{ "best_metric": null, "best_model_checkpoint": null, "epoch": 10.0, "eval_steps": 500, "global_step": 2950, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03389830508474576, "grad_norm": 4.791459083557129, "learning_rate": 6.7567567567567575e-06, "loss": 0.2753, "step": 10 }, { "epoch": 0.06779661016949153, "grad_norm": 2.2864952087402344, "learning_rate": 1.3513513513513515e-05, "loss": 0.1531, "step": 20 }, { "epoch": 0.1016949152542373, "grad_norm": 0.3792467415332794, "learning_rate": 2.0270270270270273e-05, "loss": 0.0584, "step": 30 }, { "epoch": 0.13559322033898305, "grad_norm": 0.4453708231449127, "learning_rate": 2.702702702702703e-05, "loss": 0.035, "step": 40 }, { "epoch": 0.1694915254237288, "grad_norm": 0.6060521602630615, "learning_rate": 3.3783783783783784e-05, "loss": 0.0155, "step": 50 }, { "epoch": 0.2033898305084746, "grad_norm": 0.4304051995277405, "learning_rate": 4.0540540540540545e-05, "loss": 0.0089, "step": 60 }, { "epoch": 0.23728813559322035, "grad_norm": 0.43952998518943787, "learning_rate": 4.72972972972973e-05, "loss": 0.0089, "step": 70 }, { "epoch": 0.2711864406779661, "grad_norm": 0.3579324781894684, "learning_rate": 5.405405405405406e-05, "loss": 0.0073, "step": 80 }, { "epoch": 0.3050847457627119, "grad_norm": 0.221815824508667, "learning_rate": 6.0810810810810814e-05, "loss": 0.0061, "step": 90 }, { "epoch": 0.3389830508474576, "grad_norm": 0.4653712809085846, "learning_rate": 6.756756756756757e-05, "loss": 0.0046, "step": 100 }, { "epoch": 0.3728813559322034, "grad_norm": 0.20366229116916656, "learning_rate": 7.432432432432433e-05, "loss": 0.0047, "step": 110 }, { "epoch": 0.4067796610169492, "grad_norm": 0.30185437202453613, "learning_rate": 8.108108108108109e-05, "loss": 0.0048, "step": 120 }, { "epoch": 0.4406779661016949, "grad_norm": 0.3719215989112854, "learning_rate": 8.783783783783784e-05, "loss": 0.0045, "step": 130 }, { "epoch": 0.4745762711864407, "grad_norm": 0.14779077470302582, "learning_rate": 9.45945945945946e-05, "loss": 0.0058, "step": 140 }, { "epoch": 0.5084745762711864, "grad_norm": 0.525753915309906, "learning_rate": 9.999987429188866e-05, "loss": 0.0043, "step": 150 }, { "epoch": 0.5423728813559322, "grad_norm": 0.3979560136795044, "learning_rate": 9.999547457436221e-05, "loss": 0.0051, "step": 160 }, { "epoch": 0.576271186440678, "grad_norm": 0.42477351427078247, "learning_rate": 9.998479008335589e-05, "loss": 0.0037, "step": 170 }, { "epoch": 0.6101694915254238, "grad_norm": 0.19697535037994385, "learning_rate": 9.996782216198338e-05, "loss": 0.0052, "step": 180 }, { "epoch": 0.6440677966101694, "grad_norm": 0.23994311690330505, "learning_rate": 9.994457294322858e-05, "loss": 0.0043, "step": 190 }, { "epoch": 0.6779661016949152, "grad_norm": 0.3949716091156006, "learning_rate": 9.991504534967746e-05, "loss": 0.0043, "step": 200 }, { "epoch": 0.711864406779661, "grad_norm": 0.5126088261604309, "learning_rate": 9.987924309315075e-05, "loss": 0.0029, "step": 210 }, { "epoch": 0.7457627118644068, "grad_norm": 0.34605440497398376, "learning_rate": 9.983717067423721e-05, "loss": 0.0024, "step": 220 }, { "epoch": 0.7796610169491526, "grad_norm": 0.3350368142127991, "learning_rate": 9.978883338172798e-05, "loss": 0.0028, "step": 230 }, { "epoch": 0.8135593220338984, "grad_norm": 0.4186588525772095, "learning_rate": 9.973423729195168e-05, "loss": 0.0037, "step": 240 }, { "epoch": 0.847457627118644, "grad_norm": 0.4036902189254761, "learning_rate": 9.967338926801067e-05, "loss": 0.0031, "step": 250 }, { "epoch": 0.8813559322033898, "grad_norm": 0.3429702818393707, "learning_rate": 9.960629695891814e-05, "loss": 0.0035, "step": 260 }, { "epoch": 0.9152542372881356, "grad_norm": 0.22962354123592377, "learning_rate": 9.953296879863677e-05, "loss": 0.0039, "step": 270 }, { "epoch": 0.9491525423728814, "grad_norm": 0.3024398386478424, "learning_rate": 9.945341400501838e-05, "loss": 0.0025, "step": 280 }, { "epoch": 0.9830508474576272, "grad_norm": 0.29498374462127686, "learning_rate": 9.936764257864525e-05, "loss": 0.0022, "step": 290 }, { "epoch": 1.0169491525423728, "grad_norm": 0.2633371651172638, "learning_rate": 9.927566530157298e-05, "loss": 0.0046, "step": 300 }, { "epoch": 1.0508474576271187, "grad_norm": 0.1469630002975464, "learning_rate": 9.917749373597504e-05, "loss": 0.003, "step": 310 }, { "epoch": 1.0847457627118644, "grad_norm": 0.2827948033809662, "learning_rate": 9.907314022268946e-05, "loss": 0.0025, "step": 320 }, { "epoch": 1.11864406779661, "grad_norm": 0.21969269216060638, "learning_rate": 9.896261787966739e-05, "loss": 0.0029, "step": 330 }, { "epoch": 1.152542372881356, "grad_norm": 0.20057785511016846, "learning_rate": 9.884594060032406e-05, "loss": 0.0021, "step": 340 }, { "epoch": 1.1864406779661016, "grad_norm": 0.16977977752685547, "learning_rate": 9.872312305179243e-05, "loss": 0.0029, "step": 350 }, { "epoch": 1.2203389830508475, "grad_norm": 0.25367122888565063, "learning_rate": 9.859418067307928e-05, "loss": 0.0025, "step": 360 }, { "epoch": 1.2542372881355932, "grad_norm": 