{ "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": 0.5420440435409546, "learning_rate": 1.8e-06, "loss": 0.0694, "step": 10 }, { "grad_norm": 0.7189459204673767, "learning_rate": 3.8e-06, "loss": 0.0645, "step": 20 }, { "grad_norm": 0.21701538562774658, "learning_rate": 5.8e-06, "loss": 0.0317, "step": 30 }, { "grad_norm": 0.29029083251953125, "learning_rate": 7.8e-06, "loss": 0.0315, "step": 40 }, { "grad_norm": 0.10109149664640427, "learning_rate": 9.800000000000001e-06, "loss": 0.0273, "step": 50 }, { "grad_norm": 0.19640426337718964, "learning_rate": 1.18e-05, "loss": 0.0319, "step": 60 }, { "grad_norm": 0.1066979169845581, "learning_rate": 1.3800000000000002e-05, "loss": 0.0251, "step": 70 }, { "grad_norm": 0.12810614705085754, "learning_rate": 1.58e-05, "loss": 0.024, "step": 80 }, { "grad_norm": 0.08170574903488159, "learning_rate": 1.78e-05, "loss": 0.0232, "step": 90 }, { "grad_norm": 0.16492632031440735, "learning_rate": 1.9800000000000004e-05, "loss": 0.0283, "step": 100 }, { "grad_norm": 0.14535479247570038, "learning_rate": 2.18e-05, "loss": 0.0231, "step": 110 }, { "grad_norm": 0.17147088050842285, "learning_rate": 2.38e-05, "loss": 0.0315, "step": 120 }, { "grad_norm": 0.10768885165452957, "learning_rate": 2.58e-05, "loss": 0.0194, "step": 130 }, { "grad_norm": 0.10331500321626663, "learning_rate": 2.7800000000000005e-05, "loss": 0.0203, "step": 140 }, { "grad_norm": 0.15777619183063507, "learning_rate": 2.98e-05, "loss": 0.0285, "step": 150 }, { "grad_norm": 0.11859080195426941, "learning_rate": 3.18e-05, "loss": 0.0234, "step": 160 }, { "grad_norm": 0.20379579067230225, "learning_rate": 3.38e-05, "loss": 0.0246, "step": 170 }, { "grad_norm": 0.08920297771692276, "learning_rate": 3.58e-05, "loss": 0.0209, "step": 180 }, { "grad_norm": 0.11533823609352112, "learning_rate": 3.7800000000000004e-05, "loss": 0.0261, "step": 190 }, { "grad_norm": 0.1448919177055359, "learning_rate": 3.9800000000000005e-05, "loss": 0.0264, "step": 200 }, { "grad_norm": 0.1489187777042389, "learning_rate": 4.18e-05, "loss": 0.0249, "step": 210 }, { "grad_norm": 0.1333455890417099, "learning_rate": 4.38e-05, "loss": 0.0281, "step": 220 }, { "grad_norm": 0.09986919909715652, "learning_rate": 4.58e-05, "loss": 0.0241, "step": 230 }, { "grad_norm": 0.15185418725013733, "learning_rate": 4.78e-05, "loss": 0.0279, "step": 240 }, { "grad_norm": 0.16237834095954895, "learning_rate": 4.9800000000000004e-05, "loss": 0.0294, "step": 250 }, { "grad_norm": 0.25100767612457275, "learning_rate": 5.1800000000000005e-05, "loss": 0.0292, "step": 260 }, { "grad_norm": 0.1342630535364151, "learning_rate": 5.380000000000001e-05, "loss": 0.0291, "step": 270 }, { "grad_norm": 0.1496191918849945, "learning_rate": 5.580000000000001e-05, "loss": 0.0332, "step": 280 }, { "grad_norm": 0.1313694268465042, "learning_rate": 5.7799999999999995e-05, "loss": 0.0251, "step": 290 }, { "grad_norm": 0.13200165331363678, "learning_rate": 5.9800000000000003e-05, "loss": 0.0307, "step": 300 }, { "grad_norm": 