 4.7514450867052025, "grad_norm": 0.022275064140558243, "learning_rate": 5.935589871174208e-07, "loss": 0.0028, "step": 411 }, { "epoch": 4.763005780346821, "grad_norm": 0.016609348356723785, "learning_rate": 5.328315962444874e-07, "loss": 0.0029, "step": 412 }, { "epoch": 4.77456647398844, "grad_norm": 0.015365025959908962, "learning_rate": 4.753640168271456e-07, "loss": 0.0029, "step": 413 }, { "epoch": 4.786127167630058, "grad_norm": 0.014568125829100609, "learning_rate": 4.2116003589179887e-07, "loss": 0.0024, "step": 414 }, { "epoch": 4.797687861271676, "grad_norm": 0.014286377467215061, "learning_rate": 3.702232253986804e-07, "loss": 0.0024, "step": 415 }, { "epoch": 4.797687861271676, "eval_loss": 0.008439392782747746, "eval_runtime": 10.0373, "eval_samples_per_second": 4.981, "eval_steps_per_second": 1.295, "step": 415 }, { "epoch": 4.809248554913295, "grad_norm": 0.015557860024273396, "learning_rate": 3.2255694200643006e-07, "loss": 0.0027, "step": 416 }, { "epoch": 4.820809248554913, "grad_norm": 0.01427092682570219, "learning_rate": 2.78164326850916e-07, "loss": 0.0023, "step": 417 }, { "epoch": 4.832369942196532, "grad_norm": 0.01711309514939785, "learning_rate": 2.3704830533821108e-07, "loss": 0.0031, "step": 418 }, { "epoch": 4.84393063583815, "grad_norm": 0.015366033650934696, "learning_rate": 1.9921158695184738e-07, "loss": 0.0026, "step": 419 }, { "epoch": 4.855491329479769, "grad_norm": 0.013661878183484077, "learning_rate": 1.6465666507425315e-07, "loss": 0.0024, "step": 420 }, { "epoch": 4.855491329479769, "eval_loss": 0.008424060419201851, "eval_runtime": 10.0452, "eval_samples_per_second": 4.978, "eval_steps_per_second": 1.294, "step": 420 }, { "epoch": 4.867052023121388, "grad_norm": 0.01569373905658722, "learning_rate": 1.333858168224178e-07, "loss": 0.0027, "step": 421 }, { "epoch": 4.878612716763006, "grad_norm": 0.02022898755967617, "learning_rate": 1.0540110289786742e-07, "loss": 0.0032, "step": 422 }, { "epoch": 4.890173410404624, "grad_norm": 0.013800419867038727, "learning_rate": 8.07043674508623e-08, "loss": 0.0024, "step": 423 }, { "epoch": 4.901734104046243, "grad_norm": 0.017323607578873634, "learning_rate": 5.929723795884967e-08, "loss": 0.0029, "step": 424 }, { "epoch": 4.913294797687861, "grad_norm": 0.014345725998282433, "learning_rate": 4.1181125119221787e-08, "loss": 0.0026, "step": 425 }, { "epoch": 4.913294797687861, "eval_loss": 0.008420780301094055, "eval_runtime": 10.0258, "eval_samples_per_second": 4.987, "eval_steps_per_second": 1.297, "step": 425 }, { "epoch": 4.924855491329479, "grad_norm": 0.014356693252921104, "learning_rate": 2.6357222756384636e-08, "loss": 0.0025, "step": 426 }, { "epoch": 4.936416184971098, "grad_norm": 0.014968520030379295, "learning_rate": 1.482650774303207e-08, "loss": 0.0028, "step": 427 }, { "epoch": 4.9479768786127165, "grad_norm": 0.02624007686972618, "learning_rate": 6.589739935819461e-09, "loss": 0.0033, "step": 428 }, { "epoch": 4.959537572254336, "grad_norm": 0.016497237607836723, "learning_rate": 1.6474621252704493e-09, "loss": 0.0031, "step": 429 }, { "epoch": 4.971098265895954, "grad_norm": 0.01508865226060152, "learning_rate": 0.0, "loss": 0.0029, "step": 430 }, { "epoch": 4.971098265895954, "eval_loss": 0.008429310284554958, "eval_runtime": 10.0331, "eval_samples_per_second": 4.983, "eval_steps_per_second": 1.296, "step": 430 }, { "epoch": 4.971098265895954, "step": 430, "total_flos": 1.0207257247425331e+18, "train_loss": 0.0087554031284526, "train_runtime": 8412.5192, "train_samples_per_second": 1.645, "train_steps_per_second": 0.051 } ], "logging_steps": 1, "max_steps": 430, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.0207257247425331e+18, "train_batch_size": 4, "trial_name": null, "trial_params": null }