0.25058019161224365, "learning_rate": 9.84591296731245e-05, "loss": 0.0024, "step": 370 }, { "epoch": 1.288135593220339, "grad_norm": 0.15482617914676666, "learning_rate": 9.831798702876352e-05, "loss": 0.0027, "step": 380 }, { "epoch": 1.3220338983050848, "grad_norm": 0.4145447313785553, "learning_rate": 9.817077048259314e-05, "loss": 0.0034, "step": 390 }, { "epoch": 1.3559322033898304, "grad_norm": 0.21151414513587952, "learning_rate": 9.801749854074122e-05, "loss": 0.0031, "step": 400 }, { "epoch": 1.3898305084745763, "grad_norm": 0.19668647646903992, "learning_rate": 9.785819047054035e-05, "loss": 0.0015, "step": 410 }, { "epoch": 1.423728813559322, "grad_norm": 0.12844844162464142, "learning_rate": 9.769286629810572e-05, "loss": 0.0021, "step": 420 }, { "epoch": 1.457627118644068, "grad_norm": 0.10224123299121857, "learning_rate": 9.752154680581783e-05, "loss": 0.0017, "step": 430 }, { "epoch": 1.4915254237288136, "grad_norm": 0.13156643509864807, "learning_rate": 9.73442535297099e-05, "loss": 0.0022, "step": 440 }, { "epoch": 1.5254237288135593, "grad_norm": 0.09552246332168579, "learning_rate": 9.716100875676071e-05, "loss": 0.0032, "step": 450 }, { "epoch": 1.559322033898305, "grad_norm": 0.2560286819934845, "learning_rate": 9.697183552209288e-05, "loss": 0.004, "step": 460 }, { "epoch": 1.5932203389830508, "grad_norm": 0.17713002860546112, "learning_rate": 9.677675760607734e-05, "loss": 0.0024, "step": 470 }, { "epoch": 1.6271186440677967, "grad_norm": 0.14444805681705475, "learning_rate": 9.657579953134383e-05, "loss": 0.0035, "step": 480 }, { "epoch": 1.6610169491525424, "grad_norm": 0.26531147956848145, "learning_rate": 9.636898655969837e-05, "loss": 0.0023, "step": 490 }, { "epoch": 1.694915254237288, "grad_norm": 0.1393326222896576, "learning_rate": 9.615634468894752e-05, "loss": 0.001, "step": 500 }, { "epoch": 1.7288135593220337, "grad_norm": 0.22181938588619232, "learning_rate": 9.593790064963046e-05, "loss": 0.0018, "step": 510 }, { "epoch": 1.7627118644067796, "grad_norm": 0.1423530876636505, "learning_rate": 9.571368190165863e-05, "loss": 0.0029, "step": 520 }, { "epoch": 1.7966101694915255, "grad_norm": 0.1219254806637764, "learning_rate": 9.548371663086392e-05, "loss": 0.0022, "step": 530 }, { "epoch": 1.8305084745762712, "grad_norm": 0.1542597860097885, "learning_rate": 9.524803374545548e-05, "loss": 0.0035, "step": 540 }, { "epoch": 1.8644067796610169, "grad_norm": 0.16393965482711792, "learning_rate": 9.500666287238573e-05, "loss": 0.0021, "step": 550 }, { "epoch": 1.8983050847457628, "grad_norm": 0.35059654712677, "learning_rate": 9.475963435362614e-05, "loss": 0.0027, "step": 560 }, { "epoch": 1.9322033898305084, "grad_norm": 0.20335566997528076, "learning_rate": 9.450697924235294e-05, "loss": 0.0023, "step": 570 }, { "epoch": 1.9661016949152543, "grad_norm": 0.15571418404579163, "learning_rate": 9.424872929904358e-05, "loss": 0.0012, "step": 580 }, { "epoch": 2.0, "grad_norm": 0.29340890049934387, "learning_rate": 9.398491698748422e-05, "loss": 0.0022, "step": 590 }, { "epoch": 2.0338983050847457, "grad_norm": 0.2681172788143158, "learning_rate": 9.371557547068878e-05, "loss": 0.0013, "step": 600 }, { "epoch": 2.0677966101694913, "grad_norm": 0.16841526329517365, "learning_rate": 9.344073860673015e-05, "loss": 0.0012, "step": 610 }, { "epoch": 2.1016949152542375, "grad_norm": 0.07565703988075256, "learning_rate": 9.316044094448392e-05, "loss": 0.0013, "step": 620 }, { "epoch": 2.135593220338983, "grad_norm": 0.1882319450378418, "learning_rate": 9.287471771928555e-05, "loss": 0.0022, "step": 630 }, { "epoch": 2.169491525423729, "grad_norm": 0.07069575041532516, "learning_rate": 9.25836048485008e-05, "loss": 0.0007, "step": 640 }, { "epoch": 2.2033898305084745, "grad_norm": 0.16358712315559387, "learning_rate": 9.228713892701077e-05, "loss": 0.003, "step": 650 }, { "epoch": 2.23728813559322, "grad_norm": 0.13663357496261597, "learning_rate": 9.198535722261181e-05, "loss": 0.0019, "step": 660 }, { "epoch": 2.2711864406779663, "grad_norm": 0.13056139647960663, "learning_rate": 9.167829767133047e-05, "loss": 0.0013, "step": 670 }, { "epoch": 2.305084745762712, "grad_norm": 0.12309998273849487, "learning_rate": 9.136599887265483e-05, "loss": 0.0016, "step": 680 }, { "epoch": 2.3389830508474576, "grad_norm": 0.13602758944034576, "learning_rate": 9.104850008468221e-05, "loss": 0.0007, "step": 690 }, { "epoch": 2.3728813559322033, "grad_norm": 0.08036898076534271, "learning_rate": 9.072584121918425e-05, "loss": 0.0023, "step": 700 }, { "epoch": 2.406779661016949, "grad_norm": 0.08389495313167572, "learning_rate": 9.039806283658962e-05, "loss": 0.0014, "step": 710 }, { "epoch": 2.440677966101695, "grad_norm": 0.12567836046218872, "learning_rate": 9.006520614088535e-05, "loss": 0.001, "step": 720 }, { "epoch": 2.4745762711864407, "grad_norm": 0.15499231219291687, "learning_rate": 8.972731297443722e-05, "loss": 0.0018, "step": 730 }, { "epoch": 2.5084745762711864, "grad_norm": 0.09496314823627472, "learning_rate": 8.938442581272983e-05, "loss": 0.0031, "step": 740 }, { "epoch": 2.542372881355932, "grad_norm": 0.1453184187412262, "learning_rate": 8.903658775902723e-05, "loss": 0.0014, "step": 750 }, { "epoch": 2.576271186440678, "grad_norm": 0.11160401254892349, "learning_rate": 8.868384253895445e-05, "loss": 0.001, "step": 760 }, { "epoch": 2.610169491525424, "grad_norm": 0.10584841668605804, "learning_rate": 8.832623449500095e-05, "loss": 0.0016, "step": 770 }, { "epoch": 2.6440677966101696, "grad_norm": 0.094113789498806, "learning_rate": 8.796380858094643e-05, "loss": 0.0013, "step": 780 }, { "epoch": 2.6779661016949152, "grad_norm": 0.21183228492736816, "learning_rate": 8.759661035620992e-05, "loss": 0.0015, "step": 790 }, { "epoch": 2.711864406779661, "grad_norm": 0.14965838193893433, "learning_rate": 8.722468598012245e-05, "loss": 0.0018, "step": 800 }, { "epoch": 2.7457627118644066, "grad_norm": 0.08837207406759262, "learning_rate": 8.684808220612478e-05, "loss": 0.0008, "step": 810 }, { "epoch": 2.7796610169491527, "grad_norm": 0.11261831223964691, "learning_rate": 8.646684637588991e-05, "loss": 0.0022, "step": 820 }, { "epoch": 2.8135593220338984, "grad_norm": 0.221408873796463, "learning_rate": 8.60810264133721e-05, "loss": 0.0014, "step": 830 }, { "epoch": 2.847457627118644, "grad_norm": 0.37261417508125305, "learning_rate": 8.56906708187824e-05, "loss": 0.0023, "step": 840 }, { "epoch": 2.8813559322033897, "grad_norm": 0.12965640425682068, "learning_rate": 8.529582866249186e-05, "loss": 0.002, "step": 850 }, { "epoch": 2.915254237288136, "grad_norm": 0.12147118151187897, "learning_rate": 8.489654957886306e-05, "loss": 0.0015, "step": 860 }, { "epoch": 2.9491525423728815, "grad_norm": 0.13188952207565308, "learning_rate": 8.449288376001075e-05, "loss": 0.0013, "step": 870 }, { "epoch": 2.983050847457627, "grad_norm": 0.12438294291496277, "learning_rate": 8.40848819494923e-05, "loss": 0.0013, "step": 880 }, { "epoch": 3.016949152542373, "grad_norm": 0.309495747089386, "learning_rate": 8.367259543592892e-05, "loss": 0.0012, "step": 890 }, { "epoch": 3.0508474576271185, "grad_norm": 0.24989376962184906, "learning_rate": 8.325607604655839e-05, "loss": 0.001, "step": 900 }, { "epoch": 3.084745762711864, "grad_norm": 0.1818067878484726, "learning_rate": 8.283537614071988e-05, "loss": 0.001, "step": 910 }, { "epoch": 3.1186440677966103, "grad_norm": 0.13010039925575256, "learning_rate": 8.241054860327216e-05, "loss": 0.0013, "step": 920 }, { "epoch": 3.152542372881356, "grad_norm": 0.15505361557006836, "learning_rate": 8.198164683794556e-05, "loss": 0.0029, "step": 930 }, { "epoch": 3.1864406779661016, "grad_norm": 0.07065195590257645, "learning_rate": 8.154872476062868e-05, "loss": 0.0014, "step": 940 }, { "epoch": 3.2203389830508473, "grad_norm": 0.28758659958839417, "learning_rate": 8.111183679259095e-05, "loss": 0.002, "step": 950 }, { "epoch": 3.2542372881355934, "grad_norm": 0.18301035463809967, "learning_rate": 8.067103785364139e-05, "loss": 0.0014, "step": 960 }, { "epoch": 3.288135593220339, "grad_norm": 0.09719489514827728, "learning_rate": 8.022638335522483e-05, "loss": 0.0025, "step": 970 }, { "epoch": 3.3220338983050848, "grad_norm": 0.2065349668264389, "learning_rate": 7.977792919345633e-05, "loss": 0.0014, "step": 980 }, { "epoch": 3.3559322033898304, "grad_norm": 0.10643105208873749, "learning_rate": 7.932573174209463e-05, "loss": 0.0017, "step": 990 }, { "epoch": 3.389830508474576, "grad_norm": 0.11610566079616547, "learning_rate": 7.886984784545566e-05, "loss": 0.0008, "step": 1000 }, { "epoch": 3.423728813559322, "grad_norm": 0.11425777524709702, "learning_rate": 7.84103348112667e-05, "loss": 0.0017, "step": 1010 }, { "epoch": 3.457627118644068, "grad_norm": 0.08603143692016602, "learning_rate": 7.794725040346251e-05, "loss": 0.0011, "step": 1020 }, { "epoch": 3.4915254237288136, "grad_norm": 0.14157338440418243, "learning_rate": 7.748065283492396e-05, "loss": 0.0019, "step": 1030 }, { "epoch": 3.5254237288135593, "grad_norm": 0.3015165627002716, "learning_rate": 7.701060076016024e-05, "loss": 0.0016, "step": 1040 }, { "epoch": 3.559322033898305, "grad_norm": 0.09349218755960464, "learning_rate": 7.653715326793569e-05, "loss": 0.0009, "step": 1050 }, { "epoch": 3.593220338983051, "grad_norm": 0.13891376554965973, "learning_rate": 7.606036987384184e-05, "loss": 0.0009, "step": 1060 }, { "epoch": 3.6271186440677967, "grad_norm": 0.10969799011945724, "learning_rate": 7.558031051281592e-05, "loss": 0.0016, "step": 1070 }, { "epoch": 3.6610169491525424, "grad_norm": 0.16311612725257874, "learning_rate": 7.509703553160666e-05, "loss": 0.0015, "step": 1080 }, { "epoch": 3.694915254237288, "grad_norm": 0.14392238855361938, "learning_rate": 7.461060568118822e-05, "loss": 0.0009, "step": 1090 }, { "epoch": 3.7288135593220337, "grad_norm": 0.1457260549068451, "learning_rate": 7.412108210912345e-05, "loss": 0.0013, "step": 1100 }, { "epoch": 3.7627118644067794, "grad_norm": 0.16048628091812134, "learning_rate": 7.362852635187721e-05, "loss": 0.0017, "step": 1110 }, { "epoch": 3.7966101694915255, "grad_norm": 