0.16884654760360718, "learning_rate": 6.18e-05, "loss": 0.0266, "step": 310 }, { "grad_norm": 0.21465812623500824, "learning_rate": 6.38e-05, "loss": 0.0309, "step": 320 }, { "grad_norm": 0.22422099113464355, "learning_rate": 6.58e-05, "loss": 0.0345, "step": 330 }, { "grad_norm": 0.16468451917171478, "learning_rate": 6.780000000000001e-05, "loss": 0.0269, "step": 340 }, { "grad_norm": 0.1649431437253952, "learning_rate": 6.98e-05, "loss": 0.0285, "step": 350 }, { "grad_norm": 0.18796072900295258, "learning_rate": 7.18e-05, "loss": 0.0294, "step": 360 }, { "grad_norm": 0.1406984180212021, "learning_rate": 7.38e-05, "loss": 0.0262, "step": 370 }, { "grad_norm": 0.12663526833057404, "learning_rate": 7.58e-05, "loss": 0.0287, "step": 380 }, { "grad_norm": 0.17990556359291077, "learning_rate": 7.780000000000001e-05, "loss": 0.0358, "step": 390 }, { "grad_norm": 0.14493271708488464, "learning_rate": 7.98e-05, "loss": 0.0262, "step": 400 }, { "grad_norm": 0.18220767378807068, "learning_rate": 8.18e-05, "loss": 0.0307, "step": 410 }, { "grad_norm": 0.2229013890028, "learning_rate": 8.38e-05, "loss": 0.0393, "step": 420 }, { "grad_norm": 0.21163704991340637, "learning_rate": 8.58e-05, "loss": 0.0327, "step": 430 }, { "grad_norm": 0.1903618574142456, "learning_rate": 8.78e-05, "loss": 0.034, "step": 440 }, { "grad_norm": 0.19045576453208923, "learning_rate": 8.98e-05, "loss": 0.0334, "step": 450 }, { "grad_norm": 0.12919433414936066, "learning_rate": 9.180000000000001e-05, "loss": 0.0276, "step": 460 }, { "grad_norm": 0.22944214940071106, "learning_rate": 9.38e-05, "loss": 0.0366, "step": 470 }, { "grad_norm": 0.2297431230545044, "learning_rate": 9.58e-05, "loss": 0.031, "step": 480 }, { "grad_norm": 0.11911984533071518, "learning_rate": 9.78e-05, "loss": 0.0349, "step": 490 }, { "grad_norm": 0.14332818984985352, "learning_rate": 9.98e-05, "loss": 0.0298, "step": 500 }, { "grad_norm": 0.23667897284030914, "learning_rate": 9.9999778549206e-05, "loss": 0.0325, "step": 510 }, { "grad_norm": 0.1407025307416916, "learning_rate": 9.999901304280685e-05, "loss": 0.0275, "step": 520 }, { "grad_norm": 0.2208714336156845, "learning_rate": 9.999770075521164e-05, "loss": 0.0351, "step": 530 }, { "grad_norm": 0.18774442374706268, "learning_rate": 9.99958417007713e-05, "loss": 0.0415, "step": 540 }, { "grad_norm": 0.17202721536159515, "learning_rate": 9.999343589981615e-05, "loss": 0.0279, "step": 550 }, { "grad_norm": 0.1869555562734604, "learning_rate": 9.999048337865568e-05, "loss": 0.0325, "step": 560 }, { "grad_norm": 0.184967041015625, "learning_rate": 9.998698416957815e-05, "loss": 0.0426, "step": 570 }, { "grad_norm": 0.16213007271289825, "learning_rate": 9.998293831085037e-05, "loss": 0.0329, "step": 580 }, { "grad_norm": 0.1410742849111557, "learning_rate": 9.997834584671719e-05, "loss": 0.0313, "step": 590 }, { "grad_norm": 0.1963675171136856, "learning_rate": 9.997320682740107e-05, "loss": 0.0341, "step": 600 }, { "grad_norm": 0.18096579611301422, "learning_rate": 