{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.2, "eval_steps": 500, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.8061772584915161, "learning_rate": 1.8e-06, "loss": 1.2102, "step": 10 }, { "grad_norm": 1.4747154712677002, "learning_rate": 3.8e-06, "loss": 1.1989, "step": 20 }, { "grad_norm": 0.6578592658042908, "learning_rate": 5.8e-06, "loss": 1.0852, "step": 30 }, { "grad_norm": 0.8031267523765564, "learning_rate": 7.8e-06, "loss": 1.061, "step": 40 }, { "grad_norm": 0.7423774600028992, "learning_rate": 9.800000000000001e-06, "loss": 1.0392, "step": 50 }, { "grad_norm": 0.3366369605064392, "learning_rate": 1.18e-05, "loss": 1.0494, "step": 60 }, { "grad_norm": 0.6181372404098511, "learning_rate": 1.3800000000000002e-05, "loss": 1.0226, "step": 70 }, { "grad_norm": 1.1166492700576782, "learning_rate": 1.58e-05, "loss": 1.021, "step": 80 }, { "grad_norm": 3.875643253326416, "learning_rate": 1.78e-05, "loss": 0.9186, "step": 90 }, { "grad_norm": 3.100308895111084, "learning_rate": 1.9800000000000004e-05, "loss": 0.7894, "step": 100 }, { "grad_norm": 3.7580819129943848, "learning_rate": 2.18e-05, "loss": 0.6621, "step": 110 }, { "grad_norm": 4.687175750732422, "learning_rate": 2.38e-05, "loss": 0.5106, "step": 120 }, { "grad_norm": 4.228991985321045, "learning_rate": 2.58e-05, "loss": 0.3378, "step": 130 }, { "grad_norm": 2.555137872695923, "learning_rate": 2.7800000000000005e-05, "loss": 0.2668, "step": 140 }, { "grad_norm": 3.8804194927215576, "learning_rate": 2.98e-05, "loss": 0.2177, "step": 150 }, { "grad_norm": 3.5343244075775146, "learning_rate": 3.18e-05, "loss": 0.1956, "step": 160 }, { "grad_norm": 2.94913649559021, "learning_rate": 3.38e-05, "loss": 0.1707, "step": 170 }, { "grad_norm": 3.1313822269439697, "learning_rate": 3.58e-05, "loss": 0.1722, "step": 180 }, { "grad_norm": 1.553078293800354, "learning_rate": 3.7800000000000004e-05, "loss": 0.1514, "step": 190 }, { "grad_norm": 2.466740131378174, "learning_rate": 3.9800000000000005e-05, "loss": 0.144, "step": 200 }, { "grad_norm": 1.6171531677246094, "learning_rate": 4.18e-05, "loss": 0.1403, "step": 210 }, { "grad_norm": 3.0060226917266846, "learning_rate": 4.38e-05, "loss": 0.1418, "step": 220 }, { "grad_norm": 1.5442951917648315, "learning_rate": 4.58e-05, "loss": 0.1284, "step": 230 }, { "grad_norm": 2.284505605697632, "learning_rate": 4.78e-05, "loss": 0.132, "step": 240 }, { "grad_norm": 1.6954904794692993, "learning_rate": 4.9800000000000004e-05, "loss": 0.1254, "step": 250 }, { "grad_norm": 2.5462610721588135, "learning_rate": 5.1800000000000005e-05, "loss": 0.1209, "step": 260 }, { "grad_norm": 2.6769697666168213, "learning_rate": 5.380000000000001e-05, "loss": 0.133, "step": 270 }, { "grad_norm": 1.9513802528381348, "learning_rate": 5.580000000000001e-05, "loss": 0.1226, "step": 280 }, { "grad_norm": 1.6414059400558472, "learning_rate": 5.7799999999999995e-05, "loss": 0.1515, "step": 290 }, { "grad_norm": 2.1393585205078125, "learning_rate": 