0.18469738960266113, "learning_rate": 7.31330003270808e-05, "loss": 0.0014, "step": 1120 }, { "epoch": 3.830508474576271, "grad_norm": 0.180945485830307, "learning_rate": 7.263456632574845e-05, "loss": 0.0015, "step": 1130 }, { "epoch": 3.864406779661017, "grad_norm": 0.14967080950737, "learning_rate": 7.213328700444696e-05, "loss": 0.0013, "step": 1140 }, { "epoch": 3.898305084745763, "grad_norm": 0.25463271141052246, "learning_rate": 7.162922537741937e-05, "loss": 0.0023, "step": 1150 }, { "epoch": 3.9322033898305087, "grad_norm": 0.24937133491039276, "learning_rate": 7.112244480866356e-05, "loss": 0.0018, "step": 1160 }, { "epoch": 3.9661016949152543, "grad_norm": 0.1597055047750473, "learning_rate": 7.061300900396703e-05, "loss": 0.0032, "step": 1170 }, { "epoch": 4.0, "grad_norm": 0.14016273617744446, "learning_rate": 7.010098200289859e-05, "loss": 0.0014, "step": 1180 }, { "epoch": 4.033898305084746, "grad_norm": 0.11648664623498917, "learning_rate": 6.958642817075823e-05, "loss": 0.001, "step": 1190 }, { "epoch": 4.067796610169491, "grad_norm": 0.23266953229904175, "learning_rate": 6.906941219048584e-05, "loss": 0.001, "step": 1200 }, { "epoch": 4.101694915254237, "grad_norm": 0.11199619621038437, "learning_rate": 6.854999905453021e-05, "loss": 0.0008, "step": 1210 }, { "epoch": 4.135593220338983, "grad_norm": 0.09246116131544113, "learning_rate": 6.802825405667905e-05, "loss": 0.001, "step": 1220 }, { "epoch": 4.169491525423728, "grad_norm": 0.08600598573684692, "learning_rate": 6.750424278385103e-05, "loss": 0.0008, "step": 1230 }, { "epoch": 4.203389830508475, "grad_norm": 0.09983401745557785, "learning_rate": 6.697803110785115e-05, "loss": 0.0009, "step": 1240 }, { "epoch": 4.237288135593221, "grad_norm": 0.09594856947660446, "learning_rate": 6.644968517709009e-05, "loss": 0.0015, "step": 1250 }, { "epoch": 4.271186440677966, "grad_norm": 0.08599961549043655, "learning_rate": 6.591927140826902e-05, "loss": 0.0007, "step": 1260 }, { "epoch": 4.305084745762712, "grad_norm": 0.16801735758781433, "learning_rate": 6.538685647803049e-05, "loss": 0.001, "step": 1270 }, { "epoch": 4.338983050847458, "grad_norm": 0.04569881409406662, "learning_rate": 6.485250731457678e-05, "loss": 0.0007, "step": 1280 }, { "epoch": 4.372881355932203, "grad_norm": 0.04355693981051445, "learning_rate": 6.431629108925646e-05, "loss": 0.0006, "step": 1290 }, { "epoch": 4.406779661016949, "grad_norm": 0.1598712056875229, "learning_rate": 6.377827520812061e-05, "loss": 0.0013, "step": 1300 }, { "epoch": 4.440677966101695, "grad_norm": 0.1439727395772934, "learning_rate": 6.323852730344935e-05, "loss": 0.0006, "step": 1310 }, { "epoch": 4.47457627118644, "grad_norm": 0.07059933245182037, "learning_rate": 6.269711522525006e-05, "loss": 0.0004, "step": 1320 }, { "epoch": 4.508474576271187, "grad_norm": 0.2104925960302353, "learning_rate": 6.215410703272805e-05, "loss": 0.0019, "step": 1330 }, { "epoch": 4.5423728813559325, "grad_norm": 0.3595544993877411, "learning_rate": 6.160957098573119e-05, "loss": 0.0026, "step": 1340 }, { "epoch": 4.576271186440678, "grad_norm": 0.09160851687192917, "learning_rate": 6.10635755361691e-05, "loss": 0.0012, "step": 1350 }, { "epoch": 4.610169491525424, "grad_norm": 0.15141859650611877, "learning_rate": 6.05161893194083e-05, "loss": 0.0012, "step": 1360 }, { "epoch": 4.6440677966101696, "grad_norm": 0.24170026183128357, "learning_rate": 5.9967481145644265e-05, "loss": 0.0009, "step": 1370 }, { "epoch": 4.677966101694915, "grad_norm": 0.12173440307378769, "learning_rate": 5.941751999125149e-05, "loss": 0.0009, "step": 1380 }, { "epoch": 4.711864406779661, "grad_norm": 0.2746393084526062, "learning_rate": 5.886637499011278e-05, "loss": 0.0008, "step": 1390 }, { "epoch": 4.745762711864407, "grad_norm": 0.06491386145353317, "learning_rate": 5.831411542492854e-05, "loss": 0.0006, "step": 1400 }, { "epoch": 4.779661016949152, "grad_norm": 0.10497017949819565, "learning_rate": 5.7760810718507487e-05, "loss": 0.0013, "step": 1410 }, { "epoch": 4.813559322033898, "grad_norm": 0.08166036754846573, "learning_rate": 5.720653042503978e-05, "loss": 0.0007, "step": 1420 }, { "epoch": 4.847457627118644, "grad_norm": 0.08851595968008041, "learning_rate": 5.665134422135354e-05, "loss": 0.0011, "step": 1430 }, { "epoch": 4.88135593220339, "grad_norm": 0.07782018929719925, "learning_rate": 5.6095321898156016e-05, "loss": 0.0014, "step": 1440 }, { "epoch": 4.915254237288136, "grad_norm": 0.06927744299173355, "learning_rate": 5.553853335126039e-05, "loss": 0.0004, "step": 1450 }, { "epoch": 4.9491525423728815, "grad_norm": 0.09779927879571915, "learning_rate": 5.498104857279941e-05, "loss": 0.0006, "step": 1460 }, { "epoch": 4.983050847457627, "grad_norm": 0.18358203768730164, "learning_rate": 5.4422937642426906e-05, "loss": 0.0009, "step": 1470 }, { "epoch": 5.016949152542373, "grad_norm": 0.1410626918077469, "learning_rate": 5.3864270718508305e-05, "loss": 0.0009, "step": 1480 }, { "epoch": 5.0508474576271185, "grad_norm": 0.17358587682247162, "learning_rate": 5.330511802930117e-05, "loss": 