9.996752130910149e-05, "loss": 0.0378, "step": 610 }, { "grad_norm": 0.16188250482082367, "learning_rate": 9.99612893539944e-05, "loss": 0.0355, "step": 620 }, { "grad_norm": 0.20988130569458008, "learning_rate": 9.995451103023144e-05, "loss": 0.0327, "step": 630 }, { "grad_norm": 0.13830691576004028, "learning_rate": 9.994718641193928e-05, "loss": 0.0303, "step": 640 }, { "grad_norm": 0.10543777048587799, "learning_rate": 9.993931557921874e-05, "loss": 0.0311, "step": 650 }, { "grad_norm": 0.15627624094486237, "learning_rate": 9.993089861814402e-05, "loss": 0.0311, "step": 660 }, { "grad_norm": 0.14420993626117706, "learning_rate": 9.992193562076166e-05, "loss": 0.0256, "step": 670 }, { "grad_norm": 0.18450821936130524, "learning_rate": 9.991242668508954e-05, "loss": 0.04, "step": 680 }, { "grad_norm": 0.16379375755786896, "learning_rate": 9.990237191511587e-05, "loss": 0.0284, "step": 690 }, { "grad_norm": 0.14060579240322113, "learning_rate": 9.989177142079802e-05, "loss": 0.0276, "step": 700 }, { "grad_norm": 0.22451449930667877, "learning_rate": 9.988062531806126e-05, "loss": 0.0337, "step": 710 }, { "grad_norm": 0.13399986922740936, "learning_rate": 9.986893372879762e-05, "loss": 0.0292, "step": 720 }, { "grad_norm": 0.14917080104351044, "learning_rate": 9.985669678086443e-05, "loss": 0.0369, "step": 730 }, { "grad_norm": 0.2764376699924469, "learning_rate": 9.984391460808298e-05, "loss": 0.0334, "step": 740 }, { "grad_norm": 0.15942853689193726, "learning_rate": 9.983058735023709e-05, "loss": 0.0314, "step": 750 }, { "grad_norm": 0.19267232716083527, "learning_rate": 9.98167151530715e-05, "loss": 0.0302, "step": 760 }, { "grad_norm": 0.1857871413230896, "learning_rate": 9.980229816829034e-05, "loss": 0.0324, "step": 770 }, { "grad_norm": 0.09672777354717255, "learning_rate": 9.978733655355544e-05, "loss": 0.0336, "step": 780 }, { "grad_norm": 0.1687411516904831, "learning_rate": 9.977183047248464e-05, "loss": 0.0264, "step": 790 }, { "grad_norm": 0.15346349775791168, "learning_rate": 9.975578009464992e-05, "loss": 0.0342, "step": 800 }, { "grad_norm": 0.12507975101470947, "learning_rate": 9.97391855955757e-05, "loss": 0.0282, "step": 810 }, { "grad_norm": 0.1824805736541748, "learning_rate": 9.972204715673669e-05, "loss": 0.0475, "step": 820 }, { "grad_norm": 0.1841856837272644, "learning_rate": 9.970436496555617e-05, "loss": 0.0357, "step": 830 }, { "grad_norm": 0.19906769692897797, "learning_rate": 9.968613921540373e-05, "loss": 0.0295, "step": 840 }, { "grad_norm": 0.15382325649261475, "learning_rate": 9.966737010559326e-05, "loss": 0.0308, "step": 850 }, { "grad_norm": 0.16068829596042633, "learning_rate": 9.964805784138072e-05, "loss": 0.0353, "step": 860 }, { "grad_norm": 0.12692637741565704, "learning_rate": 9.962820263396195e-05, "loss": 0.0269, "step": 870 }, { "grad_norm": 0.17217664420604706, "learning_rate": 9.960780470047033e-05, "loss": 0.0329, "step": 880 }, { "grad_norm": 0.1513223648071289, "learning_rate": 9.958686426397437e-05, "loss": 0.0297, "step": 890 }, { "grad_norm": 0.14868152141571045, "learning_rate": 9.956538155347534e-05, "loss": 0.028, "step": 900 }, { "grad_norm": 0.16328009963035583, "learning_rate": 9.95433568039047e-05, "loss": 0.0308, "step": 910 }, { "grad_norm": 0.15237754583358765, "learning_rate": 9.952079025612162e-05, "loss": 0.0342, "step": 920 }, { "grad_norm": 0.17242884635925293, "learning_rate": 9.949768215691022e-05, "loss": 0.0404, "step": 930 }, { "grad_norm": 0.15023668110370636, "learning_rate": 9.9474032758977e-05, "loss": 0.0228, "step": 940 }, { "grad_norm": 0.16494718194007874, "learning_rate": 9.944984232094794e-05, "loss": 0.0287, "step": 950 }, { "grad_norm": 0.149717316031456, "learning_rate": 9.942511110736584e-05, "loss": 0.0337, "step": 960 }, { "grad_norm": 0.14683957397937775, "learning_rate": 9.939983938868726e-05, "loss": 0.0281, "step": 970 }, { "grad_norm": 0.18098339438438416, "learning_rate": 9.93740274412797e-05, "loss": 0.0309, "step": 980 }, { "grad_norm": 0.17116370797157288, "learning_rate": 9.934767554741846e-05, "loss": 0.0291, "step": 990 }, { "grad_norm": 0.13913299143314362, "learning_rate": 9.932078399528361e-05, "loss": 0.0267, "step": 1000 }, { "grad_norm": 0.1686856746673584, "learning_rate": 9.929335307895689e-05, "loss": 0.0363, "step": 1010 }, { "grad_norm": 0.13024087250232697, "learning_rate": 9.926538309841839e-05, "loss": 0.0285, "step": 1020 }, { "grad_norm": 0.17685119807720184, "learning_rate": 9.923687435954334e-05, "loss": 0.0287, "step": 1030 }, { "grad_norm": 0.20717529952526093, "learning_rate": 9.920782717409873e-05, "loss": 0.027, "step": 1040 }, { "grad_norm": 0.15997067093849182, "learning_rate": 9.917824185973994e-05, "loss": 0.0283, "step": 1050 }, { "grad_norm": 0.13489104807376862, "learning_rate": 9.914811874000723e-05, "loss": 0.0339, "step": 1060 }, { "grad_norm": 0.1213482990860939, "learning_rate": 9.911745814432218e-05, "loss": 0.0247, "step": 1070 }, { "grad_norm": 0.13703866302967072, "learning_rate": 9.90862604079842e-05, "loss": 0.0288, "step": 1080 }, { "grad_norm": 0.18348318338394165, "learning_rate": 9.90545258721667e-05, "loss": 0.0297, "step": 1090 }, { "grad_norm": 0.11433199793100357, "learning_rate": 9.90222548839135e-05, "loss": 0.0268, "step": 1100 }, { "grad_norm": 0.17286507785320282, "learning_rate": 9.898944779613495e-05, "loss": 0.0267, "step": 1110 }, { "grad_norm": 0.1405310034751892, "learning_rate": 9.89561049676041e-05, "loss": 0.0264, "step": 1120 }, { "grad_norm": 0.10237786918878555, "learning_rate": 9.89222267629528e-05, "loss": 0.0272, "step": 1130 }, { "grad_norm": 0.17079147696495056, "learning_rate": 9.888781355266763e-05, "loss": 0.0239, "step": 1140 }, { "grad_norm": 0.1960420310497284, "learning_rate": 9.885286571308598e-05, "loss": 0.0292, "step": 1150 }, { "grad_norm": 0.1341986507177353, "learning_rate": 9.881738362639182e-05, "loss": 0.0237, "step": 1160 }, { "grad_norm": 0.14957411587238312, "learning_rate": 9.878136768061154e-05, "loss": 0.0304, "step": 1170 }, { "grad_norm": 