5.9800000000000003e-05, "loss": 0.1242, "step": 300 }, { "grad_norm": 0.8399069309234619, "learning_rate": 6.18e-05, "loss": 0.1385, "step": 310 }, { "grad_norm": 2.0428855419158936, "learning_rate": 6.38e-05, "loss": 0.1246, "step": 320 }, { "grad_norm": 2.1637940406799316, "learning_rate": 6.58e-05, "loss": 0.1297, "step": 330 }, { "grad_norm": 1.5418416261672974, "learning_rate": 6.780000000000001e-05, "loss": 0.1129, "step": 340 }, { "grad_norm": 1.2646944522857666, "learning_rate": 6.98e-05, "loss": 0.1005, "step": 350 }, { "grad_norm": 1.169583797454834, "learning_rate": 7.18e-05, "loss": 0.1103, "step": 360 }, { "grad_norm": 1.2223219871520996, "learning_rate": 7.38e-05, "loss": 0.1167, "step": 370 }, { "grad_norm": 2.094849109649658, "learning_rate": 7.58e-05, "loss": 0.1131, "step": 380 }, { "grad_norm": 1.0914831161499023, "learning_rate": 7.780000000000001e-05, "loss": 0.1183, "step": 390 }, { "grad_norm": 1.2861425876617432, "learning_rate": 7.98e-05, "loss": 0.1113, "step": 400 }, { "grad_norm": 0.8093202710151672, "learning_rate": 8.18e-05, "loss": 0.117, "step": 410 }, { "grad_norm": 1.6367533206939697, "learning_rate": 8.38e-05, "loss": 0.1055, "step": 420 }, { "grad_norm": 1.0690555572509766, "learning_rate": 8.58e-05, "loss": 0.1075, "step": 430 }, { "grad_norm": 1.3714529275894165, "learning_rate": 8.78e-05, "loss": 0.104, "step": 440 }, { "grad_norm": 1.5832537412643433, "learning_rate": 8.98e-05, "loss": 0.1128, "step": 450 }, { "grad_norm": 0.7846147418022156, "learning_rate": 9.180000000000001e-05, "loss": 0.1119, "step": 460 }, { "grad_norm": 1.241227388381958, "learning_rate": 9.38e-05, "loss": 0.1172, "step": 470 }, { "grad_norm": 0.9083560705184937, "learning_rate": 9.58e-05, "loss": 0.1111, "step": 480 }, { "grad_norm": 0.7720792889595032, "learning_rate": 9.78e-05, "loss": 0.0982, "step": 490 }, { "grad_norm": 1.0341143608093262, "learning_rate": 9.98e-05, "loss": 0.0984, "step": 500 }, { "grad_norm": 1.1317723989486694, "learning_rate": 9.9999778549206e-05, "loss": 0.1106, "step": 510 }, { "grad_norm": 0.9916303753852844, "learning_rate": 9.999901304280685e-05, "loss": 0.0987, "step": 520 }, { "grad_norm": 0.898222029209137, "learning_rate": 9.999770075521164e-05, "loss": 0.1017, "step": 530 }, { "grad_norm": 1.3211407661437988, "learning_rate": 9.99958417007713e-05, "loss": 0.0906, "step": 540 }, { "grad_norm": 1.2681704759597778, "learning_rate": 9.999343589981615e-05, "loss": 0.1007, "step": 550 }, { "grad_norm": 1.1410036087036133, "learning_rate": 9.999048337865568e-05, "loss": 0.0908, "step": 560 }, { "grad_norm": 0.619777262210846, "learning_rate": 9.998698416957815e-05, "loss": 0.0889, "step": 570 }, { "grad_norm": 0.9319987893104553, "learning_rate": 9.998293831085037e-05, "loss": 0.0947, "step": 580 }, { "grad_norm": 0.7254066467285156, "learning_rate": 9.997834584671719e-05, "loss": 0.1014, "step": 590 }, { "grad_norm": 0.5764051675796509, "learning_rate": 9.997320682740107e-05, "loss": 0.0991, "step": 600 }, { "grad_norm": 0.6951457262039185, "learning_rate": 9.996752130910149e-05, "loss": 0.1116, "step": 610 }, { "grad_norm": 0.7899103760719299, "learning_rate": 9.99612893539944e-05, "loss": 0.0923, "step": 620 }, { "grad_norm": 0.5182093977928162, "learning_rate": 9.995451103023144e-05, "loss": 0.075, "step": 630 }, { "grad_norm": 0.5825569033622742, "learning_rate": 9.994718641193928e-05, "loss": 0.0834, "step": 640 }, { "grad_norm": 0.5975721478462219, "learning_rate": 9.993931557921874e-05, "loss": 0.0808, "step": 650 }, { "grad_norm": 0.6993137001991272, "learning_rate": 9.993089861814402e-05, "loss": 0.0857, "step": 660 }, { "grad_norm": 0.7876319885253906, "learning_rate": 9.992193562076166e-05, "loss": 0.1006, "step": 670 }, { "grad_norm": 0.668891429901123, "learning_rate": 9.991242668508954e-05, "loss": 0.0899, "step": 680 }, { "grad_norm": 0.5712555646896362, "learning_rate": 9.990237191511587e-05, "loss": 0.0839, "step": 690 }, { "grad_norm": 0.43553826212882996, "learning_rate": 9.989177142079802e-05, "loss": 0.0856, "step": 700 }, { "grad_norm": 0.7408167123794556, "learning_rate": 9.988062531806126e-05, "loss": 0.0957, "step": 710 }, { "grad_norm": 0.6338667273521423, "learning_rate": 9.986893372879762e-05, "loss": 0.0849, "step": 720 }, { "grad_norm": 0.708287239074707, "learning_rate": 9.985669678086443e-05, "loss": 0.0938, "step": 730 }, { "grad_norm": 0.7147543430328369, "learning_rate": 9.984391460808298e-05, "loss": 0.0731, "step": 740 }, { "grad_norm": 0.5941300392150879, "learning_rate": 9.983058735023709e-05, "loss": 0.0834, "step": 750 }, { "grad_norm": 0.5001762509346008, "learning_rate": 9.98167151530715e-05, "loss": 0.0704, "step": 760 }, { "grad_norm": 0.4520997703075409, "learning_rate": 9.980229816829034e-05, "loss": 0.0725, "step": 770 }, { "grad_norm": 0.3239596486091614, "learning_rate": 9.978733655355544e-05, "loss": 0.0759, "step": 780 }, { "grad_norm": 0.6231990456581116, "learning_rate": 9.977183047248464e-05, "loss": 0.0692, "step": 790 }, { "grad_norm": 0.5697351098060608, "learning_rate": 9.975578009464992e-05, "loss": 0.0815, "step": 800 }, { "grad_norm": 0.5554783344268799, "learning_rate": 9.97391855955757e-05, "loss": 0.0846, "step": 810 }, { "grad_norm": 0.8224487900733948, "learning_rate": 9.972204715673669e-05, "loss": 0.0775, "step": 820 }, { "grad_norm": 0.6740953326225281, "learning_rate": 9.970436496555617e-05, "loss": 0.0798, "step": 830 }, { "grad_norm": 0.6196176409721375, "learning_rate": 9.968613921540373e-05, "loss": 0.0656, "step": 840 }, { "grad_norm": 0.7281718850135803, "learning_rate": 9.966737010559326e-05, "loss": 0.0818, "step": 850 }, { "grad_norm": 0.9248817563056946, "learning_rate": 9.964805784138072e-05, "loss": 0.0881, "step": 860 }, { "grad_norm": 0.8059525489807129, "learning_rate": 9.962820263396195e-05, "loss": 0.0822, "step": 870 }, { "grad_norm": 0.8395443558692932, "learning_rate": 9.960780470047033e-05, "loss": 0.073, "step": 880 }, { "grad_norm": 0.7793157696723938, "learning_rate": 