0.0011, "step": 1490 }, { "epoch": 5.084745762711864, "grad_norm": 0.1470521092414856, "learning_rate": 5.274554986412716e-05, "loss": 0.0011, "step": 1500 }, { "epoch": 5.11864406779661, "grad_norm": 0.1285671442747116, "learning_rate": 5.218563656453609e-05, "loss": 0.0012, "step": 1510 }, { "epoch": 5.1525423728813555, "grad_norm": 0.10445746779441833, "learning_rate": 5.162544851546349e-05, "loss": 0.0017, "step": 1520 }, { "epoch": 5.186440677966102, "grad_norm": 0.14065216481685638, "learning_rate": 5.106505613638284e-05, "loss": 0.0022, "step": 1530 }, { "epoch": 5.220338983050848, "grad_norm": 0.09856051951646805, "learning_rate": 5.0504529872453256e-05, "loss": 0.002, "step": 1540 }, { "epoch": 5.254237288135593, "grad_norm": 0.1747041493654251, "learning_rate": 4.9943940185664124e-05, "loss": 0.001, "step": 1550 }, { "epoch": 5.288135593220339, "grad_norm": 0.08963986486196518, "learning_rate": 4.9383357545977497e-05, "loss": 0.0007, "step": 1560 }, { "epoch": 5.322033898305085, "grad_norm": 0.09163827449083328, "learning_rate": 4.882285242246957e-05, "loss": 0.001, "step": 1570 }, { "epoch": 5.3559322033898304, "grad_norm": 0.08859282732009888, "learning_rate": 4.8262495274472225e-05, "loss": 0.0006, "step": 1580 }, { "epoch": 5.389830508474576, "grad_norm": 0.06251803785562515, "learning_rate": 4.770235654271582e-05, "loss": 0.0004, "step": 1590 }, { "epoch": 5.423728813559322, "grad_norm": 0.06271404027938843, "learning_rate": 4.7142506640474274e-05, "loss": 0.0005, "step": 1600 }, { "epoch": 5.4576271186440675, "grad_norm": 0.16893021762371063, "learning_rate": 4.658301594471367e-05, "loss": 0.0009, "step": 1610 }, { "epoch": 5.491525423728813, "grad_norm": 0.04200820252299309, "learning_rate": 4.602395478724539e-05, "loss": 0.0006, "step": 1620 }, { "epoch": 5.52542372881356, "grad_norm": 0.030414093285799026, "learning_rate": 4.546539344588486e-05, "loss": 0.0005, "step": 1630 }, { "epoch": 5.559322033898305, "grad_norm": 0.10116266459226608, "learning_rate": 4.490740213561727e-05, "loss": 0.0011, "step": 1640 }, { "epoch": 5.593220338983051, "grad_norm": 0.12049903720617294, "learning_rate": 4.435005099977093e-05, "loss": 0.0006, "step": 1650 }, { "epoch": 5.627118644067797, "grad_norm": 0.0699400007724762, "learning_rate": 4.379341010119992e-05, "loss": 0.0005, "step": 1660 }, { "epoch": 5.661016949152542, "grad_norm": 0.06926112622022629, "learning_rate": 4.323754941347654e-05, "loss": 0.0006, "step": 1670 }, { "epoch": 5.694915254237288, "grad_norm": 0.06657879799604416, "learning_rate": 4.268253881209532e-05, "loss": 0.0005, "step": 1680 }, { "epoch": 5.728813559322034, "grad_norm": 0.04170970991253853, "learning_rate": 4.2128448065689065e-05, "loss": 0.0005, "step": 1690 }, { "epoch": 5.762711864406779, "grad_norm": 0.09546169638633728, "learning_rate": 4.157534682725856e-05, "loss": 0.0004, "step": 1700 }, { "epoch": 5.796610169491525, "grad_norm": 0.14321397244930267, "learning_rate": 4.102330462541662e-05, "loss": 0.0005, "step": 1710 }, { "epoch": 5.830508474576272, "grad_norm": 0.06389990448951721, "learning_rate": 4.047239085564794e-05, "loss": 0.0005, "step": 1720 }, { "epoch": 5.864406779661017, "grad_norm": 0.03475102782249451, "learning_rate": 3.9922674771585546e-05, "loss": 0.0003, "step": 1730 }, { "epoch": 5.898305084745763, "grad_norm": 0.08064165711402893, "learning_rate": 3.937422547630519e-05, "loss": 0.0015, "step": 1740 }, { "epoch": 5.932203389830509, "grad_norm": 0.05471855029463768, "learning_rate": 3.882711191363859e-05, "loss": 0.0008, "step": 1750 }, { "epoch": 5.966101694915254, "grad_norm": 0.25674968957901, "learning_rate": 3.828140285950676e-05, "loss": 0.0015, "step": 1760 }, { "epoch": 6.0, "grad_norm": 0.03591851517558098, "learning_rate": 3.773716691327432e-05, "loss": 0.0007, "step": 1770 }, { "epoch": 6.033898305084746, "grad_norm": 0.228873610496521, "learning_rate": 3.7194472489126176e-05, "loss": 0.001, "step": 1780 }, { "epoch": 6.067796610169491, "grad_norm": 0.10522802919149399, "learning_rate": 3.665338780746736e-05, "loss": 0.0007, "step": 1790 }, { "epoch": 6.101694915254237, "grad_norm": 0.1092459186911583, "learning_rate": 3.611398088634721e-05, "loss": 0.0004, "step": 1800 }, { "epoch": 6.135593220338983, "grad_norm": 0.15313047170639038, "learning_rate": 3.5576319532909145e-05, "loss": 0.0004, "step": 1810 }, { "epoch": 6.169491525423728, "grad_norm": 0.04173111170530319, "learning_rate": 3.5040471334866695e-05, "loss": 0.0004, "step": 1820 }, { "epoch": 6.203389830508475, "grad_norm": 0.02953781560063362, "learning_rate": 3.450650365200741e-05, "loss": 0.0003, "step": 1830 }, { "epoch": 6.237288135593221, "grad_norm": 0.04410599172115326, "learning_rate": 3.397448360772516e-05, "loss": 0.0006, "step": 1840 }, { "epoch": 6.271186440677966, "grad_norm": 0.09171520918607712, "learning_rate": 3.344447808058232e-05, "loss": 0.0005, "step": 1850 }, { "epoch": 6.305084745762712, "grad_norm": 0.051552463322877884, "learning_rate": 3.291655369590269e-05, "loss": 0.0006, "step": 1860 }, { "epoch": 6.338983050847458, "grad_norm": 