0.23255684971809387, "learning_rate": 9.874481826960979e-05, "loss": 0.0292, "step": 1180 }, { "grad_norm": 0.12505985796451569, "learning_rate": 9.870773579308503e-05, "loss": 0.0304, "step": 1190 }, { "grad_norm": 0.13537172973155975, "learning_rate": 9.867012065656533e-05, "loss": 0.0304, "step": 1200 }, { "grad_norm": 0.16207456588745117, "learning_rate": 9.863197327140376e-05, "loss": 0.035, "step": 1210 }, { "grad_norm": 0.22213716804981232, "learning_rate": 9.859329405477403e-05, "loss": 0.031, "step": 1220 }, { "grad_norm": 0.18120811879634857, "learning_rate": 9.855408342966585e-05, "loss": 0.0267, "step": 1230 }, { "grad_norm": 0.18635620176792145, "learning_rate": 9.851434182488033e-05, "loss": 0.0308, "step": 1240 }, { "grad_norm": 0.14335626363754272, "learning_rate": 9.84740696750253e-05, "loss": 0.0229, "step": 1250 }, { "grad_norm": 0.12551842629909515, "learning_rate": 9.843326742051055e-05, "loss": 0.0224, "step": 1260 }, { "grad_norm": 0.1251303106546402, "learning_rate": 9.839193550754297e-05, "loss": 0.0232, "step": 1270 }, { "grad_norm": 0.1565120369195938, "learning_rate": 9.835007438812177e-05, "loss": 0.03, "step": 1280 }, { "grad_norm": 0.15832260251045227, "learning_rate": 9.830768452003341e-05, "loss": 0.0271, "step": 1290 }, { "grad_norm": 0.13791042566299438, "learning_rate": 9.826476636684671e-05, "loss": 0.0276, "step": 1300 }, { "grad_norm": 0.10483084619045258, "learning_rate": 9.822132039790773e-05, "loss": 0.0248, "step": 1310 }, { "grad_norm": 0.14374341070652008, "learning_rate": 9.817734708833461e-05, "loss": 0.0269, "step": 1320 }, { "grad_norm": 0.16665680706501007, "learning_rate": 9.813284691901243e-05, "loss": 0.0307, "step": 1330 }, { "grad_norm": 0.21279016137123108, "learning_rate": 9.808782037658792e-05, "loss": 0.0333, "step": 1340 }, { "grad_norm": 0.1460651457309723, "learning_rate": 9.804226795346411e-05, "loss": 0.0242, "step": 1350 }, { "grad_norm": 0.20963597297668457, "learning_rate": 9.799619014779503e-05, "loss": 0.0321, "step": 1360 }, { "grad_norm": 0.19664771854877472, "learning_rate": 9.794958746348013e-05, "loss": 0.0257, "step": 1370 }, { "grad_norm": 0.2412680685520172, "learning_rate": 9.790246041015896e-05, "loss": 0.0265, "step": 1380 }, { "grad_norm": 0.2061948925256729, "learning_rate": 9.785480950320538e-05, "loss": 0.0289, "step": 1390 }, { "grad_norm": 0.1481916904449463, "learning_rate": 9.78066352637221e-05, "loss": 0.027, "step": 1400 }, { "grad_norm": 0.15597012639045715, "learning_rate": 9.775793821853488e-05, "loss": 0.0373, "step": 1410 }, { "grad_norm": 0.14834840595722198, "learning_rate": 9.77087189001868e-05, "loss": 0.0296, "step": 1420 }, { "grad_norm": 0.18620194494724274, "learning_rate": 9.765897784693243e-05, "loss": 0.0287, "step": 1430 }, { "grad_norm": 0.13414987921714783, "learning_rate": 9.760871560273197e-05, "loss": 0.0227, "step": 1440 }, { "grad_norm": 0.1416088491678238, "learning_rate": 9.755793271724526e-05, "loss": 0.0346, "step": 1450 }, { "grad_norm": 0.10656988620758057, "learning_rate": 9.750662974582584e-05, "loss": 0.027, "step": 1460 }, { "grad_norm": 0.12058360874652863, "learning_rate": 9.745480724951473e-05, "loss": 0.0199, "step": 1470 }, { "grad_norm": 0.21359531581401825, "learning_rate": 9.740246579503447e-05, "loss": 0.0251, "step": 1480 }, { "grad_norm": 0.15765275061130524, "learning_rate": 9.734960595478284e-05, "loss": 0.0245, "step": 1490 }, { "grad_norm": 0.13638831675052643, "learning_rate": 9.729622830682657e-05, "loss": 0.031, "step": 1500 }, { "grad_norm": 0.14469914138317108, "learning_rate": 9.724233343489504e-05, "loss": 0.0237, "step": 1510 }, { "grad_norm": 0.12858904898166656, "learning_rate": 9.718792192837396e-05, "loss": 0.028, "step": 1520 }, { "grad_norm": 0.18743933737277985, "learning_rate": 9.713299438229886e-05, "loss": 0.0266, "step": 1530 }, { "grad_norm": 0.17122581601142883, "learning_rate": 9.707755139734855e-05, "loss": 0.0249, "step": 1540 }, { "grad_norm": 0.125565305352211, "learning_rate": 9.702159357983866e-05, "loss": 0.0313, "step": 1550 }, { "grad_norm": 0.11599742621183395, "learning_rate": 9.696512154171492e-05, "loss": 0.0229, "step": 1560 }, { "grad_norm": 0.17274866998195648, "learning_rate": 9.690813590054645e-05, "loss": 0.0226, "step": 1570 }, { "grad_norm": 0.10541734844446182, "learning_rate": 9.685063727951914e-05, "loss": 0.0247, "step": 1580 }, { "grad_norm": 0.27482670545578003, "learning_rate": 9.679262630742865e-05, "loss": 0.0271, "step": 1590 }, { "grad_norm": 0.1570490002632141, "learning_rate": 9.673410361867373e-05, "loss": 0.0282, "step": 1600 }, { "grad_norm": 0.1765255182981491, "learning_rate": 9.667506985324909e-05, "loss": 0.0292, "step": 1610 }, { "grad_norm": 0.12382631003856659, "learning_rate": 9.661552565673855e-05, "loss": 0.0276, "step": 1620 }, { "grad_norm": 0.1488092988729477, "learning_rate": 9.655547168030789e-05, "loss": 0.0297, "step": 1630 }, { "grad_norm": 0.17621779441833496, "learning_rate": 9.649490858069777e-05, "loss": 0.0246, "step": 1640 }, { "grad_norm": 0.1317397505044937, "learning_rate": 9.643383702021658e-05, "loss": 0.0228, "step": 1650 }, { "grad_norm": 0.18702450394630432, "learning_rate": 9.637225766673307e-05, "loss": 0.0326, "step": 1660 }, { "grad_norm": 0.15152408182621002, "learning_rate": 9.631017119366922e-05, "loss": 0.025, "step": 1670 }, { "grad_norm": 0.14151324331760406, "learning_rate": 9.624757827999273e-05, "loss": 0.0234, "step": 1680 }, { "grad_norm": 0.17372380197048187, "learning_rate": 9.618447961020971e-05, "loss": 0.0361, "step": 1690 }, { "grad_norm": 0.14488667249679565, "learning_rate": 9.612087587435707e-05, "loss": 0.0248, "step": 1700 }, { "grad_norm": 0.1468273550271988, "learning_rate": 9.605676776799508e-05, "loss": 0.0236, "step": 1710 }, { "grad_norm": 0.16720327734947205, "learning_rate": 9.599215599219973e-05, "loss": 0.0299, "step": 1720 }, { "grad_norm": 0.10409308969974518, "learning_rate": 9.592704125355505e-05, "loss": 0.0224, "step": 1730 }, { "grad_norm": 0.22400495409965515, "learning_rate": 