9.958686426397437e-05, "loss": 0.0694, "step": 890 }, { "grad_norm": 0.447717547416687, "learning_rate": 9.956538155347534e-05, "loss": 0.0794, "step": 900 }, { "grad_norm": 0.3848801553249359, "learning_rate": 9.95433568039047e-05, "loss": 0.0629, "step": 910 }, { "grad_norm": 0.5611248016357422, "learning_rate": 9.952079025612162e-05, "loss": 0.0769, "step": 920 }, { "grad_norm": 0.49911221861839294, "learning_rate": 9.949768215691022e-05, "loss": 0.0634, "step": 930 }, { "grad_norm": 0.42823612689971924, "learning_rate": 9.9474032758977e-05, "loss": 0.0582, "step": 940 }, { "grad_norm": 0.856852650642395, "learning_rate": 9.944984232094794e-05, "loss": 0.0779, "step": 950 }, { "grad_norm": 0.6261837482452393, "learning_rate": 9.942511110736584e-05, "loss": 0.0618, "step": 960 }, { "grad_norm": 0.5489150881767273, "learning_rate": 9.939983938868726e-05, "loss": 0.0646, "step": 970 }, { "grad_norm": 0.6689404249191284, "learning_rate": 9.93740274412797e-05, "loss": 0.0764, "step": 980 }, { "grad_norm": 0.5648400187492371, "learning_rate": 9.934767554741846e-05, "loss": 0.079, "step": 990 }, { "grad_norm": 0.48364317417144775, "learning_rate": 9.932078399528361e-05, "loss": 0.0703, "step": 1000 }, { "grad_norm": 0.6004571914672852, "learning_rate": 9.929335307895689e-05, "loss": 0.0713, "step": 1010 }, { "grad_norm": 0.5344667434692383, "learning_rate": 9.926538309841839e-05, "loss": 0.0677, "step": 1020 }, { "grad_norm": 0.6448832750320435, "learning_rate": 9.923687435954334e-05, "loss": 0.0681, "step": 1030 }, { "grad_norm": 0.4730415940284729, "learning_rate": 9.920782717409873e-05, "loss": 0.0694, "step": 1040 }, { "grad_norm": 0.28815484046936035, "learning_rate": 9.917824185973994e-05, "loss": 0.0548, "step": 1050 }, { "grad_norm": 0.6332595348358154, "learning_rate": 9.914811874000723e-05, "loss": 0.0776, "step": 1060 }, { "grad_norm": 0.7720545530319214, "learning_rate": 9.911745814432218e-05, "loss": 0.0738, "step": 1070 }, { "grad_norm": 0.4502904713153839, "learning_rate": 9.90862604079842e-05, "loss": 0.0738, "step": 1080 }, { "grad_norm": 0.5622782707214355, "learning_rate": 9.90545258721667e-05, "loss": 0.0507, "step": 1090 }, { "grad_norm": 0.7295961380004883, "learning_rate": 9.90222548839135e-05, "loss": 0.0684, "step": 1100 }, { "grad_norm": 0.40541696548461914, "learning_rate": 9.898944779613495e-05, "loss": 0.0661, "step": 1110 }, { "grad_norm": 0.6888296604156494, "learning_rate": 9.89561049676041e-05, "loss": 0.0671, "step": 1120 }, { "grad_norm": 0.4445127844810486, "learning_rate": 9.89222267629528e-05, "loss": 0.0771, "step": 1130 }, { "grad_norm": 0.6876557469367981, "learning_rate": 9.888781355266763e-05, "loss": 0.0749, "step": 1140 }, { "grad_norm": 0.6139089465141296, "learning_rate": 9.885286571308598e-05, "loss": 0.0721, "step": 1150 }, { "grad_norm": 0.4543188810348511, "learning_rate": 9.881738362639182e-05, "loss": 0.0733, "step": 1160 }, { "grad_norm": 0.7434316873550415, "learning_rate": 