0.03626188263297081, "learning_rate": 3.239077681739618e-05, "loss": 0.0005, "step": 1870 }, { "epoch": 6.372881355932203, "grad_norm": 0.09546960145235062, "learning_rate": 3.186721353881648e-05, "loss": 0.0004, "step": 1880 }, { "epoch": 6.406779661016949, "grad_norm": 0.0609428696334362, "learning_rate": 3.134592967565263e-05, "loss": 0.0003, "step": 1890 }, { "epoch": 6.440677966101695, "grad_norm": 0.05546706169843674, "learning_rate": 3.082699075685553e-05, "loss": 0.0003, "step": 1900 }, { "epoch": 6.47457627118644, "grad_norm": 0.03381851688027382, "learning_rate": 3.0310462016600504e-05, "loss": 0.0006, "step": 1910 }, { "epoch": 6.508474576271187, "grad_norm": 0.0640668272972107, "learning_rate": 2.9796408386086962e-05, "loss": 0.0003, "step": 1920 }, { "epoch": 6.5423728813559325, "grad_norm": 0.07631280273199081, "learning_rate": 2.9284894485376057e-05, "loss": 0.0004, "step": 1930 }, { "epoch": 6.576271186440678, "grad_norm": 0.1757967323064804, "learning_rate": 2.8775984615267504e-05, "loss": 0.0007, "step": 1940 }, { "epoch": 6.610169491525424, "grad_norm": 0.05827037990093231, "learning_rate": 2.8269742749216566e-05, "loss": 0.0005, "step": 1950 }, { "epoch": 6.6440677966101696, "grad_norm": 0.07589293271303177, "learning_rate": 2.7766232525292103e-05, "loss": 0.0003, "step": 1960 }, { "epoch": 6.677966101694915, "grad_norm": 0.0697827935218811, "learning_rate": 2.7265517238176847e-05, "loss": 0.0008, "step": 1970 }, { "epoch": 6.711864406779661, "grad_norm": 0.11972369253635406, "learning_rate": 2.676765983121089e-05, "loss": 0.0005, "step": 1980 }, { "epoch": 6.745762711864407, "grad_norm": 0.0417722649872303, "learning_rate": 2.627272288847915e-05, "loss": 0.0003, "step": 1990 }, { "epoch": 6.779661016949152, "grad_norm": 0.04584512487053871, "learning_rate": 2.578076862694426e-05, "loss": 0.0007, "step": 2000 }, { "epoch": 6.813559322033898, "grad_norm": 0.07800233364105225, "learning_rate": 2.5291858888625368e-05, "loss": 0.0004, "step": 2010 }, { "epoch": 6.847457627118644, "grad_norm": 0.07211495190858841, "learning_rate": 2.4806055132824185e-05, "loss": 0.0003, "step": 2020 }, { "epoch": 6.88135593220339, "grad_norm": 0.06364759802818298, "learning_rate": 2.4323418428399203e-05, "loss": 0.0003, "step": 2030 }, { "epoch": 6.915254237288136, "grad_norm": 0.07082319259643555, "learning_rate": 2.384400944608886e-05, "loss": 0.0007, "step": 2040 }, { "epoch": 6.9491525423728815, "grad_norm": 0.06931556761264801, "learning_rate": 2.336788845088478e-05, "loss": 0.0004, "step": 2050 }, { "epoch": 6.983050847457627, "grad_norm": 0.15574094653129578, "learning_rate": 2.289511529445616e-05, "loss": 0.0006, "step": 2060 }, { "epoch": 7.016949152542373, "grad_norm": 0.04960203543305397, "learning_rate": 2.242574940762589e-05, "loss": 0.0003, "step": 2070 }, { "epoch": 7.0508474576271185, "grad_norm": 0.049310192465782166, "learning_rate": 2.195984979289974e-05, "loss": 0.0005, "step": 2080 }, { "epoch": 7.084745762711864, "grad_norm": 0.06411266326904297, "learning_rate": 2.149747501704939e-05, "loss": 0.0004, "step": 2090 }, { "epoch": 7.11864406779661, "grad_norm": 0.06982172280550003, "learning_rate": 2.1038683203750092e-05, "loss": 0.0005, "step": 2100 }, { "epoch": 7.1525423728813555, "grad_norm": 0.11477413028478622, "learning_rate": 2.058353202627417e-05, "loss": 0.0004, "step": 2110 }, { "epoch": 7.186440677966102, "grad_norm": 0.05071744695305824, "learning_rate": 2.0132078700241158e-05, "loss": 0.0003, "step": 2120 }, { "epoch": 7.220338983050848, "grad_norm": 0.06909611821174622, "learning_rate": 1.968437997642526e-05, "loss": 0.0004, "step": 2130 }, { "epoch": 7.254237288135593, "grad_norm": 0.08218582719564438, "learning_rate": 1.924049213362153e-05, "loss": 0.0005, "step": 2140 }, { "epoch": 7.288135593220339, "grad_norm": 0.10386521369218826, "learning_rate": 1.8800470971571184e-05, "loss": 0.0003, "step": 2150 }, { "epoch": 7.322033898305085, "grad_norm": 0.052584659308195114, "learning_rate": 1.836437180394715e-05, "loss": 0.0002, "step": 2160 }, { "epoch": 7.3559322033898304, "grad_norm": 0.0438753217458725, "learning_rate": 1.793224945140086e-05, "loss": 0.0003, "step": 2170 }, { "epoch": 7.389830508474576, "grad_norm": 0.04817179590463638, "learning_rate": 1.750415823467082e-05, "loss": 0.0002, "step": 2180 }, { "epoch": 7.423728813559322, "grad_norm": 0.09476613253355026, "learning_rate": 1.7080151967754142e-05, "loss": 0.0004, "step": 2190 }, { "epoch": 7.4576271186440675, "grad_norm": 0.04253845289349556, "learning_rate": 1.666028395114185e-05, "loss": 0.0004, "step": 2200 }, { "epoch": 7.491525423728813, "grad_norm": 0.055023740977048874, "learning_rate": 1.62446069651185e-05, "loss": 0.0009, "step": 2210 }, { "epoch": 7.52542372881356, "grad_norm": 0.11626702547073364, "learning_rate": 1.5833173263127426e-05, "loss": 0.0007, "step": 2220 }, { "epoch": 7.559322033898305, "grad_norm": 0.1153387799859047, "learning_rate": 1.5426034565202142e-05, "loss": 0.0007, "step": 2230 }, { "epoch": 7.593220338983051, "grad_norm": 