9.586142426414538e-05, "loss": 0.0298, "step": 1740 }, { "grad_norm": 0.16129139065742493, "learning_rate": 9.57953057415476e-05, "loss": 0.0216, "step": 1750 }, { "grad_norm": 0.15150457620620728, "learning_rate": 9.572868640882328e-05, "loss": 0.0255, "step": 1760 }, { "grad_norm": 0.27912425994873047, "learning_rate": 9.56615669945108e-05, "loss": 0.0265, "step": 1770 }, { "grad_norm": 0.15684591233730316, "learning_rate": 9.55939482326173e-05, "loss": 0.0231, "step": 1780 }, { "grad_norm": 0.1414581537246704, "learning_rate": 9.552583086261069e-05, "loss": 0.0298, "step": 1790 }, { "grad_norm": 0.12455716729164124, "learning_rate": 9.545721562941168e-05, "loss": 0.0311, "step": 1800 }, { "grad_norm": 0.1510583609342575, "learning_rate": 9.538810328338543e-05, "loss": 0.0245, "step": 1810 }, { "grad_norm": 0.12179003655910492, "learning_rate": 9.531849458033349e-05, "loss": 0.0195, "step": 1820 }, { "grad_norm": 0.15041644871234894, "learning_rate": 9.524839028148547e-05, "loss": 0.0237, "step": 1830 }, { "grad_norm": 0.18131820857524872, "learning_rate": 9.517779115349077e-05, "loss": 0.027, "step": 1840 }, { "grad_norm": 0.13325054943561554, "learning_rate": 9.510669796841014e-05, "loss": 0.0222, "step": 1850 }, { "grad_norm": 0.14594273269176483, "learning_rate": 9.503511150370727e-05, "loss": 0.0252, "step": 1860 }, { "grad_norm": 0.1101432591676712, "learning_rate": 9.496303254224024e-05, "loss": 0.0196, "step": 1870 }, { "grad_norm": 0.11899697780609131, "learning_rate": 9.489046187225306e-05, "loss": 0.0238, "step": 1880 }, { "grad_norm": 0.1834312528371811, "learning_rate": 9.481740028736692e-05, "loss": 0.027, "step": 1890 }, { "grad_norm": 0.1628856211900711, "learning_rate": 9.474384858657164e-05, "loss": 0.033, "step": 1900 }, { "grad_norm": 0.16760051250457764, "learning_rate": 9.466980757421679e-05, "loss": 0.0237, "step": 1910 }, { "grad_norm": 0.15349121391773224, "learning_rate": 9.459527806000305e-05, "loss": 0.0201, "step": 1920 }, { "grad_norm": 0.18761931359767914, "learning_rate": 9.452026085897325e-05, "loss": 0.0263, "step": 1930 }, { "grad_norm": 0.13588790595531464, "learning_rate": 9.444475679150348e-05, "loss": 0.0199, "step": 1940 }, { "grad_norm": 0.11293373256921768, "learning_rate": 9.436876668329411e-05, "loss": 0.0265, "step": 1950 }, { "grad_norm": 0.15826141834259033, "learning_rate": 9.429229136536079e-05, "loss": 0.0253, "step": 1960 }, { "grad_norm": 0.1520809531211853, "learning_rate": 9.421533167402534e-05, "loss": 0.0245, "step": 1970 }, { "grad_norm": 0.12347332388162613, "learning_rate": 9.413788845090666e-05, "loss": 0.0294, "step": 1980 }, { "grad_norm": 0.13870365917682648, "learning_rate": 9.405996254291136e-05, "loss": 0.022, "step": 1990 }, { "grad_norm": 0.1662704199552536, "learning_rate": 9.398155480222474e-05, "loss": 0.0267, "step": 2000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2000, "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": 16, "trial_name": null, "trial_params": null }