9.878136768061154e-05, "loss": 0.0743, "step": 1170 }, { "grad_norm": 0.39948761463165283, "learning_rate": 9.874481826960979e-05, "loss": 0.0712, "step": 1180 }, { "grad_norm": 0.5060223937034607, "learning_rate": 9.870773579308503e-05, "loss": 0.0742, "step": 1190 }, { "grad_norm": 0.5455248951911926, "learning_rate": 9.867012065656533e-05, "loss": 0.0759, "step": 1200 }, { "grad_norm": 0.573907732963562, "learning_rate": 9.863197327140376e-05, "loss": 0.0726, "step": 1210 }, { "grad_norm": 0.5230447053909302, "learning_rate": 9.859329405477403e-05, "loss": 0.0569, "step": 1220 }, { "grad_norm": 0.61860191822052, "learning_rate": 9.855408342966585e-05, "loss": 0.0701, "step": 1230 }, { "grad_norm": 0.811663806438446, "learning_rate": 9.851434182488033e-05, "loss": 0.0636, "step": 1240 }, { "grad_norm": 0.6513891220092773, "learning_rate": 9.84740696750253e-05, "loss": 0.0789, "step": 1250 }, { "grad_norm": 0.3722531199455261, "learning_rate": 9.843326742051055e-05, "loss": 0.0634, "step": 1260 }, { "grad_norm": 0.4310877323150635, "learning_rate": 9.839193550754297e-05, "loss": 0.0775, "step": 1270 }, { "grad_norm": 0.5046648979187012, "learning_rate": 9.835007438812177e-05, "loss": 0.0753, "step": 1280 }, { "grad_norm": 0.4245837926864624, "learning_rate": 9.830768452003341e-05, "loss": 0.0717, "step": 1290 }, { "grad_norm": 0.5359390377998352, "learning_rate": 9.826476636684671e-05, "loss": 0.0687, "step": 1300 }, { "grad_norm": 0.8046008944511414, "learning_rate": 9.822132039790773e-05, "loss": 0.0687, "step": 1310 }, { "grad_norm": 0.7594823837280273, "learning_rate": 9.817734708833461e-05, "loss": 0.0752, "step": 1320 }, { "grad_norm": 0.7191202044487, "learning_rate": 9.813284691901243e-05, "loss": 0.0634, "step": 1330 }, { "grad_norm": 0.6475830674171448, "learning_rate": 9.808782037658792e-05, "loss": 0.0696, "step": 1340 }, { "grad_norm": 0.3510691225528717, "learning_rate": 9.804226795346411e-05, "loss": 0.0621, "step": 1350 }, { "grad_norm": 0.5146788954734802, "learning_rate": 9.799619014779503e-05, "loss": 0.0654, "step": 1360 }, { "grad_norm": 0.4277644157409668, "learning_rate": 9.794958746348013e-05, "loss": 0.0527, "step": 1370 }, { "grad_norm": 0.2788071930408478, "learning_rate": 9.790246041015896e-05, "loss": 0.0628, "step": 1380 }, { "grad_norm": 0.35217198729515076, "learning_rate": 9.785480950320538e-05, "loss": 0.0576, "step": 1390 }, { "grad_norm": 0.4901021420955658, "learning_rate": 9.78066352637221e-05, "loss": 0.0684, "step": 1400 }, { "grad_norm": 0.6155123114585876, "learning_rate": 9.775793821853488e-05, "loss": 0.0536, "step": 1410 }, { "grad_norm": 0.32494059205055237, "learning_rate": 9.77087189001868e-05, "loss": 0.0631, "step": 1420 }, { "grad_norm": 0.8440774083137512, "learning_rate": 9.765897784693243e-05, "loss": 0.0645, "step": 1430 }, { "grad_norm": 0.5711706876754761, "learning_rate": 9.760871560273197e-05, "loss": 0.0716, "step": 1440 }, { "grad_norm": 0.6301909685134888, "learning_rate": 