0.07169368118047714, "learning_rate": 1.5023242051464675e-05, "loss": 0.0004, "step": 2240 }, { "epoch": 7.627118644067797, "grad_norm": 0.061480406671762466, "learning_rate": 1.4624846355692001e-05, "loss": 0.0004, "step": 2250 }, { "epoch": 7.661016949152542, "grad_norm": 0.0765593945980072, "learning_rate": 1.4230897558950951e-05, "loss": 0.001, "step": 2260 }, { "epoch": 7.694915254237288, "grad_norm": 0.06186467036604881, "learning_rate": 1.384144518330268e-05, "loss": 0.0003, "step": 2270 }, { "epoch": 7.728813559322034, "grad_norm": 0.02641032636165619, "learning_rate": 1.3456538185577466e-05, "loss": 0.0003, "step": 2280 }, { "epoch": 7.762711864406779, "grad_norm": 0.08123563975095749, "learning_rate": 1.3076224951220411e-05, "loss": 0.0009, "step": 2290 }, { "epoch": 7.796610169491525, "grad_norm": 0.09227631241083145, "learning_rate": 1.2700553288209126e-05, "loss": 0.0003, "step": 2300 }, { "epoch": 7.830508474576272, "grad_norm": 0.03893280774354935, "learning_rate": 1.2329570421043957e-05, "loss": 0.0003, "step": 2310 }, { "epoch": 7.864406779661017, "grad_norm": 0.04822898656129837, "learning_rate": 1.1963322984811453e-05, "loss": 0.0008, "step": 2320 }, { "epoch": 7.898305084745763, "grad_norm": 0.05023796483874321, "learning_rate": 1.1601857019322093e-05, "loss": 0.0003, "step": 2330 }, { "epoch": 7.932203389830509, "grad_norm": 0.09390699863433838, "learning_rate": 1.1245217963322763e-05, "loss": 0.0003, "step": 2340 }, { "epoch": 7.966101694915254, "grad_norm": 0.016306394711136818, "learning_rate": 1.0893450648784737e-05, "loss": 0.002, "step": 2350 }, { "epoch": 8.0, "grad_norm": 0.04842328652739525, "learning_rate": 1.0546599295268061e-05, "loss": 0.0003, "step": 2360 }, { "epoch": 8.033898305084746, "grad_norm": 0.034694500267505646, "learning_rate": 1.0204707504362826e-05, "loss": 0.0003, "step": 2370 }, { "epoch": 8.067796610169491, "grad_norm": 0.049791429191827774, "learning_rate": 9.867818254208123e-06, "loss": 0.0007, "step": 2380 }, { "epoch": 8.101694915254237, "grad_norm": 0.02616948075592518, "learning_rate": 9.535973894089401e-06, "loss": 0.0004, "step": 2390 }, { "epoch": 8.135593220338983, "grad_norm": 0.08042062073945999, "learning_rate": 9.209216139114935e-06, "loss": 0.0002, "step": 2400 }, { "epoch": 8.169491525423728, "grad_norm": 0.03833793103694916, "learning_rate": 8.887586064971858e-06, "loss": 0.0004, "step": 2410 }, { "epoch": 8.203389830508474, "grad_norm": 0.04440383240580559, "learning_rate": 8.571124102762768e-06, "loss": 0.0003, "step": 2420 }, { "epoch": 8.23728813559322, "grad_norm": 0.026959413662552834, "learning_rate": 8.259870033923151e-06, "loss": 0.0012, "step": 2430 }, { "epoch": 8.271186440677965, "grad_norm": 0.037755124270915985, "learning_rate": 7.95386298522065e-06, "loss": 0.0004, "step": 2440 }, { "epoch": 8.305084745762711, "grad_norm": 0.1117471233010292, "learning_rate": 7.653141423836584e-06, "loss": 0.0011, "step": 2450 }, { "epoch": 8.338983050847457, "grad_norm": 0.09815044701099396, "learning_rate": 7.357743152530272e-06, "loss": 0.0006, "step": 2460 }, { "epoch": 8.372881355932204, "grad_norm": 0.09394808858633041, "learning_rate": 7.067705304887074e-06, "loss": 0.0005, "step": 2470 }, { "epoch": 8.40677966101695, "grad_norm": 0.06960106641054153, "learning_rate": 6.783064340650364e-06, "loss": 0.0003, "step": 2480 }, { "epoch": 8.440677966101696, "grad_norm": 0.035113442689180374, "learning_rate": 6.503856041138323e-06, "loss": 0.0005, "step": 2490 }, { "epoch": 8.474576271186441, "grad_norm": 0.02795569784939289, "learning_rate": 6.230115504745953e-06, "loss": 0.0012, "step": 2500 }, { "epoch": 8.508474576271187, "grad_norm": 0.02803714945912361, "learning_rate": 5.961877142533056e-06, "loss": 0.0004, "step": 2510 }, { "epoch": 8.542372881355933, "grad_norm": 0.009161763824522495, "learning_rate": 5.699174673898394e-06, "loss": 0.0003, "step": 2520 }, { "epoch": 8.576271186440678, "grad_norm": 0.026831788942217827, "learning_rate": 5.4420411223410574e-06, "loss": 0.0004, "step": 2530 }, { "epoch": 8.610169491525424, "grad_norm": 0.011004606261849403, "learning_rate": 5.190508811309091e-06, "loss": 0.0002, "step": 2540 }, { "epoch": 8.64406779661017, "grad_norm": 0.027460021898150444, "learning_rate": 4.944609360136271e-06, "loss": 0.0002, "step": 2550 }, { "epoch": 8.677966101694915, "grad_norm": 0.04089871421456337, "learning_rate": 4.704373680067325e-06, "loss": 0.0002, "step": 2560 }, { "epoch": 8.711864406779661, "grad_norm": 0.03699534758925438, "learning_rate": 4.469831970372146e-06, "loss": 0.0002, "step": 2570 }, { "epoch": 8.745762711864407, "grad_norm": 0.021349631249904633, "learning_rate": 4.241013714549597e-06, "loss": 0.0002, "step": 2580 }, { "epoch": 8.779661016949152, "grad_norm": 0.02745555341243744, "learning_rate": 4.017947676621187e-06, "loss": 0.0004, "step": 2590 }, { "epoch": 8.813559322033898, "grad_norm": 0.013199915178120136, "learning_rate": 3.800661897515245e-06, "loss": 0.0002, "step": 2600 }, { "epoch": 8.847457627118644, "grad_norm": 0.026064835488796234, "learning_rate": 3.589183691541981e-06, "loss": 0.0002, "step": 2610 }, { "epoch": 8.88135593220339, "grad_norm": 0.01884126476943493, "learning_rate": 3.3835396429599152e-06, "loss": 0.0003, "step": 2620 }, { "epoch": 8.915254237288135, "grad_norm": 0.03649914637207985, "learning_rate": 3.18375560263402e-06, "loss": 0.0002, "step": 2630 }, { "epoch": 8.94915254237288, "grad_norm": 0.020875606685876846, "learning_rate": 2.9898566847861245e-06, "loss": 0.0003, "step": 2640 }, { "epoch": 8.983050847457626, "grad_norm": 0.05111026391386986, "learning_rate": 2.8018672638378483e-06, "loss": 0.0002, "step": 2650 }, { "epoch": 9.016949152542374, "grad_norm": 0.017365949228405952, "learning_rate": 2.619810971346587e-06, "loss": 0.0003, "step": 2660 }, { "epoch": 9.05084745762712, "grad_norm": 0.09889742732048035, "learning_rate": 2.443710693034884e-06, "loss": 0.0003, "step": 2670 }, { "epoch": 9.084745762711865, "grad_norm": 0.029447954148054123, "learning_rate": 2.2735885659134925e-06, "loss": 0.0003, "step": 2680 }, { "epoch": 9.11864406779661, "grad_norm": 0.0224969033151865, "learning_rate": 2.1094659754986477e-06, "loss": 0.0002, "step": 2690 }, { "epoch": 9.152542372881356, "grad_norm": 0.010688988491892815, "learning_rate": 1.9513635531237417e-06, "loss": 0.0003, "step": 2700 }, { "epoch": 9.186440677966102, "grad_norm": 0.059495627880096436, "learning_rate": 1.7993011733458075e-06, "loss": 0.0003, "step": 2710 }, { "epoch": 9.220338983050848, "grad_norm": 0.007177701219916344, "learning_rate": 1.6532979514471747e-06, "loss": 0.0002, "step": 2720 }, { "epoch": 9.254237288135593, "grad_norm": 0.052893899381160736, "learning_rate": 1.5133722410325457e-06, "loss": 0.0008, "step": 2730 }, { "epoch": 9.288135593220339, "grad_norm": 0.012546384707093239, "learning_rate": 1.3795416317218036e-06, "loss": 0.0007, "step": 2740 }, { "epoch": 9.322033898305085, "grad_norm": 0.008967459201812744, "learning_rate": 1.2518229469388965e-06, "loss": 0.0002, "step": 2750 }, { "epoch": 9.35593220338983, "grad_norm": 0.11077430099248886, "learning_rate": 1.1302322417970135e-06, "loss": 0.0002, "step": 2760 }, { "epoch": 9.389830508474576, "grad_norm": 0.014416457153856754, "learning_rate": 1.0147848010803318e-06, "loss": 0.0001, "step": 2770 }, { "epoch": 9.423728813559322, "grad_norm": 0.016251103952527046, "learning_rate": 9.054951373226484e-07, "loss": 0.0002, "step": 2780 }, { "epoch": 9.457627118644067, "grad_norm": 0.021497737616300583, "learning_rate": 8.023769889830002e-07, "loss": 0.0011, "step": 2790 }, { "epoch": 9.491525423728813, "grad_norm": 0.031376853585243225, "learning_rate": 7.054433187187071e-07, "loss": 0.0003, "step": 2800 }, { "epoch": 9.525423728813559, "grad_norm": 0.017223788425326347, "learning_rate": 6.147063117558471e-07, "loss": 0.0002, "step": 2810 }, { "epoch": 9.559322033898304, "grad_norm": 0.008591080084443092, "learning_rate": 5.301773743574712e-07, "loss": 0.0002, "step": 2820 }, { "epoch": 9.59322033898305, "grad_norm": 0.016840104013681412, "learning_rate": 4.518671323897938e-07, "loss": 0.0004, "step": 2830 }, { "epoch": 9.627118644067796, "grad_norm": 0.01281412597745657, "learning_rate": 3.7978542998643454e-07, "loss": 0.0001, "step": 2840 }, { "epoch": 9.661016949152543, "grad_norm": 0.01014797855168581, "learning_rate": 3.1394132831094047e-07, "loss": 0.0001, "step": 2850 }, { "epoch": 9.694915254237289, "grad_norm": 0.006094843149185181, "learning_rate": 2.5434310441773137e-07, "loss": 0.0002, "step": 2860 }, { "epoch": 9.728813559322035, "grad_norm": 0.054053328931331635, "learning_rate": 2.0099825021163144e-07, "loss": 0.0002, "step": 2870 }, { "epoch": 9.76271186440678, "grad_norm": 0.07471420615911484, "learning_rate": 1.5391347150607304e-07, "loss": 0.0005, "step": 2880 }, { "epoch": 9.796610169491526, "grad_norm": 0.1749231517314911, "learning_rate": 1.1309468718013194e-07, "loss": 0.0005, "step": 2890 }, { "epoch": 9.830508474576272, "grad_norm": 0.011661729775369167, "learning_rate": 7.854702843449469e-08, "loss": 0.0004, "step": 2900 }, { "epoch": 9.864406779661017, "grad_norm": 0.12255346029996872, "learning_rate": 5.0274838146413536e-08, "loss": 0.0009, "step": 2910 }, { "epoch": 9.898305084745763, "grad_norm": 0.008154436945915222, "learning_rate": 2.828167032379869e-08, "loss": 0.0002, "step": 2920 }, { "epoch": 9.932203389830509, "grad_norm": 0.017673006281256676, "learning_rate": 1.257028965842566e-08, "loss": 0.0001, "step": 2930 }, { "epoch": 9.966101694915254, "grad_norm": 0.009231282398104668, "learning_rate": 3.1426711784299233e-09, "loss": 0.0004, "step": 2940 }, { "epoch": 10.0, "grad_norm": 0.3303571045398712, "learning_rate": 0.0, "loss": 0.0006, "step": 2950 }, { "epoch": 10.0, "step": 2950, "total_flos": 3.087953656660627e+17, "train_loss": 0.0031311477128005884, "train_runtime": 2880.9611, "train_samples_per_second": 50.046, "train_steps_per_second": 1.024 } ], "logging_steps": 10, "max_steps": 2950, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.087953656660627e+17, "train_batch_size": 49, "trial_name": null, "trial_params": null }