9.755793271724526e-05, "loss": 0.0751, "step": 1450 }, { "grad_norm": 0.620735764503479, "learning_rate": 9.750662974582584e-05, "loss": 0.0648, "step": 1460 }, { "grad_norm": 0.43358784914016724, "learning_rate": 9.745480724951473e-05, "loss": 0.0686, "step": 1470 }, { "grad_norm": 0.5163449048995972, "learning_rate": 9.740246579503447e-05, "loss": 0.0749, "step": 1480 }, { "grad_norm": 0.4901696443557739, "learning_rate": 9.734960595478284e-05, "loss": 0.0595, "step": 1490 }, { "grad_norm": 0.7656370997428894, "learning_rate": 9.729622830682657e-05, "loss": 0.0792, "step": 1500 }, { "grad_norm": 0.4338843822479248, "learning_rate": 9.724233343489504e-05, "loss": 0.0667, "step": 1510 }, { "grad_norm": 0.32795488834381104, "learning_rate": 9.718792192837396e-05, "loss": 0.0656, "step": 1520 }, { "grad_norm": 0.4263949394226074, "learning_rate": 9.713299438229886e-05, "loss": 0.0682, "step": 1530 }, { "grad_norm": 0.3967912793159485, "learning_rate": 9.707755139734855e-05, "loss": 0.0752, "step": 1540 }, { "grad_norm": 0.3811866044998169, "learning_rate": 9.702159357983866e-05, "loss": 0.0652, "step": 1550 }, { "grad_norm": 0.5922670364379883, "learning_rate": 9.696512154171492e-05, "loss": 0.0626, "step": 1560 }, { "grad_norm": 0.3355621099472046, "learning_rate": 9.690813590054645e-05, "loss": 0.0556, "step": 1570 }, { "grad_norm": 0.3197598159313202, "learning_rate": 9.685063727951914e-05, "loss": 0.054, "step": 1580 }, { "grad_norm": 0.4919295608997345, "learning_rate": 9.679262630742865e-05, "loss": 0.0654, "step": 1590 }, { "grad_norm": 0.393349826335907, "learning_rate": 9.673410361867373e-05, "loss": 0.0672, "step": 1600 }, { "grad_norm": 0.38691771030426025, "learning_rate": 9.667506985324909e-05, "loss": 0.0537, "step": 1610 }, { "grad_norm": 0.6783390641212463, "learning_rate": 9.661552565673855e-05, "loss": 0.0556, "step": 1620 }, { "grad_norm": 0.4459956884384155, "learning_rate": 9.655547168030789e-05, "loss": 0.0677, "step": 1630 }, { "grad_norm": 0.6396137475967407, "learning_rate": 9.649490858069777e-05, "loss": 0.0683, "step": 1640 }, { "grad_norm": 0.526408314704895, "learning_rate": 9.643383702021658e-05, "loss": 0.0717, "step": 1650 }, { "grad_norm": 0.442583829164505, "learning_rate": 9.637225766673307e-05, "loss": 0.0507, "step": 1660 }, { "grad_norm": 0.4335951805114746, "learning_rate": 9.631017119366922e-05, "loss": 0.0667, "step": 1670 }, { "grad_norm": 0.39904746413230896, "learning_rate": 9.624757827999273e-05, "loss": 0.0662, "step": 1680 }, { "grad_norm": 0.38589608669281006, "learning_rate": 9.618447961020971e-05, "loss": 0.0622, "step": 1690 }, { "grad_norm": 0.4383254945278168, "learning_rate": 9.612087587435707e-05, "loss": 0.0617, "step": 1700 }, { "grad_norm": 0.4938008189201355, "learning_rate": 9.605676776799508e-05, "loss": 0.0594, "step": 1710 }, { "grad_norm": 0.5450459718704224, "learning_rate": 9.599215599219973e-05, "loss": 0.0674, "step": 1720 }, { "grad_norm": 0.5308138728141785, "learning_rate": 9.592704125355505e-05, "loss": 0.0672, "step": 1730 }, { "grad_norm": 0.43554773926734924, "learning_rate": 9.586142426414538e-05, "loss": 0.0586, "step": 1740 }, { "grad_norm": 0.4737524390220642, "learning_rate": 9.57953057415476e-05, "loss": 0.0624, "step": 1750 }, { "grad_norm": 0.6427010297775269, "learning_rate": 9.572868640882328e-05, "loss": 0.0664, "step": 1760 }, { "grad_norm": 0.46082884073257446, "learning_rate": 9.56615669945108e-05, "loss": 0.0617, "step": 1770 }, { "grad_norm": 0.558668315410614, "learning_rate": 9.55939482326173e-05, "loss": 0.0579, "step": 1780 }, { "grad_norm": 0.47442755103111267, "learning_rate": 9.552583086261069e-05, "loss": 0.0694, "step": 1790 }, { "grad_norm": 0.44084689021110535, "learning_rate": 9.545721562941168e-05, "loss": 0.0659, "step": 1800 }, { "grad_norm": 0.47534114122390747, "learning_rate": 9.538810328338543e-05, "loss": 0.064, "step": 1810 }, { "grad_norm": 0.5826574563980103, "learning_rate": 9.531849458033349e-05, "loss": 0.0567, "step": 1820 }, { "grad_norm": 0.49254634976387024, "learning_rate": 9.524839028148547e-05, "loss": 0.0613, "step": 1830 }, { "grad_norm": 0.48347222805023193, "learning_rate": 9.517779115349077e-05, "loss": 0.0597, "step": 1840 }, { "grad_norm": 0.7780311703681946, "learning_rate": 9.510669796841014e-05, "loss": 0.0681, "step": 1850 }, { "grad_norm": 0.48774492740631104, "learning_rate": 9.503511150370727e-05, "loss": 0.073, "step": 1860 }, { "grad_norm": 0.3691915273666382, "learning_rate": 9.496303254224024e-05, "loss": 0.0632, "step": 1870 }, { "grad_norm": 0.3624691069126129, "learning_rate": 9.489046187225306e-05, "loss": 0.0563, "step": 1880 }, { "grad_norm": 0.5045533180236816, "learning_rate": 9.481740028736692e-05, "loss": 0.0569, "step": 1890 }, { "grad_norm": 0.5943124294281006, "learning_rate": 9.474384858657164e-05, "loss": 0.0799, "step": 1900 }, { "grad_norm": 0.8512851595878601, "learning_rate": 9.466980757421679e-05, "loss": 0.0638, "step": 1910 }, { "grad_norm": 0.5189014077186584, "learning_rate": 9.459527806000305e-05, "loss": 0.0611, "step": 1920 }, { "grad_norm": 0.43287965655326843, "learning_rate": 9.452026085897325e-05, "loss": 0.0661, "step": 1930 }, { "grad_norm": 0.5745028853416443, "learning_rate": 9.444475679150348e-05, "loss": 0.0688, "step": 1940 }, { "grad_norm": 0.42947813868522644, "learning_rate": 9.436876668329411e-05, "loss": 0.0672, "step": 1950 }, { "grad_norm": 0.5520886182785034, "learning_rate": 9.429229136536079e-05, "loss": 0.0515, "step": 1960 }, { "grad_norm": 0.5133675932884216, "learning_rate": 9.421533167402534e-05, "loss": 0.0669, "step": 1970 }, { "grad_norm": 0.37619224190711975, "learning_rate": 9.413788845090666e-05, "loss": 0.0765, "step": 1980 }, { "grad_norm": 0.4794745147228241, "learning_rate": 9.405996254291136e-05, "loss": 0.0641, "step": 1990 }, { "grad_norm": 0.48584285378456116, "learning_rate": 9.398155480222474e-05, "loss": 0.0605, "step": 2000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }