{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.9992332216796482, "eval_steps": 500, "global_step": 9291, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00032285402962185724, "grad_norm": 1.3199406548084531, "learning_rate": 3.2258064516129035e-08, "loss": 0.4506, "step": 1 }, { "epoch": 0.0006457080592437145, "grad_norm": 1.188714527760951, "learning_rate": 6.451612903225807e-08, "loss": 0.5095, "step": 2 }, { "epoch": 0.0009685620888655717, "grad_norm": 0.9046215227514233, "learning_rate": 9.67741935483871e-08, "loss": 0.4429, "step": 3 }, { "epoch": 0.001291416118487429, "grad_norm": 1.1528494717577689, "learning_rate": 1.2903225806451614e-07, "loss": 0.4868, "step": 4 }, { "epoch": 0.001614270148109286, "grad_norm": 1.9400010133157382, "learning_rate": 1.6129032258064518e-07, "loss": 0.5047, "step": 5 }, { "epoch": 0.0019371241777311433, "grad_norm": 1.8150772460877453, "learning_rate": 1.935483870967742e-07, "loss": 0.4491, "step": 6 }, { "epoch": 0.0022599782073530004, "grad_norm": 0.9225140580284078, "learning_rate": 2.2580645161290325e-07, "loss": 0.4464, "step": 7 }, { "epoch": 0.002582832236974858, "grad_norm": 1.3873485482357817, "learning_rate": 2.580645161290323e-07, "loss": 0.4915, "step": 8 }, { "epoch": 0.002905686266596715, "grad_norm": 1.8536886471098624, "learning_rate": 2.903225806451613e-07, "loss": 0.4376, "step": 9 }, { "epoch": 0.003228540296218572, "grad_norm": 1.651108854295682, "learning_rate": 3.2258064516129035e-07, "loss": 0.4587, "step": 10 }, { "epoch": 0.0035513943258404295, "grad_norm": 1.030178041620715, "learning_rate": 3.5483870967741936e-07, "loss": 0.4832, "step": 11 }, { "epoch": 0.0038742483554622866, "grad_norm": 1.8666448187355944, "learning_rate": 3.870967741935484e-07, "loss": 0.4499, "step": 12 }, { "epoch": 0.004197102385084144, "grad_norm": 1.1438242131752092, "learning_rate": 4.1935483870967743e-07, "loss": 0.5118, "step": 13 }, { "epoch": 0.004519956414706001, "grad_norm": 0.8905573055094166, "learning_rate": 4.516129032258065e-07, "loss": 0.4288, "step": 14 }, { "epoch": 0.004842810444327858, "grad_norm": 1.5658445641970862, "learning_rate": 4.838709677419355e-07, "loss": 0.4146, "step": 15 }, { "epoch": 0.005165664473949716, "grad_norm": 1.1812103293214502, "learning_rate": 5.161290322580646e-07, "loss": 0.5346, "step": 16 }, { "epoch": 0.005488518503571573, "grad_norm": 1.883153858056491, "learning_rate": 5.483870967741936e-07, "loss": 0.4603, "step": 17 }, { "epoch": 0.00581137253319343, "grad_norm": 2.9940691045376533, "learning_rate": 5.806451612903226e-07, "loss": 0.4623, "step": 18 }, { "epoch": 0.006134226562815287, "grad_norm": 0.9654164183567651, "learning_rate": 6.129032258064516e-07, "loss": 0.432, "step": 19 }, { "epoch": 0.006457080592437144, "grad_norm": 0.9491031103636595, "learning_rate": 6.451612903225807e-07, "loss": 0.4258, "step": 20 }, { "epoch": 0.006779934622059001, "grad_norm": 1.0894469617414633, "learning_rate": 6.774193548387097e-07, "loss": 0.4691, "step": 21 }, { "epoch": 0.007102788651680859, "grad_norm": 1.105440702659253, "learning_rate": 7.096774193548387e-07, "loss": 0.4215, "step": 22 }, { "epoch": 0.007425642681302716, "grad_norm": 1.215787430614567, "learning_rate": 7.419354838709678e-07, "loss": 0.4692, "step": 23 }, { "epoch": 0.007748496710924573, "grad_norm": 1.2547113264178558, "learning_rate": 7.741935483870968e-07, "loss": 0.3913, "step": 24 }, { "epoch": 0.008071350740546431, "grad_norm": 0.9292880055668101, "learning_rate": 8.064516129032258e-07, "loss": 0.3939, "step": 25 }, { "epoch": 0.008394204770168287, "grad_norm": 1.2899443625370353, "learning_rate": 8.387096774193549e-07, "loss": 0.544, "step": 26 }, { "epoch": 0.008717058799790145, "grad_norm": 0.8883315660973733, "learning_rate": 8.709677419354839e-07, "loss": 0.4637, "step": 27 }, { "epoch": 0.009039912829412002, "grad_norm": 1.058286541040525, "learning_rate": 9.03225806451613e-07, "loss": 0.5112, "step": 28 }, { "epoch": 0.00936276685903386, "grad_norm": 1.1279470888363723, "learning_rate": 9.354838709677419e-07, "loss": 0.3983, "step": 29 }, { "epoch": 0.009685620888655716, "grad_norm": 1.1800717601470212, "learning_rate": 9.67741935483871e-07, "loss": 0.4436, "step": 30 }, { "epoch": 0.010008474918277574, "grad_norm": 1.1051802873697405, "learning_rate": 1e-06, "loss": 0.4007, "step": 31 }, { "epoch": 0.010331328947899432, "grad_norm": 0.8511591299325636, "learning_rate": 1.0322580645161291e-06, "loss": 0.4281, "step": 32 }, { "epoch": 0.010654182977521288, "grad_norm": 0.9317185121583991, "learning_rate": 1.064516129032258e-06, "loss": 0.4793, "step": 33 }, { "epoch": 0.010977037007143146, "grad_norm": 0.7479994531842168, "learning_rate": 1.0967741935483872e-06, "loss": 0.4661, "step": 34 }, { "epoch": 0.011299891036765002, "grad_norm": 1.1163532964803498, "learning_rate": 1.1290322580645162e-06, "loss": 0.4943, "step": 35 }, { "epoch": 0.01162274506638686, "grad_norm": 0.6064175355430488, "learning_rate": 1.1612903225806452e-06, "loss": 0.3749, "step": 36 }, { "epoch": 0.011945599096008718, "grad_norm": 0.7911796549070668, "learning_rate": 1.1935483870967743e-06, "loss": 0.5143, "step": 37 }, { "epoch": 0.012268453125630574, "grad_norm": 0.8049977752663156, "learning_rate": 1.2258064516129033e-06, "loss": 0.4427, "step": 38 }, { "epoch": 0.012591307155252432, "grad_norm": 1.0852531734705064, "learning_rate": 1.2580645161290324e-06, "loss": 0.4647, "step": 39 }, { "epoch": 0.012914161184874288, "grad_norm": 0.5419718829056737, "learning_rate": 1.2903225806451614e-06, "loss": 0.3984, "step": 40 }, { "epoch": 0.013237015214496146, "grad_norm": 0.5902623011479634, "learning_rate": 1.3225806451612904e-06, "loss": 0.3936, "step": 41 }, { "epoch": 0.013559869244118002, "grad_norm": 0.8225465783246161, "learning_rate": 1.3548387096774193e-06, "loss": 0.4303, "step": 42 }, { "epoch": 0.01388272327373986, "grad_norm": 0.7182204773142886, "learning_rate": 1.3870967741935483e-06, "loss": 0.4629, "step": 43 }, { "epoch": 0.014205577303361718, "grad_norm": 0.8288070182638163, "learning_rate": 1.4193548387096774e-06, "loss": 0.4227, "step": 44 }, { "epoch": 0.014528431332983574, "grad_norm": 0.8258844486993522, "learning_rate": 1.4516129032258064e-06, "loss": 0.46, "step": 45 }, { "epoch": 0.014851285362605432, "grad_norm": 0.5808360720834956, "learning_rate": 1.4838709677419356e-06, "loss": 0.4324, "step": 46 }, { "epoch": 0.015174139392227289, "grad_norm": 0.5909059422060241, "learning_rate": 1.5161290322580645e-06, "loss": 0.4247, "step": 47 }, { "epoch": 0.015496993421849146, "grad_norm": 0.6308848222927005, "learning_rate": 1.5483870967741937e-06, "loss": 0.4543, "step": 48 }, { "epoch": 0.015819847451471004, "grad_norm": 0.5695818315817106, "learning_rate": 1.5806451612903226e-06, "loss": 0.4229, "step": 49 }, { "epoch": 0.016142701481092862, "grad_norm": 0.5010467812315175, "learning_rate": 1.6129032258064516e-06, "loss": 0.42, "step": 50 }, { "epoch": 0.016465555510714717, "grad_norm": 0.6128183809834437, "learning_rate": 1.6451612903225808e-06, "loss": 0.4173, "step": 51 }, { "epoch": 0.016788409540336575, "grad_norm": 0.5559830175226359, "learning_rate": 1.6774193548387097e-06, "loss": 0.4577, "step": 52 }, { "epoch": 0.017111263569958433, "grad_norm": 0.498190857766927, "learning_rate": 1.7096774193548389e-06, "loss": 0.3536, "step": 53 }, { "epoch": 0.01743411759958029, "grad_norm": 0.7155645595327583, "learning_rate": 1.7419354838709678e-06, "loss": 0.4243, "step": 54 }, { "epoch": 0.01775697162920215, "grad_norm": 0.48188575324764876, "learning_rate": 1.774193548387097e-06, "loss": 0.3977, "step": 55 }, { "epoch": 0.018079825658824003, "grad_norm": 0.6377711568168617, "learning_rate": 1.806451612903226e-06, "loss": 0.4468, "step": 56 }, { "epoch": 0.01840267968844586, "grad_norm": 0.7857549491556688, "learning_rate": 1.8387096774193547e-06, "loss": 0.4593, "step": 57 }, { "epoch": 0.01872553371806772, "grad_norm": 0.5055361548393018, "learning_rate": 1.8709677419354839e-06, "loss": 0.436, "step": 58 }, { "epoch": 0.019048387747689577, "grad_norm": 0.5327210151316105, "learning_rate": 1.903225806451613e-06, "loss": 0.434, "step": 59 }, { "epoch": 0.01937124177731143, "grad_norm": 0.48852676809127193, "learning_rate": 1.935483870967742e-06, "loss": 0.46, "step": 60 }, { "epoch": 0.01969409580693329, "grad_norm": 0.5014230500128393, "learning_rate": 1.9677419354838714e-06, "loss": 0.4301, "step": 61 }, { "epoch": 0.020016949836555147, "grad_norm": 0.39580242519690184, "learning_rate": 2e-06, "loss": 0.3848, "step": 62 }, { "epoch": 0.020339803866177005, "grad_norm": 0.4467845504332293, "learning_rate": 2.0322580645161293e-06, "loss": 0.3976, "step": 63 }, { "epoch": 0.020662657895798863, "grad_norm": 0.7339256573521352, "learning_rate": 2.0645161290322582e-06, "loss": 0.3799, "step": 64 }, { "epoch": 0.020985511925420718, "grad_norm": 0.539109465801622, "learning_rate": 2.0967741935483868e-06, "loss": 0.398, "step": 65 }, { "epoch": 0.021308365955042576, "grad_norm": 0.7732704331378042, "learning_rate": 2.129032258064516e-06, "loss": 0.4612, "step": 66 }, { "epoch": 0.021631219984664433, "grad_norm": 0.35659572447292076, "learning_rate": 2.161290322580645e-06, "loss": 0.3917, "step": 67 }, { "epoch": 0.02195407401428629, "grad_norm": 0.4066823918392526, "learning_rate": 2.1935483870967745e-06, "loss": 0.3975, "step": 68 }, { "epoch": 0.02227692804390815, "grad_norm": 0.38309393278333814, "learning_rate": 2.225806451612903e-06, "loss": 0.3655, "step": 69 }, { "epoch": 0.022599782073530004, "grad_norm": 0.5284813445847606, "learning_rate": 2.2580645161290324e-06, "loss": 0.4232, "step": 70 }, { "epoch": 0.022922636103151862, "grad_norm": 0.5927220042231341, "learning_rate": 2.2903225806451614e-06, "loss": 0.49, "step": 71 }, { "epoch": 0.02324549013277372, "grad_norm": 0.5435989082201925, "learning_rate": 2.3225806451612903e-06, "loss": 0.3727, "step": 72 }, { "epoch": 0.023568344162395578, "grad_norm": 0.47248847107524977, "learning_rate": 2.3548387096774193e-06, "loss": 0.3427, "step": 73 }, { "epoch": 0.023891198192017436, "grad_norm": 0.44779098383187915, "learning_rate": 2.3870967741935486e-06, "loss": 0.4234, "step": 74 }, { "epoch": 0.02421405222163929, "grad_norm": 0.3705645695048192, "learning_rate": 2.4193548387096776e-06, "loss": 0.3656, "step": 75 }, { "epoch": 0.024536906251261148, "grad_norm": 0.4067226813428594, "learning_rate": 2.4516129032258066e-06, "loss": 0.4266, "step": 76 }, { "epoch": 0.024859760280883006, "grad_norm": 0.5018470745287859, "learning_rate": 2.4838709677419355e-06, "loss": 0.4504, "step": 77 }, { "epoch": 0.025182614310504864, "grad_norm": 0.5008838675070926, "learning_rate": 2.516129032258065e-06, "loss": 0.4833, "step": 78 }, { "epoch": 0.025505468340126722, "grad_norm": 0.5185678610978903, "learning_rate": 2.5483870967741934e-06, "loss": 0.3801, "step": 79 }, { "epoch": 0.025828322369748576, "grad_norm": 0.3911103712957039, "learning_rate": 2.580645161290323e-06, "loss": 0.3842, "step": 80 }, { "epoch": 0.026151176399370434, "grad_norm": 0.43608258381873505, "learning_rate": 2.6129032258064518e-06, "loss": 0.3995, "step": 81 }, { "epoch": 0.026474030428992292, "grad_norm": 0.4831864805354886, "learning_rate": 2.6451612903225807e-06, "loss": 0.4986, "step": 82 }, { "epoch": 0.02679688445861415, "grad_norm": 0.4347359631551159, "learning_rate": 2.6774193548387097e-06, "loss": 0.3835, "step": 83 }, { "epoch": 0.027119738488236005, "grad_norm": 0.45647974387453893, "learning_rate": 2.7096774193548386e-06, "loss": 0.4602, "step": 84 }, { "epoch": 0.027442592517857863, "grad_norm": 0.3915011800309515, "learning_rate": 2.741935483870968e-06, "loss": 0.4237, "step": 85 }, { "epoch": 0.02776544654747972, "grad_norm": 0.49379645562466973, "learning_rate": 2.7741935483870965e-06, "loss": 0.3195, "step": 86 }, { "epoch": 0.02808830057710158, "grad_norm": 0.4996606249772866, "learning_rate": 2.806451612903226e-06, "loss": 0.4304, "step": 87 }, { "epoch": 0.028411154606723436, "grad_norm": 0.37258154153904427, "learning_rate": 2.838709677419355e-06, "loss": 0.3743, "step": 88 }, { "epoch": 0.02873400863634529, "grad_norm": 0.49323113525868173, "learning_rate": 2.8709677419354843e-06, "loss": 0.3887, "step": 89 }, { "epoch": 0.02905686266596715, "grad_norm": 0.38667364895319417, "learning_rate": 2.9032258064516128e-06, "loss": 0.3772, "step": 90 }, { "epoch": 0.029379716695589007, "grad_norm": 0.4161747610668827, "learning_rate": 2.935483870967742e-06, "loss": 0.4031, "step": 91 }, { "epoch": 0.029702570725210865, "grad_norm": 0.5513423149940934, "learning_rate": 2.967741935483871e-06, "loss": 0.4578, "step": 92 }, { "epoch": 0.030025424754832723, "grad_norm": 0.4943354974255142, "learning_rate": 3e-06, "loss": 0.4455, "step": 93 }, { "epoch": 0.030348278784454577, "grad_norm": 0.40958856102784774, "learning_rate": 3.032258064516129e-06, "loss": 0.3666, "step": 94 }, { "epoch": 0.030671132814076435, "grad_norm": 0.4734247195007574, "learning_rate": 3.0645161290322584e-06, "loss": 0.3899, "step": 95 }, { "epoch": 0.030993986843698293, "grad_norm": 0.3812952988021828, "learning_rate": 3.0967741935483874e-06, "loss": 0.4397, "step": 96 }, { "epoch": 0.03131684087332015, "grad_norm": 0.4385164321410885, "learning_rate": 3.129032258064516e-06, "loss": 0.3701, "step": 97 }, { "epoch": 0.03163969490294201, "grad_norm": 0.4534568978862401, "learning_rate": 3.1612903225806453e-06, "loss": 0.3731, "step": 98 }, { "epoch": 0.03196254893256387, "grad_norm": 0.7232802255102924, "learning_rate": 3.1935483870967742e-06, "loss": 0.3479, "step": 99 }, { "epoch": 0.032285402962185725, "grad_norm": 0.47714438857813973, "learning_rate": 3.225806451612903e-06, "loss": 0.4049, "step": 100 }, { "epoch": 0.032608256991807576, "grad_norm": 0.4508956604892757, "learning_rate": 3.258064516129032e-06, "loss": 0.4505, "step": 101 }, { "epoch": 0.032931111021429434, "grad_norm": 0.4602299246573128, "learning_rate": 3.2903225806451615e-06, "loss": 0.3988, "step": 102 }, { "epoch": 0.03325396505105129, "grad_norm": 0.3871479148037169, "learning_rate": 3.3225806451612905e-06, "loss": 0.4304, "step": 103 }, { "epoch": 0.03357681908067315, "grad_norm": 0.40901784596339513, "learning_rate": 3.3548387096774194e-06, "loss": 0.3908, "step": 104 }, { "epoch": 0.03389967311029501, "grad_norm": 0.44145985500858614, "learning_rate": 3.3870967741935484e-06, "loss": 0.312, "step": 105 }, { "epoch": 0.034222527139916865, "grad_norm": 0.38590711377992426, "learning_rate": 3.4193548387096778e-06, "loss": 0.3195, "step": 106 }, { "epoch": 0.03454538116953872, "grad_norm": 0.5000253486794277, "learning_rate": 3.4516129032258063e-06, "loss": 0.4026, "step": 107 }, { "epoch": 0.03486823519916058, "grad_norm": 0.48753901932876303, "learning_rate": 3.4838709677419357e-06, "loss": 0.3996, "step": 108 }, { "epoch": 0.03519108922878244, "grad_norm": 0.40333422388589457, "learning_rate": 3.5161290322580646e-06, "loss": 0.3926, "step": 109 }, { "epoch": 0.0355139432584043, "grad_norm": 0.468528833453181, "learning_rate": 3.548387096774194e-06, "loss": 0.4611, "step": 110 }, { "epoch": 0.03583679728802615, "grad_norm": 0.5409235687546141, "learning_rate": 3.5806451612903225e-06, "loss": 0.4398, "step": 111 }, { "epoch": 0.036159651317648006, "grad_norm": 0.40271514223439164, "learning_rate": 3.612903225806452e-06, "loss": 0.3846, "step": 112 }, { "epoch": 0.036482505347269864, "grad_norm": 0.3827393160348145, "learning_rate": 3.645161290322581e-06, "loss": 0.4226, "step": 113 }, { "epoch": 0.03680535937689172, "grad_norm": 0.41134631595081794, "learning_rate": 3.6774193548387094e-06, "loss": 0.3958, "step": 114 }, { "epoch": 0.03712821340651358, "grad_norm": 0.546976194290893, "learning_rate": 3.709677419354839e-06, "loss": 0.4078, "step": 115 }, { "epoch": 0.03745106743613544, "grad_norm": 0.4326334491420776, "learning_rate": 3.7419354838709678e-06, "loss": 0.4034, "step": 116 }, { "epoch": 0.037773921465757296, "grad_norm": 0.44571377718095145, "learning_rate": 3.7741935483870967e-06, "loss": 0.4553, "step": 117 }, { "epoch": 0.038096775495379154, "grad_norm": 0.49468194019703793, "learning_rate": 3.806451612903226e-06, "loss": 0.3966, "step": 118 }, { "epoch": 0.03841962952500101, "grad_norm": 0.4054206066111733, "learning_rate": 3.838709677419355e-06, "loss": 0.3366, "step": 119 }, { "epoch": 0.03874248355462286, "grad_norm": 0.46132238237459505, "learning_rate": 3.870967741935484e-06, "loss": 0.3988, "step": 120 }, { "epoch": 0.03906533758424472, "grad_norm": 0.4439738174241595, "learning_rate": 3.9032258064516125e-06, "loss": 0.4552, "step": 121 }, { "epoch": 0.03938819161386658, "grad_norm": 0.42480990425251364, "learning_rate": 3.935483870967743e-06, "loss": 0.3701, "step": 122 }, { "epoch": 0.039711045643488437, "grad_norm": 0.39126649008903247, "learning_rate": 3.967741935483871e-06, "loss": 0.4402, "step": 123 }, { "epoch": 0.040033899673110294, "grad_norm": 0.3916048110173612, "learning_rate": 4e-06, "loss": 0.4471, "step": 124 }, { "epoch": 0.04035675370273215, "grad_norm": 0.3493419548128763, "learning_rate": 4.032258064516129e-06, "loss": 0.3897, "step": 125 }, { "epoch": 0.04067960773235401, "grad_norm": 0.45372530275027456, "learning_rate": 4.064516129032259e-06, "loss": 0.4683, "step": 126 }, { "epoch": 0.04100246176197587, "grad_norm": 0.49245335617669267, "learning_rate": 4.096774193548387e-06, "loss": 0.3771, "step": 127 }, { "epoch": 0.041325315791597726, "grad_norm": 0.45552242966163364, "learning_rate": 4.1290322580645165e-06, "loss": 0.4309, "step": 128 }, { "epoch": 0.041648169821219584, "grad_norm": 0.4048737444323636, "learning_rate": 4.161290322580645e-06, "loss": 0.4236, "step": 129 }, { "epoch": 0.041971023850841435, "grad_norm": 0.49438644472605614, "learning_rate": 4.1935483870967736e-06, "loss": 0.4693, "step": 130 }, { "epoch": 0.04229387788046329, "grad_norm": 0.5438100200706313, "learning_rate": 4.225806451612904e-06, "loss": 0.469, "step": 131 }, { "epoch": 0.04261673191008515, "grad_norm": 0.5083557127982168, "learning_rate": 4.258064516129032e-06, "loss": 0.3124, "step": 132 }, { "epoch": 0.04293958593970701, "grad_norm": 0.48023080834870346, "learning_rate": 4.290322580645161e-06, "loss": 0.4325, "step": 133 }, { "epoch": 0.04326243996932887, "grad_norm": 0.5815827285255464, "learning_rate": 4.32258064516129e-06, "loss": 0.4671, "step": 134 }, { "epoch": 0.043585293998950725, "grad_norm": 0.4011818165110346, "learning_rate": 4.35483870967742e-06, "loss": 0.5054, "step": 135 }, { "epoch": 0.04390814802857258, "grad_norm": 0.42963898468959205, "learning_rate": 4.387096774193549e-06, "loss": 0.368, "step": 136 }, { "epoch": 0.04423100205819444, "grad_norm": 0.44941877730574115, "learning_rate": 4.4193548387096775e-06, "loss": 0.4114, "step": 137 }, { "epoch": 0.0445538560878163, "grad_norm": 0.4000777638172532, "learning_rate": 4.451612903225806e-06, "loss": 0.3931, "step": 138 }, { "epoch": 0.04487671011743816, "grad_norm": 0.44713251894026296, "learning_rate": 4.483870967741936e-06, "loss": 0.3839, "step": 139 }, { "epoch": 0.04519956414706001, "grad_norm": 0.4678589723009284, "learning_rate": 4.516129032258065e-06, "loss": 0.4462, "step": 140 }, { "epoch": 0.045522418176681866, "grad_norm": 0.4027883146310558, "learning_rate": 4.548387096774193e-06, "loss": 0.4776, "step": 141 }, { "epoch": 0.045845272206303724, "grad_norm": 0.41951793444259833, "learning_rate": 4.580645161290323e-06, "loss": 0.3153, "step": 142 }, { "epoch": 0.04616812623592558, "grad_norm": 0.4443540311186898, "learning_rate": 4.612903225806452e-06, "loss": 0.4143, "step": 143 }, { "epoch": 0.04649098026554744, "grad_norm": 0.4123823181883054, "learning_rate": 4.645161290322581e-06, "loss": 0.4508, "step": 144 }, { "epoch": 0.0468138342951693, "grad_norm": 0.3627529803510313, "learning_rate": 4.67741935483871e-06, "loss": 0.3638, "step": 145 }, { "epoch": 0.047136688324791155, "grad_norm": 0.41764377258420515, "learning_rate": 4.7096774193548385e-06, "loss": 0.3838, "step": 146 }, { "epoch": 0.04745954235441301, "grad_norm": 0.4628212137732196, "learning_rate": 4.741935483870967e-06, "loss": 0.43, "step": 147 }, { "epoch": 0.04778239638403487, "grad_norm": 0.43825801016365, "learning_rate": 4.774193548387097e-06, "loss": 0.3778, "step": 148 }, { "epoch": 0.04810525041365672, "grad_norm": 0.4482903742279793, "learning_rate": 4.806451612903226e-06, "loss": 0.3877, "step": 149 }, { "epoch": 0.04842810444327858, "grad_norm": 0.532074659160904, "learning_rate": 4.838709677419355e-06, "loss": 0.3659, "step": 150 }, { "epoch": 0.04875095847290044, "grad_norm": 0.348174514037258, "learning_rate": 4.870967741935484e-06, "loss": 0.3519, "step": 151 }, { "epoch": 0.049073812502522296, "grad_norm": 0.42588665946739396, "learning_rate": 4.903225806451613e-06, "loss": 0.3507, "step": 152 }, { "epoch": 0.049396666532144154, "grad_norm": 0.6209020174489545, "learning_rate": 4.9354838709677425e-06, "loss": 0.3829, "step": 153 }, { "epoch": 0.04971952056176601, "grad_norm": 0.430367527602114, "learning_rate": 4.967741935483871e-06, "loss": 0.3704, "step": 154 }, { "epoch": 0.05004237459138787, "grad_norm": 0.3379346906851005, "learning_rate": 4.9999999999999996e-06, "loss": 0.3711, "step": 155 }, { "epoch": 0.05036522862100973, "grad_norm": 0.4723999881325296, "learning_rate": 5.03225806451613e-06, "loss": 0.3546, "step": 156 }, { "epoch": 0.050688082650631586, "grad_norm": 0.37733652387274935, "learning_rate": 5.064516129032258e-06, "loss": 0.3701, "step": 157 }, { "epoch": 0.051010936680253444, "grad_norm": 0.44209182769253214, "learning_rate": 5.096774193548387e-06, "loss": 0.3883, "step": 158 }, { "epoch": 0.051333790709875295, "grad_norm": 0.3810475414882496, "learning_rate": 5.129032258064516e-06, "loss": 0.3636, "step": 159 }, { "epoch": 0.05165664473949715, "grad_norm": 0.41528641245114406, "learning_rate": 5.161290322580646e-06, "loss": 0.3399, "step": 160 }, { "epoch": 0.05197949876911901, "grad_norm": 0.4537025118009885, "learning_rate": 5.193548387096774e-06, "loss": 0.3729, "step": 161 }, { "epoch": 0.05230235279874087, "grad_norm": 0.41085457826372646, "learning_rate": 5.2258064516129035e-06, "loss": 0.3351, "step": 162 }, { "epoch": 0.052625206828362726, "grad_norm": 0.4229297499314192, "learning_rate": 5.258064516129032e-06, "loss": 0.4314, "step": 163 }, { "epoch": 0.052948060857984584, "grad_norm": 0.3470989366169195, "learning_rate": 5.2903225806451614e-06, "loss": 0.4535, "step": 164 }, { "epoch": 0.05327091488760644, "grad_norm": 0.4278791337849025, "learning_rate": 5.322580645161291e-06, "loss": 0.3446, "step": 165 }, { "epoch": 0.0535937689172283, "grad_norm": 0.42321870575402504, "learning_rate": 5.354838709677419e-06, "loss": 0.3541, "step": 166 }, { "epoch": 0.05391662294685016, "grad_norm": 0.48484758324376626, "learning_rate": 5.387096774193549e-06, "loss": 0.4183, "step": 167 }, { "epoch": 0.05423947697647201, "grad_norm": 0.42330709999563615, "learning_rate": 5.419354838709677e-06, "loss": 0.4402, "step": 168 }, { "epoch": 0.05456233100609387, "grad_norm": 0.44941947825908024, "learning_rate": 5.451612903225807e-06, "loss": 0.3193, "step": 169 }, { "epoch": 0.054885185035715725, "grad_norm": 0.4337734610629833, "learning_rate": 5.483870967741936e-06, "loss": 0.4327, "step": 170 }, { "epoch": 0.05520803906533758, "grad_norm": 0.4572137584178627, "learning_rate": 5.5161290322580645e-06, "loss": 0.4295, "step": 171 }, { "epoch": 0.05553089309495944, "grad_norm": 0.3511540800928316, "learning_rate": 5.548387096774193e-06, "loss": 0.3401, "step": 172 }, { "epoch": 0.0558537471245813, "grad_norm": 0.46534944357289515, "learning_rate": 5.580645161290323e-06, "loss": 0.382, "step": 173 }, { "epoch": 0.05617660115420316, "grad_norm": 0.49486524472273213, "learning_rate": 5.612903225806452e-06, "loss": 0.5315, "step": 174 }, { "epoch": 0.056499455183825015, "grad_norm": 0.4277066658489196, "learning_rate": 5.64516129032258e-06, "loss": 0.3447, "step": 175 }, { "epoch": 0.05682230921344687, "grad_norm": 0.47890308105473917, "learning_rate": 5.67741935483871e-06, "loss": 0.3581, "step": 176 }, { "epoch": 0.05714516324306873, "grad_norm": 0.61503697897715, "learning_rate": 5.709677419354838e-06, "loss": 0.4115, "step": 177 }, { "epoch": 0.05746801727269058, "grad_norm": 0.47791155328800117, "learning_rate": 5.7419354838709685e-06, "loss": 0.3838, "step": 178 }, { "epoch": 0.05779087130231244, "grad_norm": 0.4049906760481688, "learning_rate": 5.774193548387097e-06, "loss": 0.4242, "step": 179 }, { "epoch": 0.0581137253319343, "grad_norm": 0.44906010327578566, "learning_rate": 5.8064516129032256e-06, "loss": 0.3856, "step": 180 }, { "epoch": 0.058436579361556155, "grad_norm": 0.4442182795111198, "learning_rate": 5.838709677419355e-06, "loss": 0.3283, "step": 181 }, { "epoch": 0.05875943339117801, "grad_norm": 0.43040170405960293, "learning_rate": 5.870967741935484e-06, "loss": 0.3568, "step": 182 }, { "epoch": 0.05908228742079987, "grad_norm": 0.4472126082921409, "learning_rate": 5.903225806451613e-06, "loss": 0.3238, "step": 183 }, { "epoch": 0.05940514145042173, "grad_norm": 0.4606663450501035, "learning_rate": 5.935483870967742e-06, "loss": 0.3333, "step": 184 }, { "epoch": 0.05972799548004359, "grad_norm": 0.4076765087224244, "learning_rate": 5.967741935483871e-06, "loss": 0.3902, "step": 185 }, { "epoch": 0.060050849509665445, "grad_norm": 0.7873237684723575, "learning_rate": 6e-06, "loss": 0.4217, "step": 186 }, { "epoch": 0.0603737035392873, "grad_norm": 0.36512250043515204, "learning_rate": 6.0322580645161295e-06, "loss": 0.3984, "step": 187 }, { "epoch": 0.060696557568909154, "grad_norm": 0.41472050873348754, "learning_rate": 6.064516129032258e-06, "loss": 0.3572, "step": 188 }, { "epoch": 0.06101941159853101, "grad_norm": 0.4552614116833296, "learning_rate": 6.096774193548387e-06, "loss": 0.4198, "step": 189 }, { "epoch": 0.06134226562815287, "grad_norm": 0.45812456624962106, "learning_rate": 6.129032258064517e-06, "loss": 0.4086, "step": 190 }, { "epoch": 0.06166511965777473, "grad_norm": 0.4473454414093753, "learning_rate": 6.161290322580645e-06, "loss": 0.3684, "step": 191 }, { "epoch": 0.061987973687396586, "grad_norm": 0.4586328094330001, "learning_rate": 6.193548387096775e-06, "loss": 0.3795, "step": 192 }, { "epoch": 0.062310827717018444, "grad_norm": 0.448391930723217, "learning_rate": 6.225806451612903e-06, "loss": 0.3387, "step": 193 }, { "epoch": 0.0626336817466403, "grad_norm": 0.3992789851680658, "learning_rate": 6.258064516129032e-06, "loss": 0.4105, "step": 194 }, { "epoch": 0.06295653577626216, "grad_norm": 0.3836159705274389, "learning_rate": 6.290322580645162e-06, "loss": 0.4165, "step": 195 }, { "epoch": 0.06327938980588402, "grad_norm": 0.5674679417208935, "learning_rate": 6.3225806451612906e-06, "loss": 0.4049, "step": 196 }, { "epoch": 0.06360224383550588, "grad_norm": 0.40620694974269267, "learning_rate": 6.354838709677419e-06, "loss": 0.3831, "step": 197 }, { "epoch": 0.06392509786512773, "grad_norm": 0.45168707559536747, "learning_rate": 6.3870967741935485e-06, "loss": 0.364, "step": 198 }, { "epoch": 0.06424795189474959, "grad_norm": 0.5332901239336445, "learning_rate": 6.419354838709678e-06, "loss": 0.4002, "step": 199 }, { "epoch": 0.06457080592437145, "grad_norm": 0.40871120792646665, "learning_rate": 6.451612903225806e-06, "loss": 0.4107, "step": 200 }, { "epoch": 0.06489365995399331, "grad_norm": 0.46287752421324163, "learning_rate": 6.483870967741936e-06, "loss": 0.399, "step": 201 }, { "epoch": 0.06521651398361515, "grad_norm": 0.44400558277131, "learning_rate": 6.516129032258064e-06, "loss": 0.3442, "step": 202 }, { "epoch": 0.06553936801323701, "grad_norm": 0.4568886097095703, "learning_rate": 6.548387096774194e-06, "loss": 0.3963, "step": 203 }, { "epoch": 0.06586222204285887, "grad_norm": 0.4192459514505742, "learning_rate": 6.580645161290323e-06, "loss": 0.3746, "step": 204 }, { "epoch": 0.06618507607248073, "grad_norm": 0.41487590474034647, "learning_rate": 6.612903225806452e-06, "loss": 0.4167, "step": 205 }, { "epoch": 0.06650793010210258, "grad_norm": 0.4199722040892203, "learning_rate": 6.645161290322581e-06, "loss": 0.3816, "step": 206 }, { "epoch": 0.06683078413172444, "grad_norm": 0.40629142459520934, "learning_rate": 6.67741935483871e-06, "loss": 0.327, "step": 207 }, { "epoch": 0.0671536381613463, "grad_norm": 0.43900649928832186, "learning_rate": 6.709677419354839e-06, "loss": 0.3763, "step": 208 }, { "epoch": 0.06747649219096816, "grad_norm": 0.4840169825057309, "learning_rate": 6.741935483870968e-06, "loss": 0.39, "step": 209 }, { "epoch": 0.06779934622059001, "grad_norm": 0.452414351946741, "learning_rate": 6.774193548387097e-06, "loss": 0.4129, "step": 210 }, { "epoch": 0.06812220025021187, "grad_norm": 0.49229439231241723, "learning_rate": 6.806451612903225e-06, "loss": 0.3274, "step": 211 }, { "epoch": 0.06844505427983373, "grad_norm": 0.3710765991667387, "learning_rate": 6.8387096774193555e-06, "loss": 0.3933, "step": 212 }, { "epoch": 0.06876790830945559, "grad_norm": 0.42579398368874116, "learning_rate": 6.870967741935484e-06, "loss": 0.4108, "step": 213 }, { "epoch": 0.06909076233907745, "grad_norm": 0.40347914576208876, "learning_rate": 6.903225806451613e-06, "loss": 0.351, "step": 214 }, { "epoch": 0.0694136163686993, "grad_norm": 0.5081470623553744, "learning_rate": 6.935483870967742e-06, "loss": 0.4225, "step": 215 }, { "epoch": 0.06973647039832116, "grad_norm": 0.3682476310847115, "learning_rate": 6.967741935483871e-06, "loss": 0.4011, "step": 216 }, { "epoch": 0.07005932442794302, "grad_norm": 0.5780245253492118, "learning_rate": 7e-06, "loss": 0.4655, "step": 217 }, { "epoch": 0.07038217845756488, "grad_norm": 0.4673661195967724, "learning_rate": 7.032258064516129e-06, "loss": 0.4078, "step": 218 }, { "epoch": 0.07070503248718674, "grad_norm": 0.4357379090335157, "learning_rate": 7.064516129032258e-06, "loss": 0.4107, "step": 219 }, { "epoch": 0.0710278865168086, "grad_norm": 0.42482748581875523, "learning_rate": 7.096774193548388e-06, "loss": 0.364, "step": 220 }, { "epoch": 0.07135074054643044, "grad_norm": 0.5227541407539493, "learning_rate": 7.1290322580645166e-06, "loss": 0.3427, "step": 221 }, { "epoch": 0.0716735945760523, "grad_norm": 0.44730285633288, "learning_rate": 7.161290322580645e-06, "loss": 0.3882, "step": 222 }, { "epoch": 0.07199644860567415, "grad_norm": 0.4818031297482707, "learning_rate": 7.1935483870967745e-06, "loss": 0.4354, "step": 223 }, { "epoch": 0.07231930263529601, "grad_norm": 0.4718426035571103, "learning_rate": 7.225806451612904e-06, "loss": 0.3642, "step": 224 }, { "epoch": 0.07264215666491787, "grad_norm": 0.45726506306900283, "learning_rate": 7.258064516129032e-06, "loss": 0.4159, "step": 225 }, { "epoch": 0.07296501069453973, "grad_norm": 0.41551544007035857, "learning_rate": 7.290322580645162e-06, "loss": 0.3407, "step": 226 }, { "epoch": 0.07328786472416159, "grad_norm": 0.5770381894485794, "learning_rate": 7.32258064516129e-06, "loss": 0.3894, "step": 227 }, { "epoch": 0.07361071875378344, "grad_norm": 0.5209414831886362, "learning_rate": 7.354838709677419e-06, "loss": 0.4168, "step": 228 }, { "epoch": 0.0739335727834053, "grad_norm": 0.38237491940662666, "learning_rate": 7.387096774193549e-06, "loss": 0.33, "step": 229 }, { "epoch": 0.07425642681302716, "grad_norm": 0.4439928675824317, "learning_rate": 7.419354838709678e-06, "loss": 0.3184, "step": 230 }, { "epoch": 0.07457928084264902, "grad_norm": 0.4293909750010143, "learning_rate": 7.451612903225806e-06, "loss": 0.4107, "step": 231 }, { "epoch": 0.07490213487227088, "grad_norm": 0.39876880785745616, "learning_rate": 7.4838709677419355e-06, "loss": 0.2954, "step": 232 }, { "epoch": 0.07522498890189273, "grad_norm": 0.4546982329574137, "learning_rate": 7.516129032258065e-06, "loss": 0.3225, "step": 233 }, { "epoch": 0.07554784293151459, "grad_norm": 0.37813738396914093, "learning_rate": 7.548387096774193e-06, "loss": 0.4306, "step": 234 }, { "epoch": 0.07587069696113645, "grad_norm": 0.387472613415561, "learning_rate": 7.580645161290323e-06, "loss": 0.4324, "step": 235 }, { "epoch": 0.07619355099075831, "grad_norm": 0.4326017500689359, "learning_rate": 7.612903225806452e-06, "loss": 0.3593, "step": 236 }, { "epoch": 0.07651640502038017, "grad_norm": 0.42383925074431666, "learning_rate": 7.64516129032258e-06, "loss": 0.3989, "step": 237 }, { "epoch": 0.07683925905000202, "grad_norm": 0.3895389294583021, "learning_rate": 7.67741935483871e-06, "loss": 0.3592, "step": 238 }, { "epoch": 0.07716211307962388, "grad_norm": 0.44864080428906117, "learning_rate": 7.70967741935484e-06, "loss": 0.3003, "step": 239 }, { "epoch": 0.07748496710924573, "grad_norm": 0.4618134484522457, "learning_rate": 7.741935483870968e-06, "loss": 0.3619, "step": 240 }, { "epoch": 0.07780782113886758, "grad_norm": 0.4140906683878775, "learning_rate": 7.774193548387097e-06, "loss": 0.3361, "step": 241 }, { "epoch": 0.07813067516848944, "grad_norm": 0.48977623370695483, "learning_rate": 7.806451612903225e-06, "loss": 0.4009, "step": 242 }, { "epoch": 0.0784535291981113, "grad_norm": 0.49711953327150793, "learning_rate": 7.838709677419354e-06, "loss": 0.364, "step": 243 }, { "epoch": 0.07877638322773316, "grad_norm": 0.37629128649340693, "learning_rate": 7.870967741935486e-06, "loss": 0.4272, "step": 244 }, { "epoch": 0.07909923725735502, "grad_norm": 0.3596806967731174, "learning_rate": 7.903225806451613e-06, "loss": 0.3626, "step": 245 }, { "epoch": 0.07942209128697687, "grad_norm": 0.366547459807563, "learning_rate": 7.935483870967743e-06, "loss": 0.3823, "step": 246 }, { "epoch": 0.07974494531659873, "grad_norm": 0.46387004732943643, "learning_rate": 7.96774193548387e-06, "loss": 0.3577, "step": 247 }, { "epoch": 0.08006779934622059, "grad_norm": 0.6092234905325334, "learning_rate": 8e-06, "loss": 0.3809, "step": 248 }, { "epoch": 0.08039065337584245, "grad_norm": 0.420691462123361, "learning_rate": 8.032258064516129e-06, "loss": 0.3615, "step": 249 }, { "epoch": 0.0807135074054643, "grad_norm": 0.35630692742298625, "learning_rate": 8.064516129032258e-06, "loss": 0.3776, "step": 250 }, { "epoch": 0.08103636143508616, "grad_norm": 0.4031861840412856, "learning_rate": 8.096774193548388e-06, "loss": 0.4147, "step": 251 }, { "epoch": 0.08135921546470802, "grad_norm": 0.39407006649399384, "learning_rate": 8.129032258064517e-06, "loss": 0.3815, "step": 252 }, { "epoch": 0.08168206949432988, "grad_norm": 0.3744894558983833, "learning_rate": 8.161290322580645e-06, "loss": 0.3428, "step": 253 }, { "epoch": 0.08200492352395174, "grad_norm": 0.4433876629501691, "learning_rate": 8.193548387096774e-06, "loss": 0.3772, "step": 254 }, { "epoch": 0.0823277775535736, "grad_norm": 0.42699440715669745, "learning_rate": 8.225806451612904e-06, "loss": 0.3278, "step": 255 }, { "epoch": 0.08265063158319545, "grad_norm": 0.47540960508289415, "learning_rate": 8.258064516129033e-06, "loss": 0.3671, "step": 256 }, { "epoch": 0.08297348561281731, "grad_norm": 0.45599475825354274, "learning_rate": 8.290322580645162e-06, "loss": 0.3543, "step": 257 }, { "epoch": 0.08329633964243917, "grad_norm": 0.49141476040289617, "learning_rate": 8.32258064516129e-06, "loss": 0.446, "step": 258 }, { "epoch": 0.08361919367206103, "grad_norm": 0.5621161729807557, "learning_rate": 8.35483870967742e-06, "loss": 0.3193, "step": 259 }, { "epoch": 0.08394204770168287, "grad_norm": 0.4307585370646794, "learning_rate": 8.387096774193547e-06, "loss": 0.4722, "step": 260 }, { "epoch": 0.08426490173130473, "grad_norm": 0.39560988697897814, "learning_rate": 8.419354838709678e-06, "loss": 0.3728, "step": 261 }, { "epoch": 0.08458775576092659, "grad_norm": 0.38268946393216563, "learning_rate": 8.451612903225808e-06, "loss": 0.393, "step": 262 }, { "epoch": 0.08491060979054844, "grad_norm": 0.44770712618741043, "learning_rate": 8.483870967741935e-06, "loss": 0.473, "step": 263 }, { "epoch": 0.0852334638201703, "grad_norm": 0.3455211887043239, "learning_rate": 8.516129032258065e-06, "loss": 0.3657, "step": 264 }, { "epoch": 0.08555631784979216, "grad_norm": 0.4892019876695612, "learning_rate": 8.548387096774194e-06, "loss": 0.3503, "step": 265 }, { "epoch": 0.08587917187941402, "grad_norm": 0.4429648804079717, "learning_rate": 8.580645161290322e-06, "loss": 0.3638, "step": 266 }, { "epoch": 0.08620202590903588, "grad_norm": 0.49549148358056877, "learning_rate": 8.612903225806453e-06, "loss": 0.3995, "step": 267 }, { "epoch": 0.08652487993865773, "grad_norm": 0.4141332760693083, "learning_rate": 8.64516129032258e-06, "loss": 0.3441, "step": 268 }, { "epoch": 0.08684773396827959, "grad_norm": 0.5626074228829638, "learning_rate": 8.67741935483871e-06, "loss": 0.3138, "step": 269 }, { "epoch": 0.08717058799790145, "grad_norm": 0.5307505418733098, "learning_rate": 8.70967741935484e-06, "loss": 0.2895, "step": 270 }, { "epoch": 0.08749344202752331, "grad_norm": 0.4994131684124534, "learning_rate": 8.741935483870967e-06, "loss": 0.4595, "step": 271 }, { "epoch": 0.08781629605714517, "grad_norm": 0.4687292429974628, "learning_rate": 8.774193548387098e-06, "loss": 0.3713, "step": 272 }, { "epoch": 0.08813915008676702, "grad_norm": 0.44615616226520727, "learning_rate": 8.806451612903226e-06, "loss": 0.3681, "step": 273 }, { "epoch": 0.08846200411638888, "grad_norm": 0.4289051939207573, "learning_rate": 8.838709677419355e-06, "loss": 0.4306, "step": 274 }, { "epoch": 0.08878485814601074, "grad_norm": 0.5166759990628004, "learning_rate": 8.870967741935484e-06, "loss": 0.3406, "step": 275 }, { "epoch": 0.0891077121756326, "grad_norm": 0.5512396744396314, "learning_rate": 8.903225806451612e-06, "loss": 0.3581, "step": 276 }, { "epoch": 0.08943056620525446, "grad_norm": 0.5094330269961121, "learning_rate": 8.935483870967741e-06, "loss": 0.3759, "step": 277 }, { "epoch": 0.08975342023487631, "grad_norm": 0.4136626338662641, "learning_rate": 8.967741935483873e-06, "loss": 0.3763, "step": 278 }, { "epoch": 0.09007627426449816, "grad_norm": 0.5475229729144669, "learning_rate": 9e-06, "loss": 0.3171, "step": 279 }, { "epoch": 0.09039912829412002, "grad_norm": 0.34730919789030146, "learning_rate": 9.03225806451613e-06, "loss": 0.4188, "step": 280 }, { "epoch": 0.09072198232374187, "grad_norm": 0.41493063578000455, "learning_rate": 9.064516129032257e-06, "loss": 0.3905, "step": 281 }, { "epoch": 0.09104483635336373, "grad_norm": 0.39191165387345145, "learning_rate": 9.096774193548387e-06, "loss": 0.3077, "step": 282 }, { "epoch": 0.09136769038298559, "grad_norm": 0.37292928922953456, "learning_rate": 9.129032258064518e-06, "loss": 0.4383, "step": 283 }, { "epoch": 0.09169054441260745, "grad_norm": 0.4765372955669102, "learning_rate": 9.161290322580645e-06, "loss": 0.3893, "step": 284 }, { "epoch": 0.0920133984422293, "grad_norm": 0.43242360053847717, "learning_rate": 9.193548387096775e-06, "loss": 0.3604, "step": 285 }, { "epoch": 0.09233625247185116, "grad_norm": 0.520080724680575, "learning_rate": 9.225806451612904e-06, "loss": 0.3953, "step": 286 }, { "epoch": 0.09265910650147302, "grad_norm": 0.4114820902752855, "learning_rate": 9.258064516129032e-06, "loss": 0.4145, "step": 287 }, { "epoch": 0.09298196053109488, "grad_norm": 0.5154464323004628, "learning_rate": 9.290322580645161e-06, "loss": 0.3205, "step": 288 }, { "epoch": 0.09330481456071674, "grad_norm": 0.4234345475767889, "learning_rate": 9.32258064516129e-06, "loss": 0.3772, "step": 289 }, { "epoch": 0.0936276685903386, "grad_norm": 0.5796757589509384, "learning_rate": 9.35483870967742e-06, "loss": 0.3733, "step": 290 }, { "epoch": 0.09395052261996045, "grad_norm": 0.36223700456568947, "learning_rate": 9.38709677419355e-06, "loss": 0.3766, "step": 291 }, { "epoch": 0.09427337664958231, "grad_norm": 0.43603007155042356, "learning_rate": 9.419354838709677e-06, "loss": 0.4244, "step": 292 }, { "epoch": 0.09459623067920417, "grad_norm": 1.2056232095821224, "learning_rate": 9.451612903225806e-06, "loss": 0.3941, "step": 293 }, { "epoch": 0.09491908470882603, "grad_norm": 0.48711882065461554, "learning_rate": 9.483870967741934e-06, "loss": 0.4154, "step": 294 }, { "epoch": 0.09524193873844788, "grad_norm": 0.4414889500915035, "learning_rate": 9.516129032258065e-06, "loss": 0.3345, "step": 295 }, { "epoch": 0.09556479276806974, "grad_norm": 0.4456670862813211, "learning_rate": 9.548387096774195e-06, "loss": 0.45, "step": 296 }, { "epoch": 0.0958876467976916, "grad_norm": 0.39111097215021906, "learning_rate": 9.580645161290322e-06, "loss": 0.3956, "step": 297 }, { "epoch": 0.09621050082731344, "grad_norm": 0.4084825198492268, "learning_rate": 9.612903225806452e-06, "loss": 0.4001, "step": 298 }, { "epoch": 0.0965333548569353, "grad_norm": 0.4380005233592883, "learning_rate": 9.645161290322581e-06, "loss": 0.383, "step": 299 }, { "epoch": 0.09685620888655716, "grad_norm": 0.4599992481361966, "learning_rate": 9.67741935483871e-06, "loss": 0.343, "step": 300 }, { "epoch": 0.09717906291617902, "grad_norm": 0.4123060351996824, "learning_rate": 9.70967741935484e-06, "loss": 0.4058, "step": 301 }, { "epoch": 0.09750191694580088, "grad_norm": 0.4665945449051988, "learning_rate": 9.741935483870967e-06, "loss": 0.4164, "step": 302 }, { "epoch": 0.09782477097542273, "grad_norm": 0.4423272607326496, "learning_rate": 9.774193548387097e-06, "loss": 0.3534, "step": 303 }, { "epoch": 0.09814762500504459, "grad_norm": 0.3521089894563374, "learning_rate": 9.806451612903226e-06, "loss": 0.3409, "step": 304 }, { "epoch": 0.09847047903466645, "grad_norm": 0.418823595349286, "learning_rate": 9.838709677419354e-06, "loss": 0.3746, "step": 305 }, { "epoch": 0.09879333306428831, "grad_norm": 0.47068157932105903, "learning_rate": 9.870967741935485e-06, "loss": 0.3913, "step": 306 }, { "epoch": 0.09911618709391017, "grad_norm": 0.44328657903943547, "learning_rate": 9.903225806451613e-06, "loss": 0.39, "step": 307 }, { "epoch": 0.09943904112353202, "grad_norm": 0.3875637411905789, "learning_rate": 9.935483870967742e-06, "loss": 0.3087, "step": 308 }, { "epoch": 0.09976189515315388, "grad_norm": 0.7388062546983841, "learning_rate": 9.967741935483871e-06, "loss": 0.363, "step": 309 }, { "epoch": 0.10008474918277574, "grad_norm": 0.4123784226095678, "learning_rate": 9.999999999999999e-06, "loss": 0.3498, "step": 310 }, { "epoch": 0.1004076032123976, "grad_norm": 0.4899695919728547, "learning_rate": 1.003225806451613e-05, "loss": 0.4163, "step": 311 }, { "epoch": 0.10073045724201946, "grad_norm": 0.44414475986757695, "learning_rate": 1.006451612903226e-05, "loss": 0.383, "step": 312 }, { "epoch": 0.10105331127164131, "grad_norm": 0.44898326862485755, "learning_rate": 1.0096774193548387e-05, "loss": 0.3806, "step": 313 }, { "epoch": 0.10137616530126317, "grad_norm": 0.565125241820857, "learning_rate": 1.0129032258064517e-05, "loss": 0.4005, "step": 314 }, { "epoch": 0.10169901933088503, "grad_norm": 0.6040080874886546, "learning_rate": 1.0161290322580644e-05, "loss": 0.373, "step": 315 }, { "epoch": 0.10202187336050689, "grad_norm": 0.38949396959601695, "learning_rate": 1.0193548387096774e-05, "loss": 0.3722, "step": 316 }, { "epoch": 0.10234472739012873, "grad_norm": 0.42682626259780204, "learning_rate": 1.0225806451612905e-05, "loss": 0.3632, "step": 317 }, { "epoch": 0.10266758141975059, "grad_norm": 0.4122010676151001, "learning_rate": 1.0258064516129032e-05, "loss": 0.3114, "step": 318 }, { "epoch": 0.10299043544937245, "grad_norm": 0.40234937111130203, "learning_rate": 1.0290322580645162e-05, "loss": 0.3992, "step": 319 }, { "epoch": 0.1033132894789943, "grad_norm": 0.4980647862538481, "learning_rate": 1.0322580645161291e-05, "loss": 0.3706, "step": 320 }, { "epoch": 0.10363614350861616, "grad_norm": 0.39495448988814025, "learning_rate": 1.0354838709677419e-05, "loss": 0.3644, "step": 321 }, { "epoch": 0.10395899753823802, "grad_norm": 0.766964459650835, "learning_rate": 1.0387096774193548e-05, "loss": 0.4451, "step": 322 }, { "epoch": 0.10428185156785988, "grad_norm": 0.4179876848252694, "learning_rate": 1.0419354838709678e-05, "loss": 0.2757, "step": 323 }, { "epoch": 0.10460470559748174, "grad_norm": 0.6903591855556351, "learning_rate": 1.0451612903225807e-05, "loss": 0.3522, "step": 324 }, { "epoch": 0.1049275596271036, "grad_norm": 0.39986932026445193, "learning_rate": 1.0483870967741936e-05, "loss": 0.4043, "step": 325 }, { "epoch": 0.10525041365672545, "grad_norm": 0.422604653981704, "learning_rate": 1.0516129032258064e-05, "loss": 0.411, "step": 326 }, { "epoch": 0.10557326768634731, "grad_norm": 0.5524299118402151, "learning_rate": 1.0548387096774193e-05, "loss": 0.3583, "step": 327 }, { "epoch": 0.10589612171596917, "grad_norm": 0.4552031036711427, "learning_rate": 1.0580645161290323e-05, "loss": 0.3418, "step": 328 }, { "epoch": 0.10621897574559103, "grad_norm": 0.559620226249332, "learning_rate": 1.0612903225806452e-05, "loss": 0.3515, "step": 329 }, { "epoch": 0.10654182977521288, "grad_norm": 0.3830541756179776, "learning_rate": 1.0645161290322582e-05, "loss": 0.3509, "step": 330 }, { "epoch": 0.10686468380483474, "grad_norm": 0.38356999309785855, "learning_rate": 1.067741935483871e-05, "loss": 0.3062, "step": 331 }, { "epoch": 0.1071875378344566, "grad_norm": 0.4695302767431718, "learning_rate": 1.0709677419354839e-05, "loss": 0.4035, "step": 332 }, { "epoch": 0.10751039186407846, "grad_norm": 0.4846999544156362, "learning_rate": 1.0741935483870968e-05, "loss": 0.4322, "step": 333 }, { "epoch": 0.10783324589370032, "grad_norm": 0.4288476616889714, "learning_rate": 1.0774193548387097e-05, "loss": 0.3863, "step": 334 }, { "epoch": 0.10815609992332217, "grad_norm": 0.4945937323087069, "learning_rate": 1.0806451612903227e-05, "loss": 0.4621, "step": 335 }, { "epoch": 0.10847895395294402, "grad_norm": 0.4373764883117552, "learning_rate": 1.0838709677419355e-05, "loss": 0.3478, "step": 336 }, { "epoch": 0.10880180798256588, "grad_norm": 0.542069764687924, "learning_rate": 1.0870967741935484e-05, "loss": 0.36, "step": 337 }, { "epoch": 0.10912466201218773, "grad_norm": 0.40917550686964343, "learning_rate": 1.0903225806451613e-05, "loss": 0.3746, "step": 338 }, { "epoch": 0.10944751604180959, "grad_norm": 0.5584686078742193, "learning_rate": 1.0935483870967743e-05, "loss": 0.369, "step": 339 }, { "epoch": 0.10977037007143145, "grad_norm": 0.3972948379914541, "learning_rate": 1.0967741935483872e-05, "loss": 0.3972, "step": 340 }, { "epoch": 0.11009322410105331, "grad_norm": 0.4691351477199901, "learning_rate": 1.1e-05, "loss": 0.3808, "step": 341 }, { "epoch": 0.11041607813067517, "grad_norm": 0.4968035414196864, "learning_rate": 1.1032258064516129e-05, "loss": 0.3246, "step": 342 }, { "epoch": 0.11073893216029702, "grad_norm": 0.4723034437702499, "learning_rate": 1.1064516129032258e-05, "loss": 0.2748, "step": 343 }, { "epoch": 0.11106178618991888, "grad_norm": 0.42858773147617185, "learning_rate": 1.1096774193548386e-05, "loss": 0.3502, "step": 344 }, { "epoch": 0.11138464021954074, "grad_norm": 0.4731548050849856, "learning_rate": 1.1129032258064517e-05, "loss": 0.4131, "step": 345 }, { "epoch": 0.1117074942491626, "grad_norm": 0.41772892206132295, "learning_rate": 1.1161290322580647e-05, "loss": 0.3895, "step": 346 }, { "epoch": 0.11203034827878446, "grad_norm": 0.38387450420001895, "learning_rate": 1.1193548387096774e-05, "loss": 0.3809, "step": 347 }, { "epoch": 0.11235320230840631, "grad_norm": 0.4236301578352865, "learning_rate": 1.1225806451612904e-05, "loss": 0.4283, "step": 348 }, { "epoch": 0.11267605633802817, "grad_norm": 0.47084669820302016, "learning_rate": 1.1258064516129031e-05, "loss": 0.3349, "step": 349 }, { "epoch": 0.11299891036765003, "grad_norm": 0.36290104041097065, "learning_rate": 1.129032258064516e-05, "loss": 0.3533, "step": 350 }, { "epoch": 0.11332176439727189, "grad_norm": 0.4509795483157399, "learning_rate": 1.1322580645161292e-05, "loss": 0.3389, "step": 351 }, { "epoch": 0.11364461842689375, "grad_norm": 0.48133676972247846, "learning_rate": 1.135483870967742e-05, "loss": 0.3135, "step": 352 }, { "epoch": 0.1139674724565156, "grad_norm": 0.4549970590025242, "learning_rate": 1.1387096774193549e-05, "loss": 0.3204, "step": 353 }, { "epoch": 0.11429032648613746, "grad_norm": 0.4422957833869606, "learning_rate": 1.1419354838709677e-05, "loss": 0.4089, "step": 354 }, { "epoch": 0.11461318051575932, "grad_norm": 0.4626860267017343, "learning_rate": 1.1451612903225806e-05, "loss": 0.3653, "step": 355 }, { "epoch": 0.11493603454538116, "grad_norm": 0.3985757661334416, "learning_rate": 1.1483870967741937e-05, "loss": 0.3127, "step": 356 }, { "epoch": 0.11525888857500302, "grad_norm": 0.47789144345550133, "learning_rate": 1.1516129032258065e-05, "loss": 0.4007, "step": 357 }, { "epoch": 0.11558174260462488, "grad_norm": 0.5627158484356125, "learning_rate": 1.1548387096774194e-05, "loss": 0.4134, "step": 358 }, { "epoch": 0.11590459663424674, "grad_norm": 0.5049309903684572, "learning_rate": 1.1580645161290323e-05, "loss": 0.3841, "step": 359 }, { "epoch": 0.1162274506638686, "grad_norm": 0.476773869979708, "learning_rate": 1.1612903225806451e-05, "loss": 0.3599, "step": 360 }, { "epoch": 0.11655030469349045, "grad_norm": 0.48614150969313175, "learning_rate": 1.164516129032258e-05, "loss": 0.3634, "step": 361 }, { "epoch": 0.11687315872311231, "grad_norm": 0.48890828360973704, "learning_rate": 1.167741935483871e-05, "loss": 0.3734, "step": 362 }, { "epoch": 0.11719601275273417, "grad_norm": 0.46975611631480274, "learning_rate": 1.170967741935484e-05, "loss": 0.3346, "step": 363 }, { "epoch": 0.11751886678235603, "grad_norm": 0.43714390783855983, "learning_rate": 1.1741935483870969e-05, "loss": 0.354, "step": 364 }, { "epoch": 0.11784172081197788, "grad_norm": 0.40826686116113114, "learning_rate": 1.1774193548387096e-05, "loss": 0.4038, "step": 365 }, { "epoch": 0.11816457484159974, "grad_norm": 0.5049075367854239, "learning_rate": 1.1806451612903226e-05, "loss": 0.3655, "step": 366 }, { "epoch": 0.1184874288712216, "grad_norm": 0.41773541448946305, "learning_rate": 1.1838709677419355e-05, "loss": 0.3775, "step": 367 }, { "epoch": 0.11881028290084346, "grad_norm": 0.41218002798172876, "learning_rate": 1.1870967741935484e-05, "loss": 0.3904, "step": 368 }, { "epoch": 0.11913313693046532, "grad_norm": 0.41192023878215467, "learning_rate": 1.1903225806451614e-05, "loss": 0.4468, "step": 369 }, { "epoch": 0.11945599096008717, "grad_norm": 0.4795143830028729, "learning_rate": 1.1935483870967742e-05, "loss": 0.3525, "step": 370 }, { "epoch": 0.11977884498970903, "grad_norm": 0.3889076270941558, "learning_rate": 1.1967741935483871e-05, "loss": 0.3484, "step": 371 }, { "epoch": 0.12010169901933089, "grad_norm": 0.41941147071697416, "learning_rate": 1.2e-05, "loss": 0.3128, "step": 372 }, { "epoch": 0.12042455304895275, "grad_norm": 0.45588499479660016, "learning_rate": 1.203225806451613e-05, "loss": 0.4104, "step": 373 }, { "epoch": 0.1207474070785746, "grad_norm": 0.5414891829724029, "learning_rate": 1.2064516129032259e-05, "loss": 0.441, "step": 374 }, { "epoch": 0.12107026110819645, "grad_norm": 0.40218898294757016, "learning_rate": 1.2096774193548387e-05, "loss": 0.3372, "step": 375 }, { "epoch": 0.12139311513781831, "grad_norm": 0.6842782950607994, "learning_rate": 1.2129032258064516e-05, "loss": 0.3762, "step": 376 }, { "epoch": 0.12171596916744017, "grad_norm": 0.39670333277418457, "learning_rate": 1.2161290322580646e-05, "loss": 0.4093, "step": 377 }, { "epoch": 0.12203882319706202, "grad_norm": 0.3689448514631596, "learning_rate": 1.2193548387096773e-05, "loss": 0.3483, "step": 378 }, { "epoch": 0.12236167722668388, "grad_norm": 0.40787898786773624, "learning_rate": 1.2225806451612904e-05, "loss": 0.4275, "step": 379 }, { "epoch": 0.12268453125630574, "grad_norm": 0.43735781341857155, "learning_rate": 1.2258064516129034e-05, "loss": 0.404, "step": 380 }, { "epoch": 0.1230073852859276, "grad_norm": 0.4258888757113427, "learning_rate": 1.2290322580645161e-05, "loss": 0.4464, "step": 381 }, { "epoch": 0.12333023931554946, "grad_norm": 0.5154271580280633, "learning_rate": 1.232258064516129e-05, "loss": 0.354, "step": 382 }, { "epoch": 0.12365309334517131, "grad_norm": 0.49957323036583096, "learning_rate": 1.2354838709677418e-05, "loss": 0.3929, "step": 383 }, { "epoch": 0.12397594737479317, "grad_norm": 0.3739375380844241, "learning_rate": 1.238709677419355e-05, "loss": 0.3331, "step": 384 }, { "epoch": 0.12429880140441503, "grad_norm": 0.537122369534955, "learning_rate": 1.2419354838709679e-05, "loss": 0.3858, "step": 385 }, { "epoch": 0.12462165543403689, "grad_norm": 0.5277986714789101, "learning_rate": 1.2451612903225807e-05, "loss": 0.3881, "step": 386 }, { "epoch": 0.12494450946365875, "grad_norm": 0.33883553725898347, "learning_rate": 1.2483870967741936e-05, "loss": 0.3186, "step": 387 }, { "epoch": 0.1252673634932806, "grad_norm": 0.4319709945261223, "learning_rate": 1.2516129032258064e-05, "loss": 0.4192, "step": 388 }, { "epoch": 0.12559021752290245, "grad_norm": 0.4294997989199961, "learning_rate": 1.2548387096774193e-05, "loss": 0.3367, "step": 389 }, { "epoch": 0.12591307155252432, "grad_norm": 0.5378645317796624, "learning_rate": 1.2580645161290324e-05, "loss": 0.3592, "step": 390 }, { "epoch": 0.12623592558214616, "grad_norm": 0.4636124998198839, "learning_rate": 1.2612903225806452e-05, "loss": 0.3459, "step": 391 }, { "epoch": 0.12655877961176804, "grad_norm": 0.5180095618497267, "learning_rate": 1.2645161290322581e-05, "loss": 0.345, "step": 392 }, { "epoch": 0.12688163364138988, "grad_norm": 0.5247780350901461, "learning_rate": 1.267741935483871e-05, "loss": 0.4111, "step": 393 }, { "epoch": 0.12720448767101175, "grad_norm": 0.43024268720125297, "learning_rate": 1.2709677419354838e-05, "loss": 0.3323, "step": 394 }, { "epoch": 0.1275273417006336, "grad_norm": 0.48015798169299334, "learning_rate": 1.274193548387097e-05, "loss": 0.3509, "step": 395 }, { "epoch": 0.12785019573025547, "grad_norm": 0.4673015381944562, "learning_rate": 1.2774193548387097e-05, "loss": 0.3744, "step": 396 }, { "epoch": 0.1281730497598773, "grad_norm": 0.4499318756916378, "learning_rate": 1.2806451612903226e-05, "loss": 0.4136, "step": 397 }, { "epoch": 0.12849590378949918, "grad_norm": 0.40826154223576583, "learning_rate": 1.2838709677419356e-05, "loss": 0.3234, "step": 398 }, { "epoch": 0.12881875781912103, "grad_norm": 0.37949589900527914, "learning_rate": 1.2870967741935483e-05, "loss": 0.313, "step": 399 }, { "epoch": 0.1291416118487429, "grad_norm": 0.3608361258618144, "learning_rate": 1.2903225806451613e-05, "loss": 0.3381, "step": 400 }, { "epoch": 0.12946446587836474, "grad_norm": 0.4654854405291013, "learning_rate": 1.2935483870967742e-05, "loss": 0.3531, "step": 401 }, { "epoch": 0.12978731990798661, "grad_norm": 0.36758715064466163, "learning_rate": 1.2967741935483872e-05, "loss": 0.3621, "step": 402 }, { "epoch": 0.13011017393760846, "grad_norm": 0.544307760812289, "learning_rate": 1.3000000000000001e-05, "loss": 0.3429, "step": 403 }, { "epoch": 0.1304330279672303, "grad_norm": 0.4445170906630728, "learning_rate": 1.3032258064516129e-05, "loss": 0.3503, "step": 404 }, { "epoch": 0.13075588199685217, "grad_norm": 0.46831803303847663, "learning_rate": 1.3064516129032258e-05, "loss": 0.3939, "step": 405 }, { "epoch": 0.13107873602647402, "grad_norm": 0.5394626440276178, "learning_rate": 1.3096774193548387e-05, "loss": 0.4016, "step": 406 }, { "epoch": 0.1314015900560959, "grad_norm": 0.4499101149011981, "learning_rate": 1.3129032258064517e-05, "loss": 0.3544, "step": 407 }, { "epoch": 0.13172444408571773, "grad_norm": 0.42124322390000896, "learning_rate": 1.3161290322580646e-05, "loss": 0.3551, "step": 408 }, { "epoch": 0.1320472981153396, "grad_norm": 0.3043828430816434, "learning_rate": 1.3193548387096774e-05, "loss": 0.3107, "step": 409 }, { "epoch": 0.13237015214496145, "grad_norm": 0.375130721274051, "learning_rate": 1.3225806451612903e-05, "loss": 0.3574, "step": 410 }, { "epoch": 0.13269300617458332, "grad_norm": 0.44389345553088977, "learning_rate": 1.3258064516129033e-05, "loss": 0.4037, "step": 411 }, { "epoch": 0.13301586020420517, "grad_norm": 0.3908235579547251, "learning_rate": 1.3290322580645162e-05, "loss": 0.361, "step": 412 }, { "epoch": 0.13333871423382704, "grad_norm": 0.4348610369611594, "learning_rate": 1.3322580645161291e-05, "loss": 0.4327, "step": 413 }, { "epoch": 0.13366156826344888, "grad_norm": 0.47493336246918205, "learning_rate": 1.335483870967742e-05, "loss": 0.3657, "step": 414 }, { "epoch": 0.13398442229307075, "grad_norm": 0.4254698695657229, "learning_rate": 1.3387096774193548e-05, "loss": 0.368, "step": 415 }, { "epoch": 0.1343072763226926, "grad_norm": 0.35203789406626984, "learning_rate": 1.3419354838709678e-05, "loss": 0.431, "step": 416 }, { "epoch": 0.13463013035231447, "grad_norm": 0.355232503333859, "learning_rate": 1.3451612903225805e-05, "loss": 0.324, "step": 417 }, { "epoch": 0.13495298438193631, "grad_norm": 0.5569409808820412, "learning_rate": 1.3483870967741936e-05, "loss": 0.3867, "step": 418 }, { "epoch": 0.13527583841155819, "grad_norm": 0.4694295514068729, "learning_rate": 1.3516129032258066e-05, "loss": 0.3506, "step": 419 }, { "epoch": 0.13559869244118003, "grad_norm": 0.42380252436665994, "learning_rate": 1.3548387096774194e-05, "loss": 0.3564, "step": 420 }, { "epoch": 0.1359215464708019, "grad_norm": 0.4286794071155329, "learning_rate": 1.3580645161290323e-05, "loss": 0.444, "step": 421 }, { "epoch": 0.13624440050042375, "grad_norm": 0.399734438073307, "learning_rate": 1.361290322580645e-05, "loss": 0.3677, "step": 422 }, { "epoch": 0.1365672545300456, "grad_norm": 0.46138444916729776, "learning_rate": 1.3645161290322582e-05, "loss": 0.3701, "step": 423 }, { "epoch": 0.13689010855966746, "grad_norm": 0.40464529933846854, "learning_rate": 1.3677419354838711e-05, "loss": 0.3308, "step": 424 }, { "epoch": 0.1372129625892893, "grad_norm": 0.4713683183545651, "learning_rate": 1.3709677419354839e-05, "loss": 0.3791, "step": 425 }, { "epoch": 0.13753581661891118, "grad_norm": 0.5553184503948692, "learning_rate": 1.3741935483870968e-05, "loss": 0.3108, "step": 426 }, { "epoch": 0.13785867064853302, "grad_norm": 0.5282824415811157, "learning_rate": 1.3774193548387098e-05, "loss": 0.3655, "step": 427 }, { "epoch": 0.1381815246781549, "grad_norm": 0.37582241534766886, "learning_rate": 1.3806451612903225e-05, "loss": 0.3479, "step": 428 }, { "epoch": 0.13850437870777674, "grad_norm": 0.45260708796000887, "learning_rate": 1.3838709677419356e-05, "loss": 0.3145, "step": 429 }, { "epoch": 0.1388272327373986, "grad_norm": 0.45794162813216693, "learning_rate": 1.3870967741935484e-05, "loss": 0.3588, "step": 430 }, { "epoch": 0.13915008676702045, "grad_norm": 0.38966435898119406, "learning_rate": 1.3903225806451613e-05, "loss": 0.363, "step": 431 }, { "epoch": 0.13947294079664233, "grad_norm": 0.46514861075968733, "learning_rate": 1.3935483870967743e-05, "loss": 0.3596, "step": 432 }, { "epoch": 0.13979579482626417, "grad_norm": 0.4136724889836452, "learning_rate": 1.396774193548387e-05, "loss": 0.3678, "step": 433 }, { "epoch": 0.14011864885588604, "grad_norm": 0.4432667489055111, "learning_rate": 1.4e-05, "loss": 0.3772, "step": 434 }, { "epoch": 0.14044150288550789, "grad_norm": 0.4015346988159623, "learning_rate": 1.403225806451613e-05, "loss": 0.3456, "step": 435 }, { "epoch": 0.14076435691512976, "grad_norm": 0.4093411989339642, "learning_rate": 1.4064516129032259e-05, "loss": 0.4266, "step": 436 }, { "epoch": 0.1410872109447516, "grad_norm": 0.4210209092043211, "learning_rate": 1.4096774193548388e-05, "loss": 0.3808, "step": 437 }, { "epoch": 0.14141006497437347, "grad_norm": 0.47738352150898294, "learning_rate": 1.4129032258064516e-05, "loss": 0.3293, "step": 438 }, { "epoch": 0.14173291900399532, "grad_norm": 0.3818602484338046, "learning_rate": 1.4161290322580645e-05, "loss": 0.3884, "step": 439 }, { "epoch": 0.1420557730336172, "grad_norm": 0.43496652998463425, "learning_rate": 1.4193548387096776e-05, "loss": 0.3687, "step": 440 }, { "epoch": 0.14237862706323903, "grad_norm": 0.49473225808936133, "learning_rate": 1.4225806451612904e-05, "loss": 0.3443, "step": 441 }, { "epoch": 0.14270148109286088, "grad_norm": 0.3496854379284224, "learning_rate": 1.4258064516129033e-05, "loss": 0.3434, "step": 442 }, { "epoch": 0.14302433512248275, "grad_norm": 0.3712682594470735, "learning_rate": 1.429032258064516e-05, "loss": 0.3124, "step": 443 }, { "epoch": 0.1433471891521046, "grad_norm": 0.5274204718185155, "learning_rate": 1.432258064516129e-05, "loss": 0.4197, "step": 444 }, { "epoch": 0.14367004318172646, "grad_norm": 0.47920619689918614, "learning_rate": 1.435483870967742e-05, "loss": 0.3547, "step": 445 }, { "epoch": 0.1439928972113483, "grad_norm": 0.6087525333000466, "learning_rate": 1.4387096774193549e-05, "loss": 0.3561, "step": 446 }, { "epoch": 0.14431575124097018, "grad_norm": 0.40846739595918474, "learning_rate": 1.4419354838709678e-05, "loss": 0.3727, "step": 447 }, { "epoch": 0.14463860527059202, "grad_norm": 0.4451055437511441, "learning_rate": 1.4451612903225808e-05, "loss": 0.4006, "step": 448 }, { "epoch": 0.1449614593002139, "grad_norm": 0.43968496177310085, "learning_rate": 1.4483870967741935e-05, "loss": 0.3555, "step": 449 }, { "epoch": 0.14528431332983574, "grad_norm": 0.48153738302532945, "learning_rate": 1.4516129032258065e-05, "loss": 0.3758, "step": 450 }, { "epoch": 0.1456071673594576, "grad_norm": 0.47096976381802785, "learning_rate": 1.4548387096774194e-05, "loss": 0.3347, "step": 451 }, { "epoch": 0.14593002138907946, "grad_norm": 0.7477546776411691, "learning_rate": 1.4580645161290324e-05, "loss": 0.3749, "step": 452 }, { "epoch": 0.14625287541870133, "grad_norm": 0.36778800692491925, "learning_rate": 1.4612903225806453e-05, "loss": 0.297, "step": 453 }, { "epoch": 0.14657572944832317, "grad_norm": 0.43258796764575197, "learning_rate": 1.464516129032258e-05, "loss": 0.3971, "step": 454 }, { "epoch": 0.14689858347794504, "grad_norm": 0.39805181870075046, "learning_rate": 1.467741935483871e-05, "loss": 0.4331, "step": 455 }, { "epoch": 0.1472214375075669, "grad_norm": 0.5166945731024414, "learning_rate": 1.4709677419354838e-05, "loss": 0.3638, "step": 456 }, { "epoch": 0.14754429153718876, "grad_norm": 0.418012138214176, "learning_rate": 1.4741935483870969e-05, "loss": 0.4043, "step": 457 }, { "epoch": 0.1478671455668106, "grad_norm": 0.3804529577391831, "learning_rate": 1.4774193548387098e-05, "loss": 0.3534, "step": 458 }, { "epoch": 0.14818999959643248, "grad_norm": 0.4134174783989682, "learning_rate": 1.4806451612903226e-05, "loss": 0.3631, "step": 459 }, { "epoch": 0.14851285362605432, "grad_norm": 0.38427831405534785, "learning_rate": 1.4838709677419355e-05, "loss": 0.4167, "step": 460 }, { "epoch": 0.14883570765567616, "grad_norm": 0.5955710129699301, "learning_rate": 1.4870967741935485e-05, "loss": 0.4049, "step": 461 }, { "epoch": 0.14915856168529804, "grad_norm": 0.4785338287232893, "learning_rate": 1.4903225806451612e-05, "loss": 0.3918, "step": 462 }, { "epoch": 0.14948141571491988, "grad_norm": 0.45796602903758127, "learning_rate": 1.4935483870967743e-05, "loss": 0.373, "step": 463 }, { "epoch": 0.14980426974454175, "grad_norm": 0.41995729570362195, "learning_rate": 1.4967741935483871e-05, "loss": 0.3557, "step": 464 }, { "epoch": 0.1501271237741636, "grad_norm": 0.48413222706419834, "learning_rate": 1.5e-05, "loss": 0.4426, "step": 465 }, { "epoch": 0.15044997780378547, "grad_norm": 0.4796825057822859, "learning_rate": 1.4999999524880164e-05, "loss": 0.3897, "step": 466 }, { "epoch": 0.1507728318334073, "grad_norm": 0.5394524406794519, "learning_rate": 1.4999998099520716e-05, "loss": 0.345, "step": 467 }, { "epoch": 0.15109568586302918, "grad_norm": 0.44492195668014767, "learning_rate": 1.4999995723921838e-05, "loss": 0.3205, "step": 468 }, { "epoch": 0.15141853989265103, "grad_norm": 0.3334199264559326, "learning_rate": 1.4999992398083828e-05, "loss": 0.3657, "step": 469 }, { "epoch": 0.1517413939222729, "grad_norm": 0.39212096678500113, "learning_rate": 1.4999988122007111e-05, "loss": 0.3261, "step": 470 }, { "epoch": 0.15206424795189474, "grad_norm": 0.43431662138779165, "learning_rate": 1.4999982895692227e-05, "loss": 0.3891, "step": 471 }, { "epoch": 0.15238710198151662, "grad_norm": 0.4334043518741978, "learning_rate": 1.4999976719139837e-05, "loss": 0.2909, "step": 472 }, { "epoch": 0.15270995601113846, "grad_norm": 0.43964400247526414, "learning_rate": 1.4999969592350725e-05, "loss": 0.3368, "step": 473 }, { "epoch": 0.15303281004076033, "grad_norm": 0.3691321259250556, "learning_rate": 1.4999961515325796e-05, "loss": 0.289, "step": 474 }, { "epoch": 0.15335566407038217, "grad_norm": 0.4801104448852452, "learning_rate": 1.499995248806607e-05, "loss": 0.3654, "step": 475 }, { "epoch": 0.15367851810000405, "grad_norm": 0.3843193294129677, "learning_rate": 1.499994251057269e-05, "loss": 0.3723, "step": 476 }, { "epoch": 0.1540013721296259, "grad_norm": 0.5057582588269296, "learning_rate": 1.4999931582846922e-05, "loss": 0.5278, "step": 477 }, { "epoch": 0.15432422615924776, "grad_norm": 0.39628041052082136, "learning_rate": 1.4999919704890153e-05, "loss": 0.3773, "step": 478 }, { "epoch": 0.1546470801888696, "grad_norm": 0.5885288074610662, "learning_rate": 1.4999906876703883e-05, "loss": 0.3684, "step": 479 }, { "epoch": 0.15496993421849145, "grad_norm": 0.45305955618617705, "learning_rate": 1.499989309828974e-05, "loss": 0.3409, "step": 480 }, { "epoch": 0.15529278824811332, "grad_norm": 0.412364386256826, "learning_rate": 1.4999878369649471e-05, "loss": 0.4074, "step": 481 }, { "epoch": 0.15561564227773517, "grad_norm": 0.5230252865757481, "learning_rate": 1.499986269078494e-05, "loss": 0.3513, "step": 482 }, { "epoch": 0.15593849630735704, "grad_norm": 0.4401971637657921, "learning_rate": 1.4999846061698131e-05, "loss": 0.3632, "step": 483 }, { "epoch": 0.15626135033697888, "grad_norm": 0.3686568441127252, "learning_rate": 1.4999828482391159e-05, "loss": 0.3691, "step": 484 }, { "epoch": 0.15658420436660075, "grad_norm": 0.47417490257058387, "learning_rate": 1.4999809952866244e-05, "loss": 0.4542, "step": 485 }, { "epoch": 0.1569070583962226, "grad_norm": 0.5175957687188939, "learning_rate": 1.4999790473125732e-05, "loss": 0.3945, "step": 486 }, { "epoch": 0.15722991242584447, "grad_norm": 0.4206796899529246, "learning_rate": 1.49997700431721e-05, "loss": 0.367, "step": 487 }, { "epoch": 0.15755276645546631, "grad_norm": 0.4240517815853305, "learning_rate": 1.499974866300793e-05, "loss": 0.3389, "step": 488 }, { "epoch": 0.1578756204850882, "grad_norm": 0.4692252426914756, "learning_rate": 1.4999726332635931e-05, "loss": 0.5203, "step": 489 }, { "epoch": 0.15819847451471003, "grad_norm": 0.4814370217358306, "learning_rate": 1.4999703052058934e-05, "loss": 0.4688, "step": 490 }, { "epoch": 0.1585213285443319, "grad_norm": 0.4506043510207524, "learning_rate": 1.499967882127989e-05, "loss": 0.4472, "step": 491 }, { "epoch": 0.15884418257395375, "grad_norm": 0.4357192547562614, "learning_rate": 1.4999653640301864e-05, "loss": 0.4148, "step": 492 }, { "epoch": 0.15916703660357562, "grad_norm": 0.35648220674118963, "learning_rate": 1.499962750912805e-05, "loss": 0.3558, "step": 493 }, { "epoch": 0.15948989063319746, "grad_norm": 0.41167303803827715, "learning_rate": 1.499960042776176e-05, "loss": 0.3608, "step": 494 }, { "epoch": 0.15981274466281933, "grad_norm": 0.4602682505796056, "learning_rate": 1.4999572396206422e-05, "loss": 0.4199, "step": 495 }, { "epoch": 0.16013559869244118, "grad_norm": 0.41647258136126764, "learning_rate": 1.499954341446559e-05, "loss": 0.3171, "step": 496 }, { "epoch": 0.16045845272206305, "grad_norm": 0.4759049266054579, "learning_rate": 1.4999513482542932e-05, "loss": 0.3734, "step": 497 }, { "epoch": 0.1607813067516849, "grad_norm": 0.6937432659175593, "learning_rate": 1.4999482600442246e-05, "loss": 0.3984, "step": 498 }, { "epoch": 0.16110416078130677, "grad_norm": 0.4196923320928656, "learning_rate": 1.499945076816744e-05, "loss": 0.2912, "step": 499 }, { "epoch": 0.1614270148109286, "grad_norm": 0.4453639925139063, "learning_rate": 1.4999417985722552e-05, "loss": 0.4005, "step": 500 }, { "epoch": 0.16174986884055045, "grad_norm": 0.41909087303414155, "learning_rate": 1.4999384253111729e-05, "loss": 0.3885, "step": 501 }, { "epoch": 0.16207272287017233, "grad_norm": 0.4576398699862815, "learning_rate": 1.499934957033925e-05, "loss": 0.3594, "step": 502 }, { "epoch": 0.16239557689979417, "grad_norm": 0.5115105676711947, "learning_rate": 1.4999313937409509e-05, "loss": 0.3818, "step": 503 }, { "epoch": 0.16271843092941604, "grad_norm": 0.4714992418841494, "learning_rate": 1.4999277354327017e-05, "loss": 0.3609, "step": 504 }, { "epoch": 0.16304128495903789, "grad_norm": 0.4632300709461329, "learning_rate": 1.4999239821096413e-05, "loss": 0.3308, "step": 505 }, { "epoch": 0.16336413898865976, "grad_norm": 0.4931956522762844, "learning_rate": 1.499920133772245e-05, "loss": 0.4226, "step": 506 }, { "epoch": 0.1636869930182816, "grad_norm": 0.5650437843681284, "learning_rate": 1.4999161904210004e-05, "loss": 0.364, "step": 507 }, { "epoch": 0.16400984704790347, "grad_norm": 1.1014348418690643, "learning_rate": 1.4999121520564072e-05, "loss": 0.3905, "step": 508 }, { "epoch": 0.16433270107752532, "grad_norm": 0.6765731456069007, "learning_rate": 1.4999080186789773e-05, "loss": 0.3783, "step": 509 }, { "epoch": 0.1646555551071472, "grad_norm": 0.6159572035687493, "learning_rate": 1.4999037902892337e-05, "loss": 0.4009, "step": 510 }, { "epoch": 0.16497840913676903, "grad_norm": 0.46681907338307815, "learning_rate": 1.4998994668877128e-05, "loss": 0.3936, "step": 511 }, { "epoch": 0.1653012631663909, "grad_norm": 0.44367666864778155, "learning_rate": 1.499895048474962e-05, "loss": 0.4129, "step": 512 }, { "epoch": 0.16562411719601275, "grad_norm": 0.37590717547310465, "learning_rate": 1.4998905350515412e-05, "loss": 0.3924, "step": 513 }, { "epoch": 0.16594697122563462, "grad_norm": 0.5061795498592584, "learning_rate": 1.4998859266180223e-05, "loss": 0.4136, "step": 514 }, { "epoch": 0.16626982525525646, "grad_norm": 0.4495726764787361, "learning_rate": 1.4998812231749892e-05, "loss": 0.3421, "step": 515 }, { "epoch": 0.16659267928487834, "grad_norm": 0.46005021707131705, "learning_rate": 1.4998764247230378e-05, "loss": 0.3685, "step": 516 }, { "epoch": 0.16691553331450018, "grad_norm": 1.2684378435938393, "learning_rate": 1.4998715312627759e-05, "loss": 0.3537, "step": 517 }, { "epoch": 0.16723838734412205, "grad_norm": 1.32902612437778, "learning_rate": 1.4998665427948237e-05, "loss": 0.4438, "step": 518 }, { "epoch": 0.1675612413737439, "grad_norm": 0.4942535689102828, "learning_rate": 1.4998614593198133e-05, "loss": 0.4503, "step": 519 }, { "epoch": 0.16788409540336574, "grad_norm": 0.42454281219298123, "learning_rate": 1.4998562808383884e-05, "loss": 0.3133, "step": 520 }, { "epoch": 0.1682069494329876, "grad_norm": 0.43779630269695946, "learning_rate": 1.4998510073512054e-05, "loss": 0.3528, "step": 521 }, { "epoch": 0.16852980346260946, "grad_norm": 0.4528969196927487, "learning_rate": 1.4998456388589324e-05, "loss": 0.3532, "step": 522 }, { "epoch": 0.16885265749223133, "grad_norm": 0.5228046117397752, "learning_rate": 1.4998401753622497e-05, "loss": 0.4155, "step": 523 }, { "epoch": 0.16917551152185317, "grad_norm": 0.4099242110744202, "learning_rate": 1.499834616861849e-05, "loss": 0.3355, "step": 524 }, { "epoch": 0.16949836555147504, "grad_norm": 0.5666517888139887, "learning_rate": 1.4998289633584353e-05, "loss": 0.3557, "step": 525 }, { "epoch": 0.1698212195810969, "grad_norm": 0.5168466761895454, "learning_rate": 1.4998232148527243e-05, "loss": 0.3447, "step": 526 }, { "epoch": 0.17014407361071876, "grad_norm": 0.41893076653520767, "learning_rate": 1.4998173713454446e-05, "loss": 0.3987, "step": 527 }, { "epoch": 0.1704669276403406, "grad_norm": 0.6260388561118742, "learning_rate": 1.4998114328373365e-05, "loss": 0.3735, "step": 528 }, { "epoch": 0.17078978166996248, "grad_norm": 0.5076123203328181, "learning_rate": 1.4998053993291525e-05, "loss": 0.3539, "step": 529 }, { "epoch": 0.17111263569958432, "grad_norm": 0.39117587379412483, "learning_rate": 1.499799270821657e-05, "loss": 0.3184, "step": 530 }, { "epoch": 0.1714354897292062, "grad_norm": 0.47621563545965134, "learning_rate": 1.4997930473156261e-05, "loss": 0.4284, "step": 531 }, { "epoch": 0.17175834375882804, "grad_norm": 0.9254189670711266, "learning_rate": 1.499786728811849e-05, "loss": 0.3957, "step": 532 }, { "epoch": 0.1720811977884499, "grad_norm": 0.5740556260310267, "learning_rate": 1.4997803153111258e-05, "loss": 0.4149, "step": 533 }, { "epoch": 0.17240405181807175, "grad_norm": 0.4289153113032179, "learning_rate": 1.499773806814269e-05, "loss": 0.4231, "step": 534 }, { "epoch": 0.17272690584769362, "grad_norm": 0.6059286005891557, "learning_rate": 1.4997672033221034e-05, "loss": 0.3367, "step": 535 }, { "epoch": 0.17304975987731547, "grad_norm": 0.571079978348043, "learning_rate": 1.4997605048354657e-05, "loss": 0.3423, "step": 536 }, { "epoch": 0.17337261390693734, "grad_norm": 0.4459085196221074, "learning_rate": 1.4997537113552046e-05, "loss": 0.4076, "step": 537 }, { "epoch": 0.17369546793655918, "grad_norm": 0.5413871531325136, "learning_rate": 1.4997468228821806e-05, "loss": 0.4166, "step": 538 }, { "epoch": 0.17401832196618103, "grad_norm": 0.3424021899254079, "learning_rate": 1.4997398394172665e-05, "loss": 0.3722, "step": 539 }, { "epoch": 0.1743411759958029, "grad_norm": 0.40101188152399, "learning_rate": 1.4997327609613475e-05, "loss": 0.4479, "step": 540 }, { "epoch": 0.17466403002542474, "grad_norm": 0.4613106451280924, "learning_rate": 1.4997255875153199e-05, "loss": 0.3791, "step": 541 }, { "epoch": 0.17498688405504662, "grad_norm": 0.49038614595307384, "learning_rate": 1.499718319080093e-05, "loss": 0.324, "step": 542 }, { "epoch": 0.17530973808466846, "grad_norm": 0.5618136451078143, "learning_rate": 1.4997109556565873e-05, "loss": 0.3745, "step": 543 }, { "epoch": 0.17563259211429033, "grad_norm": 1.2639120390176053, "learning_rate": 1.4997034972457361e-05, "loss": 0.3724, "step": 544 }, { "epoch": 0.17595544614391218, "grad_norm": 0.4287559703178082, "learning_rate": 1.499695943848484e-05, "loss": 0.3089, "step": 545 }, { "epoch": 0.17627830017353405, "grad_norm": 0.6411215824357789, "learning_rate": 1.4996882954657883e-05, "loss": 0.3623, "step": 546 }, { "epoch": 0.1766011542031559, "grad_norm": 0.3891185540142393, "learning_rate": 1.4996805520986182e-05, "loss": 0.3338, "step": 547 }, { "epoch": 0.17692400823277776, "grad_norm": 0.4210822988353088, "learning_rate": 1.4996727137479541e-05, "loss": 0.3396, "step": 548 }, { "epoch": 0.1772468622623996, "grad_norm": 0.6312223587874038, "learning_rate": 1.4996647804147896e-05, "loss": 0.3662, "step": 549 }, { "epoch": 0.17756971629202148, "grad_norm": 0.562243372904327, "learning_rate": 1.4996567521001298e-05, "loss": 0.487, "step": 550 }, { "epoch": 0.17789257032164332, "grad_norm": 0.49038317789528063, "learning_rate": 1.4996486288049921e-05, "loss": 0.4273, "step": 551 }, { "epoch": 0.1782154243512652, "grad_norm": 0.60369327772882, "learning_rate": 1.4996404105304053e-05, "loss": 0.3752, "step": 552 }, { "epoch": 0.17853827838088704, "grad_norm": 0.571848813154901, "learning_rate": 1.4996320972774106e-05, "loss": 0.4029, "step": 553 }, { "epoch": 0.1788611324105089, "grad_norm": 0.5152520416917985, "learning_rate": 1.4996236890470619e-05, "loss": 0.3663, "step": 554 }, { "epoch": 0.17918398644013075, "grad_norm": 0.5305079540181613, "learning_rate": 1.4996151858404239e-05, "loss": 0.4415, "step": 555 }, { "epoch": 0.17950684046975263, "grad_norm": 0.4263349595323182, "learning_rate": 1.499606587658574e-05, "loss": 0.333, "step": 556 }, { "epoch": 0.17982969449937447, "grad_norm": 0.7301336390780618, "learning_rate": 1.4995978945026018e-05, "loss": 0.3756, "step": 557 }, { "epoch": 0.18015254852899631, "grad_norm": 0.4733334177114753, "learning_rate": 1.4995891063736087e-05, "loss": 0.4184, "step": 558 }, { "epoch": 0.1804754025586182, "grad_norm": 0.5996986622062158, "learning_rate": 1.499580223272708e-05, "loss": 0.3668, "step": 559 }, { "epoch": 0.18079825658824003, "grad_norm": 0.9359263742909386, "learning_rate": 1.4995712452010251e-05, "loss": 0.3946, "step": 560 }, { "epoch": 0.1811211106178619, "grad_norm": 0.43185038960460836, "learning_rate": 1.499562172159698e-05, "loss": 0.3996, "step": 561 }, { "epoch": 0.18144396464748375, "grad_norm": 0.7369931437491288, "learning_rate": 1.4995530041498756e-05, "loss": 0.371, "step": 562 }, { "epoch": 0.18176681867710562, "grad_norm": 0.515084106310744, "learning_rate": 1.4995437411727198e-05, "loss": 0.392, "step": 563 }, { "epoch": 0.18208967270672746, "grad_norm": 0.4688328352285442, "learning_rate": 1.4995343832294043e-05, "loss": 0.3822, "step": 564 }, { "epoch": 0.18241252673634933, "grad_norm": 0.5218585089512039, "learning_rate": 1.4995249303211147e-05, "loss": 0.3107, "step": 565 }, { "epoch": 0.18273538076597118, "grad_norm": 0.572999895180753, "learning_rate": 1.4995153824490483e-05, "loss": 0.2864, "step": 566 }, { "epoch": 0.18305823479559305, "grad_norm": 0.47255289722021265, "learning_rate": 1.4995057396144152e-05, "loss": 0.34, "step": 567 }, { "epoch": 0.1833810888252149, "grad_norm": 0.578389433987258, "learning_rate": 1.4994960018184371e-05, "loss": 0.3555, "step": 568 }, { "epoch": 0.18370394285483677, "grad_norm": 0.5072023146655039, "learning_rate": 1.4994861690623476e-05, "loss": 0.3298, "step": 569 }, { "epoch": 0.1840267968844586, "grad_norm": 0.49295720419251954, "learning_rate": 1.4994762413473926e-05, "loss": 0.3649, "step": 570 }, { "epoch": 0.18434965091408048, "grad_norm": 0.45651848380435983, "learning_rate": 1.49946621867483e-05, "loss": 0.3146, "step": 571 }, { "epoch": 0.18467250494370233, "grad_norm": 0.5301410866368356, "learning_rate": 1.4994561010459295e-05, "loss": 0.4015, "step": 572 }, { "epoch": 0.1849953589733242, "grad_norm": 0.6263860472308129, "learning_rate": 1.4994458884619731e-05, "loss": 0.3689, "step": 573 }, { "epoch": 0.18531821300294604, "grad_norm": 0.8229529994175983, "learning_rate": 1.4994355809242547e-05, "loss": 0.328, "step": 574 }, { "epoch": 0.1856410670325679, "grad_norm": 0.5039132494664615, "learning_rate": 1.49942517843408e-05, "loss": 0.4519, "step": 575 }, { "epoch": 0.18596392106218976, "grad_norm": 0.4633405813346391, "learning_rate": 1.4994146809927676e-05, "loss": 0.4506, "step": 576 }, { "epoch": 0.1862867750918116, "grad_norm": 0.6574791754683984, "learning_rate": 1.4994040886016467e-05, "loss": 0.3838, "step": 577 }, { "epoch": 0.18660962912143347, "grad_norm": 0.5264832439752606, "learning_rate": 1.49939340126206e-05, "loss": 0.3754, "step": 578 }, { "epoch": 0.18693248315105532, "grad_norm": 0.5724156006445754, "learning_rate": 1.4993826189753613e-05, "loss": 0.3732, "step": 579 }, { "epoch": 0.1872553371806772, "grad_norm": 0.8068383693976912, "learning_rate": 1.4993717417429168e-05, "loss": 0.3789, "step": 580 }, { "epoch": 0.18757819121029903, "grad_norm": 0.43022334506337334, "learning_rate": 1.4993607695661045e-05, "loss": 0.3641, "step": 581 }, { "epoch": 0.1879010452399209, "grad_norm": 0.3370469748303194, "learning_rate": 1.4993497024463148e-05, "loss": 0.3261, "step": 582 }, { "epoch": 0.18822389926954275, "grad_norm": 0.4004127094104992, "learning_rate": 1.4993385403849495e-05, "loss": 0.3395, "step": 583 }, { "epoch": 0.18854675329916462, "grad_norm": 0.45293342936762987, "learning_rate": 1.499327283383423e-05, "loss": 0.4149, "step": 584 }, { "epoch": 0.18886960732878647, "grad_norm": 0.5615227731970244, "learning_rate": 1.4993159314431619e-05, "loss": 0.3289, "step": 585 }, { "epoch": 0.18919246135840834, "grad_norm": 0.4751248747949785, "learning_rate": 1.4993044845656037e-05, "loss": 0.4031, "step": 586 }, { "epoch": 0.18951531538803018, "grad_norm": 0.650796938860345, "learning_rate": 1.4992929427521995e-05, "loss": 0.3768, "step": 587 }, { "epoch": 0.18983816941765205, "grad_norm": 0.5287490862592497, "learning_rate": 1.4992813060044111e-05, "loss": 0.4172, "step": 588 }, { "epoch": 0.1901610234472739, "grad_norm": 0.482088888036767, "learning_rate": 1.4992695743237133e-05, "loss": 0.2864, "step": 589 }, { "epoch": 0.19048387747689577, "grad_norm": 0.44916742681144745, "learning_rate": 1.4992577477115919e-05, "loss": 0.3689, "step": 590 }, { "epoch": 0.1908067315065176, "grad_norm": 1.4090509289772852, "learning_rate": 1.4992458261695458e-05, "loss": 0.4008, "step": 591 }, { "epoch": 0.19112958553613948, "grad_norm": 0.5382138884442168, "learning_rate": 1.4992338096990854e-05, "loss": 0.4135, "step": 592 }, { "epoch": 0.19145243956576133, "grad_norm": 0.5434763799138503, "learning_rate": 1.499221698301733e-05, "loss": 0.3399, "step": 593 }, { "epoch": 0.1917752935953832, "grad_norm": 0.43278488010544336, "learning_rate": 1.4992094919790232e-05, "loss": 0.3223, "step": 594 }, { "epoch": 0.19209814762500504, "grad_norm": 0.5526920724057823, "learning_rate": 1.4991971907325026e-05, "loss": 0.3875, "step": 595 }, { "epoch": 0.1924210016546269, "grad_norm": 0.3944530374663526, "learning_rate": 1.4991847945637294e-05, "loss": 0.3252, "step": 596 }, { "epoch": 0.19274385568424876, "grad_norm": 0.45061349329533895, "learning_rate": 1.4991723034742744e-05, "loss": 0.3668, "step": 597 }, { "epoch": 0.1930667097138706, "grad_norm": 0.368374158104004, "learning_rate": 1.4991597174657202e-05, "loss": 0.3367, "step": 598 }, { "epoch": 0.19338956374349248, "grad_norm": 0.48902109429685453, "learning_rate": 1.4991470365396617e-05, "loss": 0.4025, "step": 599 }, { "epoch": 0.19371241777311432, "grad_norm": 0.4434245774670545, "learning_rate": 1.4991342606977052e-05, "loss": 0.4326, "step": 600 }, { "epoch": 0.1940352718027362, "grad_norm": 0.4637716141728241, "learning_rate": 1.4991213899414695e-05, "loss": 0.3912, "step": 601 }, { "epoch": 0.19435812583235804, "grad_norm": 0.42493660231923713, "learning_rate": 1.4991084242725854e-05, "loss": 0.3381, "step": 602 }, { "epoch": 0.1946809798619799, "grad_norm": 0.4490497964857586, "learning_rate": 1.4990953636926952e-05, "loss": 0.3998, "step": 603 }, { "epoch": 0.19500383389160175, "grad_norm": 0.4183725658779027, "learning_rate": 1.499082208203454e-05, "loss": 0.3602, "step": 604 }, { "epoch": 0.19532668792122362, "grad_norm": 0.44333441210235636, "learning_rate": 1.4990689578065289e-05, "loss": 0.3298, "step": 605 }, { "epoch": 0.19564954195084547, "grad_norm": 0.4132614732705298, "learning_rate": 1.4990556125035983e-05, "loss": 0.3381, "step": 606 }, { "epoch": 0.19597239598046734, "grad_norm": 0.3959375551576009, "learning_rate": 1.499042172296353e-05, "loss": 0.3482, "step": 607 }, { "epoch": 0.19629525001008918, "grad_norm": 1.0818240310389917, "learning_rate": 1.499028637186496e-05, "loss": 0.4102, "step": 608 }, { "epoch": 0.19661810403971106, "grad_norm": 0.5306526785207776, "learning_rate": 1.4990150071757422e-05, "loss": 0.3565, "step": 609 }, { "epoch": 0.1969409580693329, "grad_norm": 0.4879223077490384, "learning_rate": 1.4990012822658183e-05, "loss": 0.3925, "step": 610 }, { "epoch": 0.19726381209895477, "grad_norm": 0.5531418287535899, "learning_rate": 1.4989874624584635e-05, "loss": 0.364, "step": 611 }, { "epoch": 0.19758666612857662, "grad_norm": 0.5215901877659358, "learning_rate": 1.4989735477554286e-05, "loss": 0.4053, "step": 612 }, { "epoch": 0.1979095201581985, "grad_norm": 0.3791351641525893, "learning_rate": 1.4989595381584766e-05, "loss": 0.3586, "step": 613 }, { "epoch": 0.19823237418782033, "grad_norm": 0.49056819060544005, "learning_rate": 1.4989454336693825e-05, "loss": 0.4219, "step": 614 }, { "epoch": 0.19855522821744218, "grad_norm": 0.436863788159559, "learning_rate": 1.4989312342899332e-05, "loss": 0.3408, "step": 615 }, { "epoch": 0.19887808224706405, "grad_norm": 0.5548833163473572, "learning_rate": 1.498916940021928e-05, "loss": 0.4128, "step": 616 }, { "epoch": 0.1992009362766859, "grad_norm": 0.4458795024996622, "learning_rate": 1.4989025508671776e-05, "loss": 0.3467, "step": 617 }, { "epoch": 0.19952379030630776, "grad_norm": 0.43107558984219907, "learning_rate": 1.4988880668275057e-05, "loss": 0.3256, "step": 618 }, { "epoch": 0.1998466443359296, "grad_norm": 0.5045949378989837, "learning_rate": 1.4988734879047466e-05, "loss": 0.4076, "step": 619 }, { "epoch": 0.20016949836555148, "grad_norm": 0.38600929871177553, "learning_rate": 1.498858814100748e-05, "loss": 0.4175, "step": 620 }, { "epoch": 0.20049235239517332, "grad_norm": 0.48143743181710985, "learning_rate": 1.498844045417369e-05, "loss": 0.4025, "step": 621 }, { "epoch": 0.2008152064247952, "grad_norm": 0.3990154161571044, "learning_rate": 1.4988291818564806e-05, "loss": 0.3891, "step": 622 }, { "epoch": 0.20113806045441704, "grad_norm": 0.46587049162815747, "learning_rate": 1.498814223419966e-05, "loss": 0.4107, "step": 623 }, { "epoch": 0.2014609144840389, "grad_norm": 0.4652733222232367, "learning_rate": 1.4987991701097205e-05, "loss": 0.4116, "step": 624 }, { "epoch": 0.20178376851366076, "grad_norm": 0.4000257164764127, "learning_rate": 1.4987840219276513e-05, "loss": 0.3153, "step": 625 }, { "epoch": 0.20210662254328263, "grad_norm": 0.46627790789895174, "learning_rate": 1.4987687788756777e-05, "loss": 0.4008, "step": 626 }, { "epoch": 0.20242947657290447, "grad_norm": 0.427536836499093, "learning_rate": 1.498753440955731e-05, "loss": 0.3416, "step": 627 }, { "epoch": 0.20275233060252634, "grad_norm": 0.40991098260350506, "learning_rate": 1.4987380081697544e-05, "loss": 0.3894, "step": 628 }, { "epoch": 0.2030751846321482, "grad_norm": 0.48410917539569165, "learning_rate": 1.498722480519703e-05, "loss": 0.3874, "step": 629 }, { "epoch": 0.20339803866177006, "grad_norm": 0.3878976860897732, "learning_rate": 1.4987068580075445e-05, "loss": 0.3148, "step": 630 }, { "epoch": 0.2037208926913919, "grad_norm": 0.45702398729174964, "learning_rate": 1.4986911406352583e-05, "loss": 0.3808, "step": 631 }, { "epoch": 0.20404374672101377, "grad_norm": 0.5283429722606366, "learning_rate": 1.4986753284048353e-05, "loss": 0.4255, "step": 632 }, { "epoch": 0.20436660075063562, "grad_norm": 0.4176008965039356, "learning_rate": 1.4986594213182794e-05, "loss": 0.3526, "step": 633 }, { "epoch": 0.20468945478025746, "grad_norm": 0.3726930614214516, "learning_rate": 1.4986434193776058e-05, "loss": 0.3609, "step": 634 }, { "epoch": 0.20501230880987933, "grad_norm": 0.48576348063694746, "learning_rate": 1.4986273225848418e-05, "loss": 0.3953, "step": 635 }, { "epoch": 0.20533516283950118, "grad_norm": 0.554494650978848, "learning_rate": 1.4986111309420269e-05, "loss": 0.3924, "step": 636 }, { "epoch": 0.20565801686912305, "grad_norm": 0.48397024924757265, "learning_rate": 1.4985948444512126e-05, "loss": 0.3799, "step": 637 }, { "epoch": 0.2059808708987449, "grad_norm": 0.43718114808723973, "learning_rate": 1.4985784631144626e-05, "loss": 0.4544, "step": 638 }, { "epoch": 0.20630372492836677, "grad_norm": 0.42915843499116907, "learning_rate": 1.4985619869338521e-05, "loss": 0.4322, "step": 639 }, { "epoch": 0.2066265789579886, "grad_norm": 0.5100834050537769, "learning_rate": 1.4985454159114687e-05, "loss": 0.3485, "step": 640 }, { "epoch": 0.20694943298761048, "grad_norm": 0.4582148532072101, "learning_rate": 1.4985287500494118e-05, "loss": 0.3918, "step": 641 }, { "epoch": 0.20727228701723233, "grad_norm": 0.36853654051653234, "learning_rate": 1.498511989349793e-05, "loss": 0.3449, "step": 642 }, { "epoch": 0.2075951410468542, "grad_norm": 0.44544519402635874, "learning_rate": 1.498495133814736e-05, "loss": 0.3677, "step": 643 }, { "epoch": 0.20791799507647604, "grad_norm": 0.4645764853307449, "learning_rate": 1.4984781834463762e-05, "loss": 0.3634, "step": 644 }, { "epoch": 0.20824084910609791, "grad_norm": 0.5090186077145763, "learning_rate": 1.4984611382468614e-05, "loss": 0.3597, "step": 645 }, { "epoch": 0.20856370313571976, "grad_norm": 0.8942141296362047, "learning_rate": 1.4984439982183512e-05, "loss": 0.3384, "step": 646 }, { "epoch": 0.20888655716534163, "grad_norm": 0.4126669979585592, "learning_rate": 1.498426763363017e-05, "loss": 0.3298, "step": 647 }, { "epoch": 0.20920941119496347, "grad_norm": 0.4466186205012202, "learning_rate": 1.4984094336830424e-05, "loss": 0.3852, "step": 648 }, { "epoch": 0.20953226522458535, "grad_norm": 0.41398866650152916, "learning_rate": 1.4983920091806232e-05, "loss": 0.3661, "step": 649 }, { "epoch": 0.2098551192542072, "grad_norm": 0.44050067218720734, "learning_rate": 1.4983744898579671e-05, "loss": 0.3836, "step": 650 }, { "epoch": 0.21017797328382906, "grad_norm": 0.45275073404409283, "learning_rate": 1.4983568757172937e-05, "loss": 0.3535, "step": 651 }, { "epoch": 0.2105008273134509, "grad_norm": 0.4446315864876987, "learning_rate": 1.4983391667608345e-05, "loss": 0.4614, "step": 652 }, { "epoch": 0.21082368134307275, "grad_norm": 0.44128319691267737, "learning_rate": 1.4983213629908338e-05, "loss": 0.4195, "step": 653 }, { "epoch": 0.21114653537269462, "grad_norm": 0.4777387006073791, "learning_rate": 1.4983034644095464e-05, "loss": 0.3675, "step": 654 }, { "epoch": 0.21146938940231647, "grad_norm": 0.42032735329849075, "learning_rate": 1.498285471019241e-05, "loss": 0.4255, "step": 655 }, { "epoch": 0.21179224343193834, "grad_norm": 0.4169809948160134, "learning_rate": 1.4982673828221966e-05, "loss": 0.4275, "step": 656 }, { "epoch": 0.21211509746156018, "grad_norm": 0.47390063600495475, "learning_rate": 1.4982491998207054e-05, "loss": 0.4296, "step": 657 }, { "epoch": 0.21243795149118205, "grad_norm": 0.5412952479144771, "learning_rate": 1.4982309220170707e-05, "loss": 0.3503, "step": 658 }, { "epoch": 0.2127608055208039, "grad_norm": 0.4141120405984391, "learning_rate": 1.4982125494136089e-05, "loss": 0.4174, "step": 659 }, { "epoch": 0.21308365955042577, "grad_norm": 0.39984278401393614, "learning_rate": 1.4981940820126474e-05, "loss": 0.3711, "step": 660 }, { "epoch": 0.2134065135800476, "grad_norm": 0.5402277335741801, "learning_rate": 1.4981755198165258e-05, "loss": 0.3731, "step": 661 }, { "epoch": 0.21372936760966948, "grad_norm": 0.504969785148045, "learning_rate": 1.4981568628275963e-05, "loss": 0.381, "step": 662 }, { "epoch": 0.21405222163929133, "grad_norm": 0.5759703470157859, "learning_rate": 1.4981381110482226e-05, "loss": 0.4041, "step": 663 }, { "epoch": 0.2143750756689132, "grad_norm": 0.40554604393716065, "learning_rate": 1.4981192644807802e-05, "loss": 0.3362, "step": 664 }, { "epoch": 0.21469792969853504, "grad_norm": 0.5672965179038282, "learning_rate": 1.4981003231276576e-05, "loss": 0.3401, "step": 665 }, { "epoch": 0.21502078372815692, "grad_norm": 0.5356063425622674, "learning_rate": 1.498081286991254e-05, "loss": 0.3258, "step": 666 }, { "epoch": 0.21534363775777876, "grad_norm": 0.6368766080503767, "learning_rate": 1.4980621560739815e-05, "loss": 0.3309, "step": 667 }, { "epoch": 0.21566649178740063, "grad_norm": 0.4192929281527647, "learning_rate": 1.4980429303782641e-05, "loss": 0.4707, "step": 668 }, { "epoch": 0.21598934581702248, "grad_norm": 0.4125573764759406, "learning_rate": 1.4980236099065375e-05, "loss": 0.3422, "step": 669 }, { "epoch": 0.21631219984664435, "grad_norm": 0.5682828104460262, "learning_rate": 1.4980041946612496e-05, "loss": 0.3601, "step": 670 }, { "epoch": 0.2166350538762662, "grad_norm": 0.40049244098100045, "learning_rate": 1.4979846846448602e-05, "loss": 0.3461, "step": 671 }, { "epoch": 0.21695790790588804, "grad_norm": 0.4177096149598353, "learning_rate": 1.4979650798598414e-05, "loss": 0.3667, "step": 672 }, { "epoch": 0.2172807619355099, "grad_norm": 0.4401058426407057, "learning_rate": 1.497945380308677e-05, "loss": 0.3874, "step": 673 }, { "epoch": 0.21760361596513175, "grad_norm": 0.37529071786418966, "learning_rate": 1.4979255859938629e-05, "loss": 0.3643, "step": 674 }, { "epoch": 0.21792646999475362, "grad_norm": 0.4872001732118169, "learning_rate": 1.497905696917907e-05, "loss": 0.3902, "step": 675 }, { "epoch": 0.21824932402437547, "grad_norm": 0.46358002846014373, "learning_rate": 1.4978857130833292e-05, "loss": 0.4221, "step": 676 }, { "epoch": 0.21857217805399734, "grad_norm": 0.42253148440289606, "learning_rate": 1.4978656344926615e-05, "loss": 0.3544, "step": 677 }, { "epoch": 0.21889503208361918, "grad_norm": 0.5723460491700638, "learning_rate": 1.4978454611484477e-05, "loss": 0.4175, "step": 678 }, { "epoch": 0.21921788611324106, "grad_norm": 0.6627741505994543, "learning_rate": 1.4978251930532438e-05, "loss": 0.3482, "step": 679 }, { "epoch": 0.2195407401428629, "grad_norm": 0.4299963503830465, "learning_rate": 1.4978048302096179e-05, "loss": 0.414, "step": 680 }, { "epoch": 0.21986359417248477, "grad_norm": 0.47655207560967794, "learning_rate": 1.4977843726201497e-05, "loss": 0.3991, "step": 681 }, { "epoch": 0.22018644820210662, "grad_norm": 0.5503486885303833, "learning_rate": 1.497763820287431e-05, "loss": 0.369, "step": 682 }, { "epoch": 0.2205093022317285, "grad_norm": 0.40115679497902995, "learning_rate": 1.4977431732140665e-05, "loss": 0.355, "step": 683 }, { "epoch": 0.22083215626135033, "grad_norm": 0.5176220881253282, "learning_rate": 1.4977224314026713e-05, "loss": 0.379, "step": 684 }, { "epoch": 0.2211550102909722, "grad_norm": 0.43797520432798515, "learning_rate": 1.497701594855874e-05, "loss": 0.3757, "step": 685 }, { "epoch": 0.22147786432059405, "grad_norm": 0.489248799364668, "learning_rate": 1.497680663576314e-05, "loss": 0.2939, "step": 686 }, { "epoch": 0.22180071835021592, "grad_norm": 0.5940425350921853, "learning_rate": 1.4976596375666436e-05, "loss": 0.354, "step": 687 }, { "epoch": 0.22212357237983776, "grad_norm": 0.4030884298737031, "learning_rate": 1.4976385168295269e-05, "loss": 0.3929, "step": 688 }, { "epoch": 0.22244642640945964, "grad_norm": 0.5141711179956666, "learning_rate": 1.4976173013676396e-05, "loss": 0.3812, "step": 689 }, { "epoch": 0.22276928043908148, "grad_norm": 0.4913415852184453, "learning_rate": 1.4975959911836695e-05, "loss": 0.3649, "step": 690 }, { "epoch": 0.22309213446870332, "grad_norm": 0.5156492221645427, "learning_rate": 1.497574586280317e-05, "loss": 0.3718, "step": 691 }, { "epoch": 0.2234149884983252, "grad_norm": 0.48337230570599615, "learning_rate": 1.497553086660294e-05, "loss": 0.3655, "step": 692 }, { "epoch": 0.22373784252794704, "grad_norm": 0.43649968286916146, "learning_rate": 1.4975314923263243e-05, "loss": 0.3593, "step": 693 }, { "epoch": 0.2240606965575689, "grad_norm": 0.404795550863597, "learning_rate": 1.4975098032811439e-05, "loss": 0.4061, "step": 694 }, { "epoch": 0.22438355058719076, "grad_norm": 0.4919213843238154, "learning_rate": 1.4974880195275008e-05, "loss": 0.3291, "step": 695 }, { "epoch": 0.22470640461681263, "grad_norm": 0.4248796185427426, "learning_rate": 1.4974661410681548e-05, "loss": 0.3652, "step": 696 }, { "epoch": 0.22502925864643447, "grad_norm": 0.473974484504545, "learning_rate": 1.4974441679058783e-05, "loss": 0.4008, "step": 697 }, { "epoch": 0.22535211267605634, "grad_norm": 0.4936616302718047, "learning_rate": 1.4974221000434549e-05, "loss": 0.3704, "step": 698 }, { "epoch": 0.2256749667056782, "grad_norm": 0.46486338582035447, "learning_rate": 1.4973999374836809e-05, "loss": 0.351, "step": 699 }, { "epoch": 0.22599782073530006, "grad_norm": 0.38935056909818844, "learning_rate": 1.4973776802293638e-05, "loss": 0.3364, "step": 700 }, { "epoch": 0.2263206747649219, "grad_norm": 0.49830465536171714, "learning_rate": 1.4973553282833239e-05, "loss": 0.3536, "step": 701 }, { "epoch": 0.22664352879454377, "grad_norm": 0.3724146947755354, "learning_rate": 1.4973328816483931e-05, "loss": 0.3829, "step": 702 }, { "epoch": 0.22696638282416562, "grad_norm": 0.44330719573565047, "learning_rate": 1.4973103403274154e-05, "loss": 0.3964, "step": 703 }, { "epoch": 0.2272892368537875, "grad_norm": 0.3743859721090949, "learning_rate": 1.4972877043232466e-05, "loss": 0.3192, "step": 704 }, { "epoch": 0.22761209088340933, "grad_norm": 0.48082638152884044, "learning_rate": 1.4972649736387546e-05, "loss": 0.3533, "step": 705 }, { "epoch": 0.2279349449130312, "grad_norm": 0.4400110748545771, "learning_rate": 1.4972421482768197e-05, "loss": 0.343, "step": 706 }, { "epoch": 0.22825779894265305, "grad_norm": 0.5386677215493516, "learning_rate": 1.4972192282403336e-05, "loss": 0.3616, "step": 707 }, { "epoch": 0.22858065297227492, "grad_norm": 0.46838117075710417, "learning_rate": 1.4971962135322e-05, "loss": 0.3447, "step": 708 }, { "epoch": 0.22890350700189677, "grad_norm": 0.4023448700450532, "learning_rate": 1.4971731041553353e-05, "loss": 0.3364, "step": 709 }, { "epoch": 0.22922636103151864, "grad_norm": 0.5031792562039873, "learning_rate": 1.4971499001126674e-05, "loss": 0.367, "step": 710 }, { "epoch": 0.22954921506114048, "grad_norm": 0.5120133365506145, "learning_rate": 1.4971266014071357e-05, "loss": 0.4026, "step": 711 }, { "epoch": 0.22987206909076233, "grad_norm": 0.49332858720009853, "learning_rate": 1.4971032080416929e-05, "loss": 0.3456, "step": 712 }, { "epoch": 0.2301949231203842, "grad_norm": 0.524697500919251, "learning_rate": 1.4970797200193021e-05, "loss": 0.3563, "step": 713 }, { "epoch": 0.23051777715000604, "grad_norm": 0.40745806382591965, "learning_rate": 1.4970561373429396e-05, "loss": 0.3812, "step": 714 }, { "epoch": 0.23084063117962791, "grad_norm": 0.5566176862425689, "learning_rate": 1.4970324600155935e-05, "loss": 0.4198, "step": 715 }, { "epoch": 0.23116348520924976, "grad_norm": 1.331886598148543, "learning_rate": 1.4970086880402631e-05, "loss": 0.4014, "step": 716 }, { "epoch": 0.23148633923887163, "grad_norm": 0.5008450170716275, "learning_rate": 1.496984821419961e-05, "loss": 0.3377, "step": 717 }, { "epoch": 0.23180919326849347, "grad_norm": 0.5168332734250204, "learning_rate": 1.4969608601577102e-05, "loss": 0.4035, "step": 718 }, { "epoch": 0.23213204729811535, "grad_norm": 0.4216521358187965, "learning_rate": 1.4969368042565474e-05, "loss": 0.3348, "step": 719 }, { "epoch": 0.2324549013277372, "grad_norm": 0.4052487109561257, "learning_rate": 1.4969126537195198e-05, "loss": 0.4046, "step": 720 }, { "epoch": 0.23277775535735906, "grad_norm": 0.4351157965032379, "learning_rate": 1.4968884085496878e-05, "loss": 0.3483, "step": 721 }, { "epoch": 0.2331006093869809, "grad_norm": 0.4102192907431703, "learning_rate": 1.496864068750123e-05, "loss": 0.368, "step": 722 }, { "epoch": 0.23342346341660278, "grad_norm": 0.4696098904344873, "learning_rate": 1.496839634323909e-05, "loss": 0.3753, "step": 723 }, { "epoch": 0.23374631744622462, "grad_norm": 0.4924112015339799, "learning_rate": 1.4968151052741419e-05, "loss": 0.3745, "step": 724 }, { "epoch": 0.2340691714758465, "grad_norm": 0.48331054692760295, "learning_rate": 1.4967904816039294e-05, "loss": 0.425, "step": 725 }, { "epoch": 0.23439202550546834, "grad_norm": 0.4041977331566458, "learning_rate": 1.4967657633163913e-05, "loss": 0.3694, "step": 726 }, { "epoch": 0.2347148795350902, "grad_norm": 0.5025993513944014, "learning_rate": 1.4967409504146593e-05, "loss": 0.3376, "step": 727 }, { "epoch": 0.23503773356471205, "grad_norm": 0.622997709876866, "learning_rate": 1.4967160429018771e-05, "loss": 0.3399, "step": 728 }, { "epoch": 0.23536058759433393, "grad_norm": 0.5887355585079421, "learning_rate": 1.4966910407812006e-05, "loss": 0.4142, "step": 729 }, { "epoch": 0.23568344162395577, "grad_norm": 0.4906717351953235, "learning_rate": 1.4966659440557975e-05, "loss": 0.3924, "step": 730 }, { "epoch": 0.2360062956535776, "grad_norm": 0.3668238290390886, "learning_rate": 1.4966407527288476e-05, "loss": 0.3806, "step": 731 }, { "epoch": 0.23632914968319949, "grad_norm": 0.48483325933046933, "learning_rate": 1.4966154668035424e-05, "loss": 0.3393, "step": 732 }, { "epoch": 0.23665200371282133, "grad_norm": 0.45471982664219474, "learning_rate": 1.4965900862830857e-05, "loss": 0.4423, "step": 733 }, { "epoch": 0.2369748577424432, "grad_norm": 0.4348299484594779, "learning_rate": 1.4965646111706932e-05, "loss": 0.3609, "step": 734 }, { "epoch": 0.23729771177206505, "grad_norm": 0.535680495670231, "learning_rate": 1.4965390414695927e-05, "loss": 0.3914, "step": 735 }, { "epoch": 0.23762056580168692, "grad_norm": 0.8444083133497843, "learning_rate": 1.4965133771830235e-05, "loss": 0.3089, "step": 736 }, { "epoch": 0.23794341983130876, "grad_norm": 0.40763668028225447, "learning_rate": 1.4964876183142375e-05, "loss": 0.3335, "step": 737 }, { "epoch": 0.23826627386093063, "grad_norm": 0.40797928160577773, "learning_rate": 1.4964617648664982e-05, "loss": 0.4201, "step": 738 }, { "epoch": 0.23858912789055248, "grad_norm": 0.47789640443644166, "learning_rate": 1.4964358168430813e-05, "loss": 0.4073, "step": 739 }, { "epoch": 0.23891198192017435, "grad_norm": 0.40198570032394493, "learning_rate": 1.4964097742472744e-05, "loss": 0.3749, "step": 740 }, { "epoch": 0.2392348359497962, "grad_norm": 0.37932767063426504, "learning_rate": 1.4963836370823768e-05, "loss": 0.3385, "step": 741 }, { "epoch": 0.23955768997941806, "grad_norm": 0.5813351437570612, "learning_rate": 1.4963574053517001e-05, "loss": 0.3474, "step": 742 }, { "epoch": 0.2398805440090399, "grad_norm": 0.39754022678723844, "learning_rate": 1.4963310790585681e-05, "loss": 0.3077, "step": 743 }, { "epoch": 0.24020339803866178, "grad_norm": 0.45445159876367736, "learning_rate": 1.4963046582063162e-05, "loss": 0.4103, "step": 744 }, { "epoch": 0.24052625206828362, "grad_norm": 0.3929914817197378, "learning_rate": 1.4962781427982917e-05, "loss": 0.4, "step": 745 }, { "epoch": 0.2408491060979055, "grad_norm": 0.47103059795638313, "learning_rate": 1.4962515328378543e-05, "loss": 0.3907, "step": 746 }, { "epoch": 0.24117196012752734, "grad_norm": 0.41905145069945277, "learning_rate": 1.4962248283283754e-05, "loss": 0.4004, "step": 747 }, { "epoch": 0.2414948141571492, "grad_norm": 0.4246738033103205, "learning_rate": 1.496198029273238e-05, "loss": 0.3411, "step": 748 }, { "epoch": 0.24181766818677106, "grad_norm": 0.4153088114258945, "learning_rate": 1.4961711356758383e-05, "loss": 0.3723, "step": 749 }, { "epoch": 0.2421405222163929, "grad_norm": 0.5380856367716079, "learning_rate": 1.496144147539583e-05, "loss": 0.3776, "step": 750 }, { "epoch": 0.24246337624601477, "grad_norm": 0.549553038998934, "learning_rate": 1.4961170648678918e-05, "loss": 0.3597, "step": 751 }, { "epoch": 0.24278623027563662, "grad_norm": 0.405754080177905, "learning_rate": 1.4960898876641959e-05, "loss": 0.4179, "step": 752 }, { "epoch": 0.2431090843052585, "grad_norm": 0.5503255787975231, "learning_rate": 1.4960626159319386e-05, "loss": 0.3805, "step": 753 }, { "epoch": 0.24343193833488033, "grad_norm": 0.4615143301528694, "learning_rate": 1.4960352496745754e-05, "loss": 0.3505, "step": 754 }, { "epoch": 0.2437547923645022, "grad_norm": 0.4938149382376007, "learning_rate": 1.4960077888955733e-05, "loss": 0.4122, "step": 755 }, { "epoch": 0.24407764639412405, "grad_norm": 0.4426281850799338, "learning_rate": 1.4959802335984118e-05, "loss": 0.3731, "step": 756 }, { "epoch": 0.24440050042374592, "grad_norm": 0.4166834448614979, "learning_rate": 1.4959525837865818e-05, "loss": 0.3614, "step": 757 }, { "epoch": 0.24472335445336776, "grad_norm": 0.4350627683340346, "learning_rate": 1.4959248394635867e-05, "loss": 0.4132, "step": 758 }, { "epoch": 0.24504620848298964, "grad_norm": 0.6883642254673134, "learning_rate": 1.4958970006329418e-05, "loss": 0.3509, "step": 759 }, { "epoch": 0.24536906251261148, "grad_norm": 0.451646075954908, "learning_rate": 1.495869067298174e-05, "loss": 0.4053, "step": 760 }, { "epoch": 0.24569191654223335, "grad_norm": 0.4351332892965326, "learning_rate": 1.4958410394628227e-05, "loss": 0.3083, "step": 761 }, { "epoch": 0.2460147705718552, "grad_norm": 0.3957583962928366, "learning_rate": 1.4958129171304385e-05, "loss": 0.4182, "step": 762 }, { "epoch": 0.24633762460147707, "grad_norm": 0.6412153270982703, "learning_rate": 1.4957847003045845e-05, "loss": 0.3146, "step": 763 }, { "epoch": 0.2466604786310989, "grad_norm": 0.4616896443546052, "learning_rate": 1.4957563889888365e-05, "loss": 0.3716, "step": 764 }, { "epoch": 0.24698333266072078, "grad_norm": 0.43231834013779374, "learning_rate": 1.4957279831867807e-05, "loss": 0.298, "step": 765 }, { "epoch": 0.24730618669034263, "grad_norm": 0.4463555695674332, "learning_rate": 1.4956994829020164e-05, "loss": 0.3933, "step": 766 }, { "epoch": 0.2476290407199645, "grad_norm": 0.4792685282998694, "learning_rate": 1.4956708881381547e-05, "loss": 0.3356, "step": 767 }, { "epoch": 0.24795189474958634, "grad_norm": 0.45240337821214605, "learning_rate": 1.4956421988988182e-05, "loss": 0.4079, "step": 768 }, { "epoch": 0.2482747487792082, "grad_norm": 0.5350648009542132, "learning_rate": 1.4956134151876418e-05, "loss": 0.3566, "step": 769 }, { "epoch": 0.24859760280883006, "grad_norm": 0.5032583937659203, "learning_rate": 1.4955845370082725e-05, "loss": 0.4157, "step": 770 }, { "epoch": 0.2489204568384519, "grad_norm": 0.6089552736301479, "learning_rate": 1.495555564364369e-05, "loss": 0.4044, "step": 771 }, { "epoch": 0.24924331086807378, "grad_norm": 0.46542630896814785, "learning_rate": 1.4955264972596023e-05, "loss": 0.3774, "step": 772 }, { "epoch": 0.24956616489769562, "grad_norm": 0.4181531104938061, "learning_rate": 1.4954973356976551e-05, "loss": 0.4351, "step": 773 }, { "epoch": 0.2498890189273175, "grad_norm": 0.3856048307567118, "learning_rate": 1.4954680796822223e-05, "loss": 0.3637, "step": 774 }, { "epoch": 0.25021187295693936, "grad_norm": 0.4626397227810222, "learning_rate": 1.49543872921701e-05, "loss": 0.3667, "step": 775 }, { "epoch": 0.2505347269865612, "grad_norm": 0.3797404342469076, "learning_rate": 1.4954092843057374e-05, "loss": 0.3997, "step": 776 }, { "epoch": 0.25085758101618305, "grad_norm": 0.40898491742952425, "learning_rate": 1.495379744952135e-05, "loss": 0.3127, "step": 777 }, { "epoch": 0.2511804350458049, "grad_norm": 0.39781584970196326, "learning_rate": 1.4953501111599452e-05, "loss": 0.369, "step": 778 }, { "epoch": 0.2515032890754268, "grad_norm": 0.4027645302539348, "learning_rate": 1.4953203829329228e-05, "loss": 0.399, "step": 779 }, { "epoch": 0.25182614310504864, "grad_norm": 0.4453866348033251, "learning_rate": 1.4952905602748341e-05, "loss": 0.3175, "step": 780 }, { "epoch": 0.2521489971346705, "grad_norm": 0.45803865879307787, "learning_rate": 1.495260643189458e-05, "loss": 0.3583, "step": 781 }, { "epoch": 0.2524718511642923, "grad_norm": 0.4116543576910319, "learning_rate": 1.4952306316805846e-05, "loss": 0.3213, "step": 782 }, { "epoch": 0.2527947051939142, "grad_norm": 0.5071242725429222, "learning_rate": 1.4952005257520165e-05, "loss": 0.3597, "step": 783 }, { "epoch": 0.25311755922353607, "grad_norm": 0.4856178406649537, "learning_rate": 1.4951703254075677e-05, "loss": 0.3123, "step": 784 }, { "epoch": 0.2534404132531579, "grad_norm": 0.41858086359412117, "learning_rate": 1.4951400306510651e-05, "loss": 0.3438, "step": 785 }, { "epoch": 0.25376326728277976, "grad_norm": 0.44258402599058616, "learning_rate": 1.4951096414863465e-05, "loss": 0.423, "step": 786 }, { "epoch": 0.2540861213124016, "grad_norm": 0.6785503519727456, "learning_rate": 1.4950791579172628e-05, "loss": 0.4272, "step": 787 }, { "epoch": 0.2544089753420235, "grad_norm": 0.4291311082894695, "learning_rate": 1.4950485799476756e-05, "loss": 0.3539, "step": 788 }, { "epoch": 0.25473182937164535, "grad_norm": 0.39433323674959586, "learning_rate": 1.4950179075814592e-05, "loss": 0.4031, "step": 789 }, { "epoch": 0.2550546834012672, "grad_norm": 0.48086782114229903, "learning_rate": 1.4949871408225e-05, "loss": 0.4062, "step": 790 }, { "epoch": 0.25537753743088903, "grad_norm": 0.9838864050371834, "learning_rate": 1.4949562796746959e-05, "loss": 0.3314, "step": 791 }, { "epoch": 0.25570039146051093, "grad_norm": 0.4642812963244718, "learning_rate": 1.494925324141957e-05, "loss": 0.3777, "step": 792 }, { "epoch": 0.2560232454901328, "grad_norm": 0.3942738772442868, "learning_rate": 1.4948942742282056e-05, "loss": 0.3952, "step": 793 }, { "epoch": 0.2563460995197546, "grad_norm": 0.5092257373054465, "learning_rate": 1.494863129937375e-05, "loss": 0.4047, "step": 794 }, { "epoch": 0.25666895354937647, "grad_norm": 0.5935298145929452, "learning_rate": 1.494831891273412e-05, "loss": 0.3997, "step": 795 }, { "epoch": 0.25699180757899837, "grad_norm": 0.43394155640481663, "learning_rate": 1.4948005582402738e-05, "loss": 0.4436, "step": 796 }, { "epoch": 0.2573146616086202, "grad_norm": 0.35722448638249243, "learning_rate": 1.4947691308419305e-05, "loss": 0.3977, "step": 797 }, { "epoch": 0.25763751563824205, "grad_norm": 0.573100567404854, "learning_rate": 1.4947376090823641e-05, "loss": 0.4111, "step": 798 }, { "epoch": 0.2579603696678639, "grad_norm": 0.36928849845681233, "learning_rate": 1.494705992965568e-05, "loss": 0.4112, "step": 799 }, { "epoch": 0.2582832236974858, "grad_norm": 0.4080414825482638, "learning_rate": 1.4946742824955484e-05, "loss": 0.3511, "step": 800 }, { "epoch": 0.25860607772710764, "grad_norm": 0.5490207237169996, "learning_rate": 1.4946424776763226e-05, "loss": 0.4057, "step": 801 }, { "epoch": 0.2589289317567295, "grad_norm": 0.5847671155033782, "learning_rate": 1.4946105785119202e-05, "loss": 0.4506, "step": 802 }, { "epoch": 0.25925178578635133, "grad_norm": 0.37979539342337026, "learning_rate": 1.4945785850063828e-05, "loss": 0.3918, "step": 803 }, { "epoch": 0.25957463981597323, "grad_norm": 0.4321417074496389, "learning_rate": 1.4945464971637644e-05, "loss": 0.3575, "step": 804 }, { "epoch": 0.2598974938455951, "grad_norm": 0.5312177441689341, "learning_rate": 1.4945143149881299e-05, "loss": 0.3537, "step": 805 }, { "epoch": 0.2602203478752169, "grad_norm": 0.4403169093688816, "learning_rate": 1.4944820384835568e-05, "loss": 0.4025, "step": 806 }, { "epoch": 0.26054320190483876, "grad_norm": 0.4471660781195587, "learning_rate": 1.4944496676541349e-05, "loss": 0.3785, "step": 807 }, { "epoch": 0.2608660559344606, "grad_norm": 0.5889680362361569, "learning_rate": 1.494417202503965e-05, "loss": 0.3756, "step": 808 }, { "epoch": 0.2611889099640825, "grad_norm": 0.46437627972782425, "learning_rate": 1.494384643037161e-05, "loss": 0.3127, "step": 809 }, { "epoch": 0.26151176399370435, "grad_norm": 0.36388031880690375, "learning_rate": 1.4943519892578475e-05, "loss": 0.3797, "step": 810 }, { "epoch": 0.2618346180233262, "grad_norm": 0.44079644405077534, "learning_rate": 1.4943192411701622e-05, "loss": 0.3933, "step": 811 }, { "epoch": 0.26215747205294804, "grad_norm": 0.39309399607912543, "learning_rate": 1.4942863987782538e-05, "loss": 0.336, "step": 812 }, { "epoch": 0.26248032608256994, "grad_norm": 0.43273367285404446, "learning_rate": 1.4942534620862839e-05, "loss": 0.3891, "step": 813 }, { "epoch": 0.2628031801121918, "grad_norm": 0.4596730521403857, "learning_rate": 1.494220431098425e-05, "loss": 0.3517, "step": 814 }, { "epoch": 0.2631260341418136, "grad_norm": 0.546783571904674, "learning_rate": 1.4941873058188624e-05, "loss": 0.324, "step": 815 }, { "epoch": 0.26344888817143547, "grad_norm": 0.53201739410015, "learning_rate": 1.494154086251793e-05, "loss": 0.3245, "step": 816 }, { "epoch": 0.26377174220105737, "grad_norm": 0.4551789359403826, "learning_rate": 1.4941207724014257e-05, "loss": 0.3829, "step": 817 }, { "epoch": 0.2640945962306792, "grad_norm": 0.4511714713371321, "learning_rate": 1.494087364271981e-05, "loss": 0.4093, "step": 818 }, { "epoch": 0.26441745026030106, "grad_norm": 0.4198476838931148, "learning_rate": 1.4940538618676923e-05, "loss": 0.344, "step": 819 }, { "epoch": 0.2647403042899229, "grad_norm": 0.46719352649804874, "learning_rate": 1.4940202651928037e-05, "loss": 0.3575, "step": 820 }, { "epoch": 0.2650631583195448, "grad_norm": 0.3469121685472725, "learning_rate": 1.493986574251572e-05, "loss": 0.3043, "step": 821 }, { "epoch": 0.26538601234916664, "grad_norm": 0.38992424978091134, "learning_rate": 1.493952789048266e-05, "loss": 0.3906, "step": 822 }, { "epoch": 0.2657088663787885, "grad_norm": 0.4523306233098873, "learning_rate": 1.4939189095871662e-05, "loss": 0.3116, "step": 823 }, { "epoch": 0.26603172040841033, "grad_norm": 0.43126610138961335, "learning_rate": 1.4938849358725649e-05, "loss": 0.3571, "step": 824 }, { "epoch": 0.26635457443803223, "grad_norm": 0.5283160628059298, "learning_rate": 1.4938508679087665e-05, "loss": 0.3386, "step": 825 }, { "epoch": 0.2666774284676541, "grad_norm": 0.5799961656541867, "learning_rate": 1.4938167057000877e-05, "loss": 0.3562, "step": 826 }, { "epoch": 0.2670002824972759, "grad_norm": 0.3872236106763124, "learning_rate": 1.4937824492508564e-05, "loss": 0.4271, "step": 827 }, { "epoch": 0.26732313652689776, "grad_norm": 0.39472971236236115, "learning_rate": 1.4937480985654132e-05, "loss": 0.3325, "step": 828 }, { "epoch": 0.2676459905565196, "grad_norm": 0.4512973107988881, "learning_rate": 1.4937136536481099e-05, "loss": 0.3663, "step": 829 }, { "epoch": 0.2679688445861415, "grad_norm": 0.36383920629341926, "learning_rate": 1.4936791145033109e-05, "loss": 0.3572, "step": 830 }, { "epoch": 0.26829169861576335, "grad_norm": 0.45794025572346847, "learning_rate": 1.4936444811353922e-05, "loss": 0.4145, "step": 831 }, { "epoch": 0.2686145526453852, "grad_norm": 0.5064427387822612, "learning_rate": 1.4936097535487418e-05, "loss": 0.3625, "step": 832 }, { "epoch": 0.26893740667500704, "grad_norm": 0.4299750406308721, "learning_rate": 1.4935749317477597e-05, "loss": 0.4035, "step": 833 }, { "epoch": 0.26926026070462894, "grad_norm": 0.5678710266984479, "learning_rate": 1.4935400157368579e-05, "loss": 0.3364, "step": 834 }, { "epoch": 0.2695831147342508, "grad_norm": 0.40690904235698483, "learning_rate": 1.4935050055204597e-05, "loss": 0.3844, "step": 835 }, { "epoch": 0.26990596876387263, "grad_norm": 0.4291779088235278, "learning_rate": 1.4934699011030012e-05, "loss": 0.3409, "step": 836 }, { "epoch": 0.27022882279349447, "grad_norm": 0.4407262534606147, "learning_rate": 1.4934347024889301e-05, "loss": 0.3128, "step": 837 }, { "epoch": 0.27055167682311637, "grad_norm": 0.5676482891152783, "learning_rate": 1.493399409682706e-05, "loss": 0.3836, "step": 838 }, { "epoch": 0.2708745308527382, "grad_norm": 0.4012122754813037, "learning_rate": 1.4933640226888005e-05, "loss": 0.3574, "step": 839 }, { "epoch": 0.27119738488236006, "grad_norm": 0.5132717062656185, "learning_rate": 1.4933285415116968e-05, "loss": 0.3985, "step": 840 }, { "epoch": 0.2715202389119819, "grad_norm": 0.4281002057860103, "learning_rate": 1.4932929661558907e-05, "loss": 0.4212, "step": 841 }, { "epoch": 0.2718430929416038, "grad_norm": 0.6102827941427885, "learning_rate": 1.4932572966258893e-05, "loss": 0.2877, "step": 842 }, { "epoch": 0.27216594697122565, "grad_norm": 0.4715283839558377, "learning_rate": 1.4932215329262119e-05, "loss": 0.3664, "step": 843 }, { "epoch": 0.2724888010008475, "grad_norm": 0.5488572320193519, "learning_rate": 1.4931856750613897e-05, "loss": 0.3521, "step": 844 }, { "epoch": 0.27281165503046934, "grad_norm": 0.430366768095245, "learning_rate": 1.493149723035966e-05, "loss": 0.3692, "step": 845 }, { "epoch": 0.2731345090600912, "grad_norm": 0.43531172957415026, "learning_rate": 1.4931136768544957e-05, "loss": 0.3779, "step": 846 }, { "epoch": 0.2734573630897131, "grad_norm": 0.4152347222173556, "learning_rate": 1.4930775365215461e-05, "loss": 0.3368, "step": 847 }, { "epoch": 0.2737802171193349, "grad_norm": 0.4644283771264317, "learning_rate": 1.4930413020416958e-05, "loss": 0.3631, "step": 848 }, { "epoch": 0.27410307114895677, "grad_norm": 0.4161184632326264, "learning_rate": 1.4930049734195357e-05, "loss": 0.4172, "step": 849 }, { "epoch": 0.2744259251785786, "grad_norm": 0.447485801614374, "learning_rate": 1.4929685506596685e-05, "loss": 0.3491, "step": 850 }, { "epoch": 0.2747487792082005, "grad_norm": 0.47377420285974925, "learning_rate": 1.4929320337667092e-05, "loss": 0.3981, "step": 851 }, { "epoch": 0.27507163323782235, "grad_norm": 0.46045229677694727, "learning_rate": 1.4928954227452842e-05, "loss": 0.3632, "step": 852 }, { "epoch": 0.2753944872674442, "grad_norm": 0.5046407361293903, "learning_rate": 1.4928587176000323e-05, "loss": 0.3364, "step": 853 }, { "epoch": 0.27571734129706604, "grad_norm": 0.4058163259163864, "learning_rate": 1.4928219183356039e-05, "loss": 0.3077, "step": 854 }, { "epoch": 0.27604019532668794, "grad_norm": 0.423192763328646, "learning_rate": 1.4927850249566611e-05, "loss": 0.3865, "step": 855 }, { "epoch": 0.2763630493563098, "grad_norm": 0.42796650356428534, "learning_rate": 1.492748037467879e-05, "loss": 0.3184, "step": 856 }, { "epoch": 0.27668590338593163, "grad_norm": 0.5600019314190188, "learning_rate": 1.4927109558739428e-05, "loss": 0.3893, "step": 857 }, { "epoch": 0.2770087574155535, "grad_norm": 0.4452416355909798, "learning_rate": 1.4926737801795516e-05, "loss": 0.3702, "step": 858 }, { "epoch": 0.2773316114451754, "grad_norm": 0.4227031677426542, "learning_rate": 1.4926365103894153e-05, "loss": 0.3564, "step": 859 }, { "epoch": 0.2776544654747972, "grad_norm": 0.3777673027814674, "learning_rate": 1.4925991465082555e-05, "loss": 0.3354, "step": 860 }, { "epoch": 0.27797731950441906, "grad_norm": 0.9077022607101818, "learning_rate": 1.4925616885408067e-05, "loss": 0.3853, "step": 861 }, { "epoch": 0.2783001735340409, "grad_norm": 0.4837843807737776, "learning_rate": 1.4925241364918144e-05, "loss": 0.4454, "step": 862 }, { "epoch": 0.2786230275636628, "grad_norm": 0.4904329446622005, "learning_rate": 1.4924864903660364e-05, "loss": 0.3583, "step": 863 }, { "epoch": 0.27894588159328465, "grad_norm": 0.3538977042144304, "learning_rate": 1.4924487501682428e-05, "loss": 0.4069, "step": 864 }, { "epoch": 0.2792687356229065, "grad_norm": 0.4030906709365191, "learning_rate": 1.4924109159032148e-05, "loss": 0.3547, "step": 865 }, { "epoch": 0.27959158965252834, "grad_norm": 0.42931877540782065, "learning_rate": 1.4923729875757461e-05, "loss": 0.403, "step": 866 }, { "epoch": 0.2799144436821502, "grad_norm": 0.41301249559996195, "learning_rate": 1.4923349651906422e-05, "loss": 0.32, "step": 867 }, { "epoch": 0.2802372977117721, "grad_norm": 0.45097047474157786, "learning_rate": 1.4922968487527204e-05, "loss": 0.3721, "step": 868 }, { "epoch": 0.2805601517413939, "grad_norm": 0.5019876746844114, "learning_rate": 1.4922586382668105e-05, "loss": 0.2993, "step": 869 }, { "epoch": 0.28088300577101577, "grad_norm": 0.42751795117967534, "learning_rate": 1.4922203337377528e-05, "loss": 0.3627, "step": 870 }, { "epoch": 0.2812058598006376, "grad_norm": 0.42641599809990566, "learning_rate": 1.4921819351704011e-05, "loss": 0.3275, "step": 871 }, { "epoch": 0.2815287138302595, "grad_norm": 0.47562875073396865, "learning_rate": 1.4921434425696202e-05, "loss": 0.3723, "step": 872 }, { "epoch": 0.28185156785988136, "grad_norm": 0.4386386496706318, "learning_rate": 1.492104855940287e-05, "loss": 0.3621, "step": 873 }, { "epoch": 0.2821744218895032, "grad_norm": 0.5754569355858813, "learning_rate": 1.4920661752872908e-05, "loss": 0.3557, "step": 874 }, { "epoch": 0.28249727591912505, "grad_norm": 0.4177445922366082, "learning_rate": 1.4920274006155318e-05, "loss": 0.376, "step": 875 }, { "epoch": 0.28282012994874695, "grad_norm": 2.651312381883615, "learning_rate": 1.491988531929923e-05, "loss": 0.3025, "step": 876 }, { "epoch": 0.2831429839783688, "grad_norm": 0.47959522807633026, "learning_rate": 1.4919495692353892e-05, "loss": 0.4326, "step": 877 }, { "epoch": 0.28346583800799063, "grad_norm": 0.5036530897765782, "learning_rate": 1.4919105125368667e-05, "loss": 0.3358, "step": 878 }, { "epoch": 0.2837886920376125, "grad_norm": 0.6742560419485244, "learning_rate": 1.4918713618393038e-05, "loss": 0.346, "step": 879 }, { "epoch": 0.2841115460672344, "grad_norm": 0.4234848763521654, "learning_rate": 1.4918321171476607e-05, "loss": 0.3755, "step": 880 }, { "epoch": 0.2844344000968562, "grad_norm": 0.5093983772992946, "learning_rate": 1.4917927784669103e-05, "loss": 0.3594, "step": 881 }, { "epoch": 0.28475725412647807, "grad_norm": 0.48190728974330604, "learning_rate": 1.4917533458020363e-05, "loss": 0.3232, "step": 882 }, { "epoch": 0.2850801081560999, "grad_norm": 0.45816679068198557, "learning_rate": 1.4917138191580349e-05, "loss": 0.3656, "step": 883 }, { "epoch": 0.28540296218572175, "grad_norm": 0.4192522701625645, "learning_rate": 1.4916741985399138e-05, "loss": 0.3339, "step": 884 }, { "epoch": 0.28572581621534365, "grad_norm": 0.5138737281158097, "learning_rate": 1.4916344839526933e-05, "loss": 0.3525, "step": 885 }, { "epoch": 0.2860486702449655, "grad_norm": 0.4780901787889241, "learning_rate": 1.4915946754014049e-05, "loss": 0.402, "step": 886 }, { "epoch": 0.28637152427458734, "grad_norm": 0.597868763219833, "learning_rate": 1.4915547728910923e-05, "loss": 0.3537, "step": 887 }, { "epoch": 0.2866943783042092, "grad_norm": 0.3977233499785006, "learning_rate": 1.4915147764268112e-05, "loss": 0.4111, "step": 888 }, { "epoch": 0.2870172323338311, "grad_norm": 0.4452496526073299, "learning_rate": 1.4914746860136292e-05, "loss": 0.4331, "step": 889 }, { "epoch": 0.28734008636345293, "grad_norm": 0.5441479358376469, "learning_rate": 1.4914345016566254e-05, "loss": 0.3951, "step": 890 }, { "epoch": 0.2876629403930748, "grad_norm": 0.3922550263116025, "learning_rate": 1.4913942233608913e-05, "loss": 0.4152, "step": 891 }, { "epoch": 0.2879857944226966, "grad_norm": 0.37855258677193565, "learning_rate": 1.49135385113153e-05, "loss": 0.4101, "step": 892 }, { "epoch": 0.2883086484523185, "grad_norm": 0.47781564413925726, "learning_rate": 1.4913133849736568e-05, "loss": 0.4138, "step": 893 }, { "epoch": 0.28863150248194036, "grad_norm": 0.3877736550209192, "learning_rate": 1.4912728248923985e-05, "loss": 0.3604, "step": 894 }, { "epoch": 0.2889543565115622, "grad_norm": 0.5675072621838361, "learning_rate": 1.491232170892894e-05, "loss": 0.3071, "step": 895 }, { "epoch": 0.28927721054118405, "grad_norm": 0.5205306499354604, "learning_rate": 1.4911914229802943e-05, "loss": 0.4192, "step": 896 }, { "epoch": 0.28960006457080595, "grad_norm": 0.48745935481957137, "learning_rate": 1.4911505811597618e-05, "loss": 0.3707, "step": 897 }, { "epoch": 0.2899229186004278, "grad_norm": 0.5698911150154825, "learning_rate": 1.4911096454364715e-05, "loss": 0.4579, "step": 898 }, { "epoch": 0.29024577263004964, "grad_norm": 0.4507545817942562, "learning_rate": 1.4910686158156096e-05, "loss": 0.419, "step": 899 }, { "epoch": 0.2905686266596715, "grad_norm": 0.7657364993692517, "learning_rate": 1.4910274923023747e-05, "loss": 0.3146, "step": 900 }, { "epoch": 0.2908914806892934, "grad_norm": 0.4978090510755483, "learning_rate": 1.4909862749019767e-05, "loss": 0.4026, "step": 901 }, { "epoch": 0.2912143347189152, "grad_norm": 0.44935643475056103, "learning_rate": 1.4909449636196383e-05, "loss": 0.3762, "step": 902 }, { "epoch": 0.29153718874853707, "grad_norm": 0.6588545280697379, "learning_rate": 1.4909035584605933e-05, "loss": 0.3166, "step": 903 }, { "epoch": 0.2918600427781589, "grad_norm": 0.448892201533572, "learning_rate": 1.4908620594300879e-05, "loss": 0.367, "step": 904 }, { "epoch": 0.29218289680778076, "grad_norm": 0.34556045723565254, "learning_rate": 1.4908204665333794e-05, "loss": 0.3287, "step": 905 }, { "epoch": 0.29250575083740266, "grad_norm": 0.49647890414049894, "learning_rate": 1.4907787797757382e-05, "loss": 0.3636, "step": 906 }, { "epoch": 0.2928286048670245, "grad_norm": 0.4066907058348886, "learning_rate": 1.4907369991624459e-05, "loss": 0.3395, "step": 907 }, { "epoch": 0.29315145889664634, "grad_norm": 0.4181009236404524, "learning_rate": 1.4906951246987957e-05, "loss": 0.3653, "step": 908 }, { "epoch": 0.2934743129262682, "grad_norm": 0.4743676611390152, "learning_rate": 1.4906531563900931e-05, "loss": 0.4251, "step": 909 }, { "epoch": 0.2937971669558901, "grad_norm": 0.5102867996504427, "learning_rate": 1.4906110942416555e-05, "loss": 0.44, "step": 910 }, { "epoch": 0.29412002098551193, "grad_norm": 0.5303547510981473, "learning_rate": 1.4905689382588123e-05, "loss": 0.4346, "step": 911 }, { "epoch": 0.2944428750151338, "grad_norm": 0.5976981195703334, "learning_rate": 1.4905266884469044e-05, "loss": 0.3521, "step": 912 }, { "epoch": 0.2947657290447556, "grad_norm": 0.4215303675197153, "learning_rate": 1.490484344811285e-05, "loss": 0.4265, "step": 913 }, { "epoch": 0.2950885830743775, "grad_norm": 0.4866554955116059, "learning_rate": 1.4904419073573185e-05, "loss": 0.3979, "step": 914 }, { "epoch": 0.29541143710399936, "grad_norm": 0.5093064629822341, "learning_rate": 1.490399376090382e-05, "loss": 0.3607, "step": 915 }, { "epoch": 0.2957342911336212, "grad_norm": 0.4929293818802504, "learning_rate": 1.4903567510158644e-05, "loss": 0.4151, "step": 916 }, { "epoch": 0.29605714516324305, "grad_norm": 0.5060662185325279, "learning_rate": 1.4903140321391657e-05, "loss": 0.3662, "step": 917 }, { "epoch": 0.29637999919286495, "grad_norm": 0.5044412084523249, "learning_rate": 1.490271219465699e-05, "loss": 0.3771, "step": 918 }, { "epoch": 0.2967028532224868, "grad_norm": 0.40296216051611033, "learning_rate": 1.490228313000888e-05, "loss": 0.4253, "step": 919 }, { "epoch": 0.29702570725210864, "grad_norm": 0.5671905100528465, "learning_rate": 1.490185312750169e-05, "loss": 0.3535, "step": 920 }, { "epoch": 0.2973485612817305, "grad_norm": 0.5225196369962574, "learning_rate": 1.4901422187189903e-05, "loss": 0.3731, "step": 921 }, { "epoch": 0.2976714153113523, "grad_norm": 0.5860281878480997, "learning_rate": 1.4900990309128114e-05, "loss": 0.3859, "step": 922 }, { "epoch": 0.2979942693409742, "grad_norm": 0.4131139839059648, "learning_rate": 1.4900557493371049e-05, "loss": 0.3942, "step": 923 }, { "epoch": 0.29831712337059607, "grad_norm": 0.445479793146334, "learning_rate": 1.4900123739973539e-05, "loss": 0.3267, "step": 924 }, { "epoch": 0.2986399774002179, "grad_norm": 0.41987890544154655, "learning_rate": 1.489968904899054e-05, "loss": 0.4489, "step": 925 }, { "epoch": 0.29896283142983976, "grad_norm": 0.4264858305890996, "learning_rate": 1.4899253420477131e-05, "loss": 0.408, "step": 926 }, { "epoch": 0.29928568545946166, "grad_norm": 0.425815911916243, "learning_rate": 1.4898816854488502e-05, "loss": 0.4034, "step": 927 }, { "epoch": 0.2996085394890835, "grad_norm": 0.563817270240366, "learning_rate": 1.4898379351079966e-05, "loss": 0.3582, "step": 928 }, { "epoch": 0.29993139351870535, "grad_norm": 0.7179502537913376, "learning_rate": 1.4897940910306957e-05, "loss": 0.3821, "step": 929 }, { "epoch": 0.3002542475483272, "grad_norm": 0.4433121305548897, "learning_rate": 1.4897501532225018e-05, "loss": 0.4302, "step": 930 }, { "epoch": 0.3005771015779491, "grad_norm": 0.7555876025783643, "learning_rate": 1.4897061216889823e-05, "loss": 0.3749, "step": 931 }, { "epoch": 0.30089995560757093, "grad_norm": 0.4044394906848734, "learning_rate": 1.489661996435716e-05, "loss": 0.3419, "step": 932 }, { "epoch": 0.3012228096371928, "grad_norm": 0.46828460698764485, "learning_rate": 1.4896177774682934e-05, "loss": 0.4208, "step": 933 }, { "epoch": 0.3015456636668146, "grad_norm": 0.4772322582341112, "learning_rate": 1.4895734647923167e-05, "loss": 0.347, "step": 934 }, { "epoch": 0.3018685176964365, "grad_norm": 0.3904386302590894, "learning_rate": 1.4895290584134004e-05, "loss": 0.3674, "step": 935 }, { "epoch": 0.30219137172605837, "grad_norm": 0.42378177069546485, "learning_rate": 1.4894845583371708e-05, "loss": 0.3894, "step": 936 }, { "epoch": 0.3025142257556802, "grad_norm": 0.44761303667766605, "learning_rate": 1.4894399645692662e-05, "loss": 0.4143, "step": 937 }, { "epoch": 0.30283707978530205, "grad_norm": 0.40173359961968985, "learning_rate": 1.4893952771153361e-05, "loss": 0.2952, "step": 938 }, { "epoch": 0.30315993381492395, "grad_norm": 0.823713939635508, "learning_rate": 1.4893504959810427e-05, "loss": 0.3924, "step": 939 }, { "epoch": 0.3034827878445458, "grad_norm": 0.3712425556684154, "learning_rate": 1.4893056211720599e-05, "loss": 0.3638, "step": 940 }, { "epoch": 0.30380564187416764, "grad_norm": 0.4193583895424693, "learning_rate": 1.4892606526940728e-05, "loss": 0.3571, "step": 941 }, { "epoch": 0.3041284959037895, "grad_norm": 0.4610490336861531, "learning_rate": 1.4892155905527789e-05, "loss": 0.4008, "step": 942 }, { "epoch": 0.30445134993341133, "grad_norm": 0.41309724815196996, "learning_rate": 1.4891704347538876e-05, "loss": 0.3298, "step": 943 }, { "epoch": 0.30477420396303323, "grad_norm": 0.46418872852433046, "learning_rate": 1.4891251853031203e-05, "loss": 0.3809, "step": 944 }, { "epoch": 0.3050970579926551, "grad_norm": 0.5059930973138959, "learning_rate": 1.4890798422062097e-05, "loss": 0.3607, "step": 945 }, { "epoch": 0.3054199120222769, "grad_norm": 0.36828491594778723, "learning_rate": 1.489034405468901e-05, "loss": 0.3446, "step": 946 }, { "epoch": 0.30574276605189876, "grad_norm": 0.46187582833189705, "learning_rate": 1.4889888750969508e-05, "loss": 0.3854, "step": 947 }, { "epoch": 0.30606562008152066, "grad_norm": 0.4198020927266921, "learning_rate": 1.4889432510961276e-05, "loss": 0.3581, "step": 948 }, { "epoch": 0.3063884741111425, "grad_norm": 0.5408210705513631, "learning_rate": 1.4888975334722122e-05, "loss": 0.382, "step": 949 }, { "epoch": 0.30671132814076435, "grad_norm": 0.5056213195385783, "learning_rate": 1.488851722230997e-05, "loss": 0.3915, "step": 950 }, { "epoch": 0.3070341821703862, "grad_norm": 0.5483247354302975, "learning_rate": 1.4888058173782859e-05, "loss": 0.3804, "step": 951 }, { "epoch": 0.3073570362000081, "grad_norm": 0.6760035689926133, "learning_rate": 1.4887598189198952e-05, "loss": 0.3565, "step": 952 }, { "epoch": 0.30767989022962994, "grad_norm": 0.42888889847931255, "learning_rate": 1.4887137268616526e-05, "loss": 0.309, "step": 953 }, { "epoch": 0.3080027442592518, "grad_norm": 0.5039700532516886, "learning_rate": 1.4886675412093981e-05, "loss": 0.3512, "step": 954 }, { "epoch": 0.3083255982888736, "grad_norm": 0.49393552352329506, "learning_rate": 1.4886212619689834e-05, "loss": 0.3591, "step": 955 }, { "epoch": 0.3086484523184955, "grad_norm": 0.7712109354596991, "learning_rate": 1.4885748891462719e-05, "loss": 0.4524, "step": 956 }, { "epoch": 0.30897130634811737, "grad_norm": 0.4340474053587945, "learning_rate": 1.4885284227471392e-05, "loss": 0.4266, "step": 957 }, { "epoch": 0.3092941603777392, "grad_norm": 0.5412929974379694, "learning_rate": 1.4884818627774723e-05, "loss": 0.398, "step": 958 }, { "epoch": 0.30961701440736106, "grad_norm": 0.3876603233724719, "learning_rate": 1.4884352092431703e-05, "loss": 0.424, "step": 959 }, { "epoch": 0.3099398684369829, "grad_norm": 0.4473508001670218, "learning_rate": 1.4883884621501441e-05, "loss": 0.344, "step": 960 }, { "epoch": 0.3102627224666048, "grad_norm": 0.38468406474153033, "learning_rate": 1.4883416215043167e-05, "loss": 0.4258, "step": 961 }, { "epoch": 0.31058557649622665, "grad_norm": 0.4663003642030701, "learning_rate": 1.4882946873116224e-05, "loss": 0.3486, "step": 962 }, { "epoch": 0.3109084305258485, "grad_norm": 0.6375611729425938, "learning_rate": 1.4882476595780081e-05, "loss": 0.3223, "step": 963 }, { "epoch": 0.31123128455547033, "grad_norm": 0.4702279193734361, "learning_rate": 1.4882005383094321e-05, "loss": 0.3539, "step": 964 }, { "epoch": 0.31155413858509223, "grad_norm": 0.49220471240567804, "learning_rate": 1.488153323511864e-05, "loss": 0.3585, "step": 965 }, { "epoch": 0.3118769926147141, "grad_norm": 0.854719160841799, "learning_rate": 1.4881060151912868e-05, "loss": 0.3848, "step": 966 }, { "epoch": 0.3121998466443359, "grad_norm": 0.5349045500004375, "learning_rate": 1.4880586133536936e-05, "loss": 0.3534, "step": 967 }, { "epoch": 0.31252270067395777, "grad_norm": 0.4212818242606985, "learning_rate": 1.4880111180050905e-05, "loss": 0.3831, "step": 968 }, { "epoch": 0.31284555470357966, "grad_norm": 0.43710905320539234, "learning_rate": 1.4879635291514952e-05, "loss": 0.3402, "step": 969 }, { "epoch": 0.3131684087332015, "grad_norm": 0.6094887034919737, "learning_rate": 1.4879158467989368e-05, "loss": 0.3223, "step": 970 }, { "epoch": 0.31349126276282335, "grad_norm": 0.4795281264612259, "learning_rate": 1.4878680709534569e-05, "loss": 0.3523, "step": 971 }, { "epoch": 0.3138141167924452, "grad_norm": 0.6761428981134174, "learning_rate": 1.4878202016211086e-05, "loss": 0.3684, "step": 972 }, { "epoch": 0.3141369708220671, "grad_norm": 0.5088865709345151, "learning_rate": 1.4877722388079566e-05, "loss": 0.4297, "step": 973 }, { "epoch": 0.31445982485168894, "grad_norm": 0.6122936956328399, "learning_rate": 1.4877241825200777e-05, "loss": 0.3604, "step": 974 }, { "epoch": 0.3147826788813108, "grad_norm": 0.4412004725290722, "learning_rate": 1.4876760327635612e-05, "loss": 0.3072, "step": 975 }, { "epoch": 0.31510553291093263, "grad_norm": 0.46901739209847665, "learning_rate": 1.4876277895445072e-05, "loss": 0.41, "step": 976 }, { "epoch": 0.31542838694055453, "grad_norm": 0.5107821305890843, "learning_rate": 1.4875794528690277e-05, "loss": 0.3995, "step": 977 }, { "epoch": 0.3157512409701764, "grad_norm": 0.37481743357310454, "learning_rate": 1.4875310227432475e-05, "loss": 0.3413, "step": 978 }, { "epoch": 0.3160740949997982, "grad_norm": 0.6067667514244226, "learning_rate": 1.4874824991733023e-05, "loss": 0.4339, "step": 979 }, { "epoch": 0.31639694902942006, "grad_norm": 0.5236363181705211, "learning_rate": 1.4874338821653402e-05, "loss": 0.3404, "step": 980 }, { "epoch": 0.3167198030590419, "grad_norm": 0.4498377269078263, "learning_rate": 1.4873851717255205e-05, "loss": 0.418, "step": 981 }, { "epoch": 0.3170426570886638, "grad_norm": 0.3383020644545434, "learning_rate": 1.4873363678600149e-05, "loss": 0.3488, "step": 982 }, { "epoch": 0.31736551111828565, "grad_norm": 0.4605969388847885, "learning_rate": 1.4872874705750071e-05, "loss": 0.4057, "step": 983 }, { "epoch": 0.3176883651479075, "grad_norm": 0.5244359018626412, "learning_rate": 1.4872384798766917e-05, "loss": 0.3799, "step": 984 }, { "epoch": 0.31801121917752934, "grad_norm": 0.5201456186873642, "learning_rate": 1.4871893957712767e-05, "loss": 0.3388, "step": 985 }, { "epoch": 0.31833407320715124, "grad_norm": 0.4478027424690723, "learning_rate": 1.4871402182649802e-05, "loss": 0.3946, "step": 986 }, { "epoch": 0.3186569272367731, "grad_norm": 0.6076748889745864, "learning_rate": 1.4870909473640329e-05, "loss": 0.3869, "step": 987 }, { "epoch": 0.3189797812663949, "grad_norm": 0.5444096855190196, "learning_rate": 1.4870415830746779e-05, "loss": 0.3501, "step": 988 }, { "epoch": 0.31930263529601677, "grad_norm": 1.1832240643646432, "learning_rate": 1.4869921254031692e-05, "loss": 0.4289, "step": 989 }, { "epoch": 0.31962548932563867, "grad_norm": 0.4327634211026097, "learning_rate": 1.4869425743557729e-05, "loss": 0.3428, "step": 990 }, { "epoch": 0.3199483433552605, "grad_norm": 0.8213718368938326, "learning_rate": 1.4868929299387677e-05, "loss": 0.3447, "step": 991 }, { "epoch": 0.32027119738488236, "grad_norm": 0.41203265845833387, "learning_rate": 1.4868431921584426e-05, "loss": 0.2979, "step": 992 }, { "epoch": 0.3205940514145042, "grad_norm": 0.3975030134549701, "learning_rate": 1.4867933610211e-05, "loss": 0.3481, "step": 993 }, { "epoch": 0.3209169054441261, "grad_norm": 0.7842167841767383, "learning_rate": 1.4867434365330532e-05, "loss": 0.4391, "step": 994 }, { "epoch": 0.32123975947374794, "grad_norm": 0.5163454765349932, "learning_rate": 1.4866934187006273e-05, "loss": 0.3382, "step": 995 }, { "epoch": 0.3215626135033698, "grad_norm": 0.5078451897559407, "learning_rate": 1.4866433075301602e-05, "loss": 0.4223, "step": 996 }, { "epoch": 0.32188546753299163, "grad_norm": 0.3836476487413843, "learning_rate": 1.486593103028e-05, "loss": 0.4067, "step": 997 }, { "epoch": 0.32220832156261353, "grad_norm": 0.535342430075141, "learning_rate": 1.4865428052005082e-05, "loss": 0.3808, "step": 998 }, { "epoch": 0.3225311755922354, "grad_norm": 0.5040748366596722, "learning_rate": 1.4864924140540572e-05, "loss": 0.4083, "step": 999 }, { "epoch": 0.3228540296218572, "grad_norm": 0.4313647930023514, "learning_rate": 1.4864419295950315e-05, "loss": 0.4133, "step": 1000 }, { "epoch": 0.32317688365147906, "grad_norm": 0.375864175175609, "learning_rate": 1.4863913518298275e-05, "loss": 0.3787, "step": 1001 }, { "epoch": 0.3234997376811009, "grad_norm": 0.6208878737962067, "learning_rate": 1.4863406807648535e-05, "loss": 0.3585, "step": 1002 }, { "epoch": 0.3238225917107228, "grad_norm": 0.48736868936383504, "learning_rate": 1.486289916406529e-05, "loss": 0.3383, "step": 1003 }, { "epoch": 0.32414544574034465, "grad_norm": 0.6296037320113339, "learning_rate": 1.4862390587612862e-05, "loss": 0.3859, "step": 1004 }, { "epoch": 0.3244682997699665, "grad_norm": 0.5099230992532999, "learning_rate": 1.4861881078355683e-05, "loss": 0.3678, "step": 1005 }, { "epoch": 0.32479115379958834, "grad_norm": 0.5495371908296495, "learning_rate": 1.486137063635831e-05, "loss": 0.3475, "step": 1006 }, { "epoch": 0.32511400782921024, "grad_norm": 2.6665255453923087, "learning_rate": 1.4860859261685416e-05, "loss": 0.3829, "step": 1007 }, { "epoch": 0.3254368618588321, "grad_norm": 0.4733041366226227, "learning_rate": 1.486034695440179e-05, "loss": 0.4392, "step": 1008 }, { "epoch": 0.3257597158884539, "grad_norm": 0.414779943658683, "learning_rate": 1.4859833714572339e-05, "loss": 0.4141, "step": 1009 }, { "epoch": 0.32608256991807577, "grad_norm": 0.41687518728689105, "learning_rate": 1.4859319542262092e-05, "loss": 0.3457, "step": 1010 }, { "epoch": 0.32640542394769767, "grad_norm": 0.4417477072798294, "learning_rate": 1.4858804437536194e-05, "loss": 0.3406, "step": 1011 }, { "epoch": 0.3267282779773195, "grad_norm": 0.411663939950446, "learning_rate": 1.4858288400459907e-05, "loss": 0.3065, "step": 1012 }, { "epoch": 0.32705113200694136, "grad_norm": 0.43682743206760466, "learning_rate": 1.4857771431098614e-05, "loss": 0.4398, "step": 1013 }, { "epoch": 0.3273739860365632, "grad_norm": 0.6310699981588803, "learning_rate": 1.485725352951781e-05, "loss": 0.4325, "step": 1014 }, { "epoch": 0.3276968400661851, "grad_norm": 0.3976102209084725, "learning_rate": 1.4856734695783118e-05, "loss": 0.343, "step": 1015 }, { "epoch": 0.32801969409580695, "grad_norm": 0.5186125887461838, "learning_rate": 1.485621492996027e-05, "loss": 0.3304, "step": 1016 }, { "epoch": 0.3283425481254288, "grad_norm": 0.4969391789643401, "learning_rate": 1.485569423211512e-05, "loss": 0.3888, "step": 1017 }, { "epoch": 0.32866540215505063, "grad_norm": 0.39433950580953514, "learning_rate": 1.485517260231364e-05, "loss": 0.3085, "step": 1018 }, { "epoch": 0.3289882561846725, "grad_norm": 0.46183537805546065, "learning_rate": 1.485465004062192e-05, "loss": 0.4212, "step": 1019 }, { "epoch": 0.3293111102142944, "grad_norm": 0.5034570134711867, "learning_rate": 1.4854126547106168e-05, "loss": 0.4238, "step": 1020 }, { "epoch": 0.3296339642439162, "grad_norm": 0.4859608614099281, "learning_rate": 1.485360212183271e-05, "loss": 0.3413, "step": 1021 }, { "epoch": 0.32995681827353807, "grad_norm": 0.496902164007767, "learning_rate": 1.4853076764867991e-05, "loss": 0.3901, "step": 1022 }, { "epoch": 0.3302796723031599, "grad_norm": 0.5724580350271591, "learning_rate": 1.485255047627857e-05, "loss": 0.3934, "step": 1023 }, { "epoch": 0.3306025263327818, "grad_norm": 0.5118976690639165, "learning_rate": 1.4852023256131127e-05, "loss": 0.3704, "step": 1024 }, { "epoch": 0.33092538036240365, "grad_norm": 0.5224006895082741, "learning_rate": 1.4851495104492466e-05, "loss": 0.3957, "step": 1025 }, { "epoch": 0.3312482343920255, "grad_norm": 0.47567257190085493, "learning_rate": 1.4850966021429496e-05, "loss": 0.3949, "step": 1026 }, { "epoch": 0.33157108842164734, "grad_norm": 0.4072297164270231, "learning_rate": 1.4850436007009253e-05, "loss": 0.3877, "step": 1027 }, { "epoch": 0.33189394245126924, "grad_norm": 0.5023105858396758, "learning_rate": 1.4849905061298894e-05, "loss": 0.329, "step": 1028 }, { "epoch": 0.3322167964808911, "grad_norm": 0.45609223491729123, "learning_rate": 1.4849373184365683e-05, "loss": 0.4257, "step": 1029 }, { "epoch": 0.33253965051051293, "grad_norm": 2.1450451857525383, "learning_rate": 1.4848840376277011e-05, "loss": 0.2869, "step": 1030 }, { "epoch": 0.3328625045401348, "grad_norm": 0.4901412861401721, "learning_rate": 1.4848306637100381e-05, "loss": 0.3995, "step": 1031 }, { "epoch": 0.3331853585697567, "grad_norm": 0.46845039164143626, "learning_rate": 1.4847771966903421e-05, "loss": 0.3312, "step": 1032 }, { "epoch": 0.3335082125993785, "grad_norm": 0.42568486570650466, "learning_rate": 1.484723636575387e-05, "loss": 0.3188, "step": 1033 }, { "epoch": 0.33383106662900036, "grad_norm": 0.6365118850700535, "learning_rate": 1.484669983371959e-05, "loss": 0.3237, "step": 1034 }, { "epoch": 0.3341539206586222, "grad_norm": 0.39403129173964596, "learning_rate": 1.484616237086856e-05, "loss": 0.3637, "step": 1035 }, { "epoch": 0.3344767746882441, "grad_norm": 0.5673944651949763, "learning_rate": 1.4845623977268871e-05, "loss": 0.3943, "step": 1036 }, { "epoch": 0.33479962871786595, "grad_norm": 0.46272248304598845, "learning_rate": 1.4845084652988742e-05, "loss": 0.3214, "step": 1037 }, { "epoch": 0.3351224827474878, "grad_norm": 0.40076012936663824, "learning_rate": 1.48445443980965e-05, "loss": 0.3826, "step": 1038 }, { "epoch": 0.33544533677710964, "grad_norm": 0.4136879837360543, "learning_rate": 1.4844003212660598e-05, "loss": 0.3977, "step": 1039 }, { "epoch": 0.3357681908067315, "grad_norm": 0.5256187609065779, "learning_rate": 1.4843461096749604e-05, "loss": 0.4023, "step": 1040 }, { "epoch": 0.3360910448363534, "grad_norm": 0.4264023300012006, "learning_rate": 1.4842918050432198e-05, "loss": 0.4115, "step": 1041 }, { "epoch": 0.3364138988659752, "grad_norm": 0.3683400276202768, "learning_rate": 1.4842374073777189e-05, "loss": 0.3075, "step": 1042 }, { "epoch": 0.33673675289559707, "grad_norm": 0.696044156272593, "learning_rate": 1.4841829166853495e-05, "loss": 0.3936, "step": 1043 }, { "epoch": 0.3370596069252189, "grad_norm": 0.4677100797583049, "learning_rate": 1.4841283329730155e-05, "loss": 0.3359, "step": 1044 }, { "epoch": 0.3373824609548408, "grad_norm": 0.4251762311992063, "learning_rate": 1.4840736562476328e-05, "loss": 0.3493, "step": 1045 }, { "epoch": 0.33770531498446266, "grad_norm": 0.4655088943627301, "learning_rate": 1.4840188865161288e-05, "loss": 0.3561, "step": 1046 }, { "epoch": 0.3380281690140845, "grad_norm": 0.4604319365657758, "learning_rate": 1.4839640237854428e-05, "loss": 0.3335, "step": 1047 }, { "epoch": 0.33835102304370634, "grad_norm": 0.515375422596295, "learning_rate": 1.4839090680625253e-05, "loss": 0.336, "step": 1048 }, { "epoch": 0.33867387707332824, "grad_norm": 0.43971754133203605, "learning_rate": 1.4838540193543398e-05, "loss": 0.3474, "step": 1049 }, { "epoch": 0.3389967311029501, "grad_norm": 0.46353333444449213, "learning_rate": 1.4837988776678606e-05, "loss": 0.3533, "step": 1050 }, { "epoch": 0.33931958513257193, "grad_norm": 0.4451158441287189, "learning_rate": 1.483743643010074e-05, "loss": 0.3218, "step": 1051 }, { "epoch": 0.3396424391621938, "grad_norm": 0.559423391079411, "learning_rate": 1.4836883153879785e-05, "loss": 0.3383, "step": 1052 }, { "epoch": 0.3399652931918157, "grad_norm": 0.4096325045458375, "learning_rate": 1.4836328948085835e-05, "loss": 0.3186, "step": 1053 }, { "epoch": 0.3402881472214375, "grad_norm": 0.3699706776112085, "learning_rate": 1.483577381278911e-05, "loss": 0.3662, "step": 1054 }, { "epoch": 0.34061100125105936, "grad_norm": 0.3371558092944219, "learning_rate": 1.4835217748059949e-05, "loss": 0.3121, "step": 1055 }, { "epoch": 0.3409338552806812, "grad_norm": 0.5236325755648975, "learning_rate": 1.4834660753968795e-05, "loss": 0.4052, "step": 1056 }, { "epoch": 0.34125670931030305, "grad_norm": 0.4976378427555188, "learning_rate": 1.4834102830586228e-05, "loss": 0.3004, "step": 1057 }, { "epoch": 0.34157956333992495, "grad_norm": 0.3718177596080029, "learning_rate": 1.483354397798293e-05, "loss": 0.3039, "step": 1058 }, { "epoch": 0.3419024173695468, "grad_norm": 0.46613512226233167, "learning_rate": 1.483298419622971e-05, "loss": 0.3258, "step": 1059 }, { "epoch": 0.34222527139916864, "grad_norm": 0.4702749378870827, "learning_rate": 1.483242348539749e-05, "loss": 0.2745, "step": 1060 }, { "epoch": 0.3425481254287905, "grad_norm": 0.4538753564214289, "learning_rate": 1.4831861845557312e-05, "loss": 0.3558, "step": 1061 }, { "epoch": 0.3428709794584124, "grad_norm": 0.4212173054076696, "learning_rate": 1.4831299276780333e-05, "loss": 0.3468, "step": 1062 }, { "epoch": 0.34319383348803423, "grad_norm": 0.4833571476603375, "learning_rate": 1.4830735779137833e-05, "loss": 0.3384, "step": 1063 }, { "epoch": 0.34351668751765607, "grad_norm": 0.5237589131050726, "learning_rate": 1.4830171352701205e-05, "loss": 0.3332, "step": 1064 }, { "epoch": 0.3438395415472779, "grad_norm": 0.5347127064035512, "learning_rate": 1.482960599754196e-05, "loss": 0.393, "step": 1065 }, { "epoch": 0.3441623955768998, "grad_norm": 0.4683873949376129, "learning_rate": 1.4829039713731728e-05, "loss": 0.356, "step": 1066 }, { "epoch": 0.34448524960652166, "grad_norm": 0.458798388471045, "learning_rate": 1.482847250134226e-05, "loss": 0.3961, "step": 1067 }, { "epoch": 0.3448081036361435, "grad_norm": 0.41399110707631265, "learning_rate": 1.4827904360445416e-05, "loss": 0.337, "step": 1068 }, { "epoch": 0.34513095766576535, "grad_norm": 0.49076269236588993, "learning_rate": 1.482733529111318e-05, "loss": 0.3439, "step": 1069 }, { "epoch": 0.34545381169538725, "grad_norm": 0.4334873481103145, "learning_rate": 1.4826765293417654e-05, "loss": 0.3588, "step": 1070 }, { "epoch": 0.3457766657250091, "grad_norm": 0.43138179284377726, "learning_rate": 1.4826194367431055e-05, "loss": 0.3564, "step": 1071 }, { "epoch": 0.34609951975463094, "grad_norm": 0.4584151684446281, "learning_rate": 1.4825622513225717e-05, "loss": 0.4019, "step": 1072 }, { "epoch": 0.3464223737842528, "grad_norm": 0.4556236620721798, "learning_rate": 1.4825049730874095e-05, "loss": 0.3838, "step": 1073 }, { "epoch": 0.3467452278138747, "grad_norm": 0.3707475538880095, "learning_rate": 1.4824476020448762e-05, "loss": 0.3811, "step": 1074 }, { "epoch": 0.3470680818434965, "grad_norm": 0.41131495354063974, "learning_rate": 1.4823901382022401e-05, "loss": 0.3947, "step": 1075 }, { "epoch": 0.34739093587311837, "grad_norm": 0.5114108772998658, "learning_rate": 1.482332581566782e-05, "loss": 0.3672, "step": 1076 }, { "epoch": 0.3477137899027402, "grad_norm": 0.32548191016064204, "learning_rate": 1.4822749321457943e-05, "loss": 0.327, "step": 1077 }, { "epoch": 0.34803664393236206, "grad_norm": 0.5301794747134944, "learning_rate": 1.4822171899465812e-05, "loss": 0.3916, "step": 1078 }, { "epoch": 0.34835949796198395, "grad_norm": 0.45183250903737804, "learning_rate": 1.4821593549764584e-05, "loss": 0.2885, "step": 1079 }, { "epoch": 0.3486823519916058, "grad_norm": 0.39924054148320404, "learning_rate": 1.4821014272427537e-05, "loss": 0.3345, "step": 1080 }, { "epoch": 0.34900520602122764, "grad_norm": 0.4684216118369971, "learning_rate": 1.4820434067528063e-05, "loss": 0.3747, "step": 1081 }, { "epoch": 0.3493280600508495, "grad_norm": 0.40863721226794253, "learning_rate": 1.4819852935139673e-05, "loss": 0.3856, "step": 1082 }, { "epoch": 0.3496509140804714, "grad_norm": 0.54086354134877, "learning_rate": 1.4819270875335994e-05, "loss": 0.3353, "step": 1083 }, { "epoch": 0.34997376811009323, "grad_norm": 0.45084416092249846, "learning_rate": 1.481868788819078e-05, "loss": 0.3463, "step": 1084 }, { "epoch": 0.3502966221397151, "grad_norm": 0.48211327652214275, "learning_rate": 1.4818103973777885e-05, "loss": 0.4224, "step": 1085 }, { "epoch": 0.3506194761693369, "grad_norm": 0.7046494184911661, "learning_rate": 1.4817519132171294e-05, "loss": 0.3014, "step": 1086 }, { "epoch": 0.3509423301989588, "grad_norm": 0.5787575940498751, "learning_rate": 1.4816933363445106e-05, "loss": 0.3887, "step": 1087 }, { "epoch": 0.35126518422858066, "grad_norm": 0.34762149869010983, "learning_rate": 1.4816346667673538e-05, "loss": 0.4009, "step": 1088 }, { "epoch": 0.3515880382582025, "grad_norm": 0.39573724412865285, "learning_rate": 1.481575904493092e-05, "loss": 0.3592, "step": 1089 }, { "epoch": 0.35191089228782435, "grad_norm": 0.4161827808861479, "learning_rate": 1.4815170495291709e-05, "loss": 0.3558, "step": 1090 }, { "epoch": 0.35223374631744625, "grad_norm": 0.5259311458249045, "learning_rate": 1.4814581018830467e-05, "loss": 0.3327, "step": 1091 }, { "epoch": 0.3525566003470681, "grad_norm": 0.7142380402369445, "learning_rate": 1.4813990615621883e-05, "loss": 0.3296, "step": 1092 }, { "epoch": 0.35287945437668994, "grad_norm": 0.4405804096134938, "learning_rate": 1.4813399285740763e-05, "loss": 0.3971, "step": 1093 }, { "epoch": 0.3532023084063118, "grad_norm": 0.4781215508801049, "learning_rate": 1.481280702926202e-05, "loss": 0.319, "step": 1094 }, { "epoch": 0.3535251624359336, "grad_norm": 0.4578148408261532, "learning_rate": 1.4812213846260698e-05, "loss": 0.3601, "step": 1095 }, { "epoch": 0.3538480164655555, "grad_norm": 0.37169979509593615, "learning_rate": 1.4811619736811952e-05, "loss": 0.3631, "step": 1096 }, { "epoch": 0.35417087049517737, "grad_norm": 0.47151460853346416, "learning_rate": 1.4811024700991055e-05, "loss": 0.3792, "step": 1097 }, { "epoch": 0.3544937245247992, "grad_norm": 0.42401591688401913, "learning_rate": 1.4810428738873396e-05, "loss": 0.4094, "step": 1098 }, { "epoch": 0.35481657855442106, "grad_norm": 0.41358943176808904, "learning_rate": 1.4809831850534481e-05, "loss": 0.3656, "step": 1099 }, { "epoch": 0.35513943258404296, "grad_norm": 0.4564520872195094, "learning_rate": 1.4809234036049937e-05, "loss": 0.3416, "step": 1100 }, { "epoch": 0.3554622866136648, "grad_norm": 0.4044094520303478, "learning_rate": 1.4808635295495509e-05, "loss": 0.3189, "step": 1101 }, { "epoch": 0.35578514064328665, "grad_norm": 0.5912626918422568, "learning_rate": 1.4808035628947051e-05, "loss": 0.3144, "step": 1102 }, { "epoch": 0.3561079946729085, "grad_norm": 0.4609618283218117, "learning_rate": 1.4807435036480544e-05, "loss": 0.3289, "step": 1103 }, { "epoch": 0.3564308487025304, "grad_norm": 0.5468376735997236, "learning_rate": 1.4806833518172079e-05, "loss": 0.3398, "step": 1104 }, { "epoch": 0.35675370273215223, "grad_norm": 0.4202101898406944, "learning_rate": 1.4806231074097872e-05, "loss": 0.304, "step": 1105 }, { "epoch": 0.3570765567617741, "grad_norm": 0.42257124108496424, "learning_rate": 1.4805627704334247e-05, "loss": 0.359, "step": 1106 }, { "epoch": 0.3573994107913959, "grad_norm": 0.537009685797824, "learning_rate": 1.4805023408957654e-05, "loss": 0.3076, "step": 1107 }, { "epoch": 0.3577222648210178, "grad_norm": 0.4727573414649979, "learning_rate": 1.4804418188044653e-05, "loss": 0.3655, "step": 1108 }, { "epoch": 0.35804511885063967, "grad_norm": 0.5674934304982233, "learning_rate": 1.4803812041671927e-05, "loss": 0.338, "step": 1109 }, { "epoch": 0.3583679728802615, "grad_norm": 0.5189355363306899, "learning_rate": 1.4803204969916274e-05, "loss": 0.4148, "step": 1110 }, { "epoch": 0.35869082690988335, "grad_norm": 0.5352443726300071, "learning_rate": 1.4802596972854606e-05, "loss": 0.3785, "step": 1111 }, { "epoch": 0.35901368093950525, "grad_norm": 0.39794539949281893, "learning_rate": 1.4801988050563961e-05, "loss": 0.3983, "step": 1112 }, { "epoch": 0.3593365349691271, "grad_norm": 0.466376949063044, "learning_rate": 1.4801378203121482e-05, "loss": 0.376, "step": 1113 }, { "epoch": 0.35965938899874894, "grad_norm": 0.4314307719166095, "learning_rate": 1.4800767430604444e-05, "loss": 0.3622, "step": 1114 }, { "epoch": 0.3599822430283708, "grad_norm": 0.3979158701059166, "learning_rate": 1.4800155733090222e-05, "loss": 0.3739, "step": 1115 }, { "epoch": 0.36030509705799263, "grad_norm": 0.45775120995796376, "learning_rate": 1.4799543110656324e-05, "loss": 0.3195, "step": 1116 }, { "epoch": 0.36062795108761453, "grad_norm": 0.3402508207051887, "learning_rate": 1.4798929563380366e-05, "loss": 0.3526, "step": 1117 }, { "epoch": 0.3609508051172364, "grad_norm": 0.5131711027514856, "learning_rate": 1.4798315091340082e-05, "loss": 0.3992, "step": 1118 }, { "epoch": 0.3612736591468582, "grad_norm": 0.32883390142513774, "learning_rate": 1.4797699694613325e-05, "loss": 0.3418, "step": 1119 }, { "epoch": 0.36159651317648006, "grad_norm": 0.46245885588943864, "learning_rate": 1.479708337327807e-05, "loss": 0.3692, "step": 1120 }, { "epoch": 0.36191936720610196, "grad_norm": 0.4691960509656908, "learning_rate": 1.4796466127412398e-05, "loss": 0.3418, "step": 1121 }, { "epoch": 0.3622422212357238, "grad_norm": 0.5946413038200642, "learning_rate": 1.4795847957094516e-05, "loss": 0.33, "step": 1122 }, { "epoch": 0.36256507526534565, "grad_norm": 0.45727619838036176, "learning_rate": 1.4795228862402743e-05, "loss": 0.4427, "step": 1123 }, { "epoch": 0.3628879292949675, "grad_norm": 0.5437949643340739, "learning_rate": 1.4794608843415521e-05, "loss": 0.4322, "step": 1124 }, { "epoch": 0.3632107833245894, "grad_norm": 0.3709702589832622, "learning_rate": 1.4793987900211406e-05, "loss": 0.3195, "step": 1125 }, { "epoch": 0.36353363735421124, "grad_norm": 0.4075374624056025, "learning_rate": 1.4793366032869064e-05, "loss": 0.3469, "step": 1126 }, { "epoch": 0.3638564913838331, "grad_norm": 0.39148477873051435, "learning_rate": 1.4792743241467291e-05, "loss": 0.4192, "step": 1127 }, { "epoch": 0.3641793454134549, "grad_norm": 0.44620133420775315, "learning_rate": 1.4792119526084993e-05, "loss": 0.4043, "step": 1128 }, { "epoch": 0.3645021994430768, "grad_norm": 0.43428394894856237, "learning_rate": 1.4791494886801192e-05, "loss": 0.3921, "step": 1129 }, { "epoch": 0.36482505347269867, "grad_norm": 0.38905008420455983, "learning_rate": 1.4790869323695031e-05, "loss": 0.3813, "step": 1130 }, { "epoch": 0.3651479075023205, "grad_norm": 0.4332562924361852, "learning_rate": 1.4790242836845766e-05, "loss": 0.3501, "step": 1131 }, { "epoch": 0.36547076153194236, "grad_norm": 0.4733412506931274, "learning_rate": 1.4789615426332773e-05, "loss": 0.3427, "step": 1132 }, { "epoch": 0.3657936155615642, "grad_norm": 0.3942931162995598, "learning_rate": 1.4788987092235543e-05, "loss": 0.3595, "step": 1133 }, { "epoch": 0.3661164695911861, "grad_norm": 0.3880520420594162, "learning_rate": 1.4788357834633687e-05, "loss": 0.3698, "step": 1134 }, { "epoch": 0.36643932362080794, "grad_norm": 0.42156764344883557, "learning_rate": 1.4787727653606929e-05, "loss": 0.3438, "step": 1135 }, { "epoch": 0.3667621776504298, "grad_norm": 0.4572987826525825, "learning_rate": 1.4787096549235115e-05, "loss": 0.3971, "step": 1136 }, { "epoch": 0.36708503168005163, "grad_norm": 0.43012371260441357, "learning_rate": 1.4786464521598201e-05, "loss": 0.3945, "step": 1137 }, { "epoch": 0.36740788570967353, "grad_norm": 0.46121796398900167, "learning_rate": 1.4785831570776268e-05, "loss": 0.3079, "step": 1138 }, { "epoch": 0.3677307397392954, "grad_norm": 0.42906431470602063, "learning_rate": 1.4785197696849507e-05, "loss": 0.3705, "step": 1139 }, { "epoch": 0.3680535937689172, "grad_norm": 0.44804898174153274, "learning_rate": 1.478456289989823e-05, "loss": 0.3486, "step": 1140 }, { "epoch": 0.36837644779853906, "grad_norm": 0.4593844089918376, "learning_rate": 1.4783927180002865e-05, "loss": 0.3455, "step": 1141 }, { "epoch": 0.36869930182816096, "grad_norm": 0.4485396212718391, "learning_rate": 1.4783290537243958e-05, "loss": 0.3484, "step": 1142 }, { "epoch": 0.3690221558577828, "grad_norm": 0.5199590404184955, "learning_rate": 1.4782652971702169e-05, "loss": 0.374, "step": 1143 }, { "epoch": 0.36934500988740465, "grad_norm": 0.4522651130555708, "learning_rate": 1.4782014483458276e-05, "loss": 0.3433, "step": 1144 }, { "epoch": 0.3696678639170265, "grad_norm": 0.3705821348751655, "learning_rate": 1.4781375072593179e-05, "loss": 0.402, "step": 1145 }, { "epoch": 0.3699907179466484, "grad_norm": 0.4365440466833068, "learning_rate": 1.4780734739187886e-05, "loss": 0.4232, "step": 1146 }, { "epoch": 0.37031357197627024, "grad_norm": 0.4547629277921382, "learning_rate": 1.4780093483323525e-05, "loss": 0.3807, "step": 1147 }, { "epoch": 0.3706364260058921, "grad_norm": 0.46747564294851, "learning_rate": 1.4779451305081348e-05, "loss": 0.4213, "step": 1148 }, { "epoch": 0.3709592800355139, "grad_norm": 0.5251488104771299, "learning_rate": 1.4778808204542714e-05, "loss": 0.3726, "step": 1149 }, { "epoch": 0.3712821340651358, "grad_norm": 0.42958104173294454, "learning_rate": 1.4778164181789103e-05, "loss": 0.3703, "step": 1150 }, { "epoch": 0.37160498809475767, "grad_norm": 0.39328408619825006, "learning_rate": 1.4777519236902114e-05, "loss": 0.4107, "step": 1151 }, { "epoch": 0.3719278421243795, "grad_norm": 0.4683681903308352, "learning_rate": 1.477687336996346e-05, "loss": 0.3556, "step": 1152 }, { "epoch": 0.37225069615400136, "grad_norm": 0.422533179536325, "learning_rate": 1.477622658105497e-05, "loss": 0.3615, "step": 1153 }, { "epoch": 0.3725735501836232, "grad_norm": 0.3840570631977079, "learning_rate": 1.4775578870258592e-05, "loss": 0.3391, "step": 1154 }, { "epoch": 0.3728964042132451, "grad_norm": 0.5152537214277311, "learning_rate": 1.477493023765639e-05, "loss": 0.3908, "step": 1155 }, { "epoch": 0.37321925824286695, "grad_norm": 0.5131284830252699, "learning_rate": 1.4774280683330546e-05, "loss": 0.3715, "step": 1156 }, { "epoch": 0.3735421122724888, "grad_norm": 0.41594132510582166, "learning_rate": 1.4773630207363356e-05, "loss": 0.3917, "step": 1157 }, { "epoch": 0.37386496630211064, "grad_norm": 0.3936105812528523, "learning_rate": 1.4772978809837237e-05, "loss": 0.3931, "step": 1158 }, { "epoch": 0.37418782033173253, "grad_norm": 0.4943965350208832, "learning_rate": 1.4772326490834717e-05, "loss": 0.3828, "step": 1159 }, { "epoch": 0.3745106743613544, "grad_norm": 0.48131864930789414, "learning_rate": 1.4771673250438444e-05, "loss": 0.3802, "step": 1160 }, { "epoch": 0.3748335283909762, "grad_norm": 0.4196380713741386, "learning_rate": 1.4771019088731185e-05, "loss": 0.3416, "step": 1161 }, { "epoch": 0.37515638242059807, "grad_norm": 0.5701564222711673, "learning_rate": 1.477036400579582e-05, "loss": 0.3198, "step": 1162 }, { "epoch": 0.37547923645021997, "grad_norm": 0.6105030490444558, "learning_rate": 1.4769708001715347e-05, "loss": 0.373, "step": 1163 }, { "epoch": 0.3758020904798418, "grad_norm": 0.44658215211984226, "learning_rate": 1.4769051076572881e-05, "loss": 0.4135, "step": 1164 }, { "epoch": 0.37612494450946365, "grad_norm": 0.37155666114438307, "learning_rate": 1.4768393230451654e-05, "loss": 0.3787, "step": 1165 }, { "epoch": 0.3764477985390855, "grad_norm": 0.39756437185214355, "learning_rate": 1.4767734463435014e-05, "loss": 0.3678, "step": 1166 }, { "epoch": 0.3767706525687074, "grad_norm": 0.4143514123493714, "learning_rate": 1.4767074775606427e-05, "loss": 0.3317, "step": 1167 }, { "epoch": 0.37709350659832924, "grad_norm": 0.6855802863226698, "learning_rate": 1.4766414167049473e-05, "loss": 0.4167, "step": 1168 }, { "epoch": 0.3774163606279511, "grad_norm": 0.3551569770235824, "learning_rate": 1.4765752637847848e-05, "loss": 0.3709, "step": 1169 }, { "epoch": 0.37773921465757293, "grad_norm": 0.39884300179415944, "learning_rate": 1.4765090188085371e-05, "loss": 0.4097, "step": 1170 }, { "epoch": 0.3780620686871948, "grad_norm": 0.3855394641206456, "learning_rate": 1.4764426817845972e-05, "loss": 0.3619, "step": 1171 }, { "epoch": 0.3783849227168167, "grad_norm": 0.38506340746364814, "learning_rate": 1.4763762527213698e-05, "loss": 0.3384, "step": 1172 }, { "epoch": 0.3787077767464385, "grad_norm": 0.496155398668093, "learning_rate": 1.4763097316272717e-05, "loss": 0.349, "step": 1173 }, { "epoch": 0.37903063077606036, "grad_norm": 0.4060234374236021, "learning_rate": 1.4762431185107305e-05, "loss": 0.3614, "step": 1174 }, { "epoch": 0.3793534848056822, "grad_norm": 0.47063676367231766, "learning_rate": 1.4761764133801864e-05, "loss": 0.3073, "step": 1175 }, { "epoch": 0.3796763388353041, "grad_norm": 0.38021519067480725, "learning_rate": 1.4761096162440906e-05, "loss": 0.3097, "step": 1176 }, { "epoch": 0.37999919286492595, "grad_norm": 0.390211669873354, "learning_rate": 1.4760427271109067e-05, "loss": 0.3486, "step": 1177 }, { "epoch": 0.3803220468945478, "grad_norm": 0.4008102641230785, "learning_rate": 1.4759757459891087e-05, "loss": 0.4092, "step": 1178 }, { "epoch": 0.38064490092416964, "grad_norm": 0.46791319295610473, "learning_rate": 1.4759086728871837e-05, "loss": 0.3965, "step": 1179 }, { "epoch": 0.38096775495379154, "grad_norm": 0.4606774196181836, "learning_rate": 1.4758415078136293e-05, "loss": 0.3985, "step": 1180 }, { "epoch": 0.3812906089834134, "grad_norm": 0.5010090223431715, "learning_rate": 1.4757742507769553e-05, "loss": 0.3598, "step": 1181 }, { "epoch": 0.3816134630130352, "grad_norm": 0.4430524133939961, "learning_rate": 1.4757069017856833e-05, "loss": 0.3309, "step": 1182 }, { "epoch": 0.38193631704265707, "grad_norm": 0.3969408539483864, "learning_rate": 1.4756394608483463e-05, "loss": 0.4061, "step": 1183 }, { "epoch": 0.38225917107227897, "grad_norm": 0.48290265879023436, "learning_rate": 1.4755719279734888e-05, "loss": 0.3605, "step": 1184 }, { "epoch": 0.3825820251019008, "grad_norm": 0.6380120007219123, "learning_rate": 1.4755043031696674e-05, "loss": 0.3987, "step": 1185 }, { "epoch": 0.38290487913152266, "grad_norm": 0.3715332304454431, "learning_rate": 1.4754365864454495e-05, "loss": 0.3326, "step": 1186 }, { "epoch": 0.3832277331611445, "grad_norm": 0.4218978641519883, "learning_rate": 1.4753687778094152e-05, "loss": 0.3553, "step": 1187 }, { "epoch": 0.3835505871907664, "grad_norm": 0.5189436763553545, "learning_rate": 1.4753008772701559e-05, "loss": 0.357, "step": 1188 }, { "epoch": 0.38387344122038825, "grad_norm": 0.45164300913587, "learning_rate": 1.4752328848362739e-05, "loss": 0.4062, "step": 1189 }, { "epoch": 0.3841962952500101, "grad_norm": 0.4210089128322713, "learning_rate": 1.4751648005163842e-05, "loss": 0.2991, "step": 1190 }, { "epoch": 0.38451914927963193, "grad_norm": 0.5715693727124064, "learning_rate": 1.475096624319113e-05, "loss": 0.4212, "step": 1191 }, { "epoch": 0.3848420033092538, "grad_norm": 0.42065733415772083, "learning_rate": 1.4750283562530979e-05, "loss": 0.3936, "step": 1192 }, { "epoch": 0.3851648573388757, "grad_norm": 0.47631329546162027, "learning_rate": 1.4749599963269887e-05, "loss": 0.3666, "step": 1193 }, { "epoch": 0.3854877113684975, "grad_norm": 0.3878161517982719, "learning_rate": 1.4748915445494461e-05, "loss": 0.4254, "step": 1194 }, { "epoch": 0.38581056539811937, "grad_norm": 0.43745005840807444, "learning_rate": 1.4748230009291429e-05, "loss": 0.3853, "step": 1195 }, { "epoch": 0.3861334194277412, "grad_norm": 0.46553562117347996, "learning_rate": 1.4747543654747639e-05, "loss": 0.3585, "step": 1196 }, { "epoch": 0.3864562734573631, "grad_norm": 0.4955132452401203, "learning_rate": 1.4746856381950045e-05, "loss": 0.3355, "step": 1197 }, { "epoch": 0.38677912748698495, "grad_norm": 0.3866864472299744, "learning_rate": 1.474616819098573e-05, "loss": 0.3372, "step": 1198 }, { "epoch": 0.3871019815166068, "grad_norm": 0.49541647455004434, "learning_rate": 1.4745479081941883e-05, "loss": 0.3959, "step": 1199 }, { "epoch": 0.38742483554622864, "grad_norm": 0.3728067161868775, "learning_rate": 1.4744789054905814e-05, "loss": 0.3424, "step": 1200 }, { "epoch": 0.38774768957585054, "grad_norm": 0.5231997440422339, "learning_rate": 1.4744098109964949e-05, "loss": 0.3197, "step": 1201 }, { "epoch": 0.3880705436054724, "grad_norm": 0.5343681702747515, "learning_rate": 1.4743406247206828e-05, "loss": 0.4224, "step": 1202 }, { "epoch": 0.38839339763509423, "grad_norm": 0.4944331562375518, "learning_rate": 1.474271346671911e-05, "loss": 0.4368, "step": 1203 }, { "epoch": 0.3887162516647161, "grad_norm": 0.40627747412446424, "learning_rate": 1.474201976858957e-05, "loss": 0.3848, "step": 1204 }, { "epoch": 0.38903910569433797, "grad_norm": 0.49348337864219133, "learning_rate": 1.4741325152906099e-05, "loss": 0.4027, "step": 1205 }, { "epoch": 0.3893619597239598, "grad_norm": 0.5085453547196854, "learning_rate": 1.4740629619756704e-05, "loss": 0.3539, "step": 1206 }, { "epoch": 0.38968481375358166, "grad_norm": 0.44255932165710976, "learning_rate": 1.4739933169229505e-05, "loss": 0.3359, "step": 1207 }, { "epoch": 0.3900076677832035, "grad_norm": 0.5197300417280519, "learning_rate": 1.4739235801412746e-05, "loss": 0.3779, "step": 1208 }, { "epoch": 0.3903305218128254, "grad_norm": 0.442868945229025, "learning_rate": 1.4738537516394777e-05, "loss": 0.4148, "step": 1209 }, { "epoch": 0.39065337584244725, "grad_norm": 0.49181023794132767, "learning_rate": 1.4737838314264072e-05, "loss": 0.3803, "step": 1210 }, { "epoch": 0.3909762298720691, "grad_norm": 0.5353581672697766, "learning_rate": 1.4737138195109222e-05, "loss": 0.3792, "step": 1211 }, { "epoch": 0.39129908390169094, "grad_norm": 0.5452575021722887, "learning_rate": 1.473643715901893e-05, "loss": 0.3783, "step": 1212 }, { "epoch": 0.3916219379313128, "grad_norm": 0.4791139467078459, "learning_rate": 1.4735735206082013e-05, "loss": 0.3759, "step": 1213 }, { "epoch": 0.3919447919609347, "grad_norm": 0.388017827498977, "learning_rate": 1.4735032336387408e-05, "loss": 0.3205, "step": 1214 }, { "epoch": 0.3922676459905565, "grad_norm": 0.40290662633554125, "learning_rate": 1.4734328550024174e-05, "loss": 0.415, "step": 1215 }, { "epoch": 0.39259050002017837, "grad_norm": 0.521695582644787, "learning_rate": 1.473362384708147e-05, "loss": 0.3825, "step": 1216 }, { "epoch": 0.3929133540498002, "grad_norm": 0.583698756187949, "learning_rate": 1.473291822764859e-05, "loss": 0.3624, "step": 1217 }, { "epoch": 0.3932362080794221, "grad_norm": 0.441834169462515, "learning_rate": 1.4732211691814928e-05, "loss": 0.3605, "step": 1218 }, { "epoch": 0.39355906210904396, "grad_norm": 0.40184998924482235, "learning_rate": 1.4731504239670006e-05, "loss": 0.3306, "step": 1219 }, { "epoch": 0.3938819161386658, "grad_norm": 0.45667684879244236, "learning_rate": 1.4730795871303453e-05, "loss": 0.3201, "step": 1220 }, { "epoch": 0.39420477016828764, "grad_norm": 0.3902747894388157, "learning_rate": 1.4730086586805022e-05, "loss": 0.3574, "step": 1221 }, { "epoch": 0.39452762419790954, "grad_norm": 0.46177720084779866, "learning_rate": 1.4729376386264575e-05, "loss": 0.3493, "step": 1222 }, { "epoch": 0.3948504782275314, "grad_norm": 0.48660346423697887, "learning_rate": 1.4728665269772099e-05, "loss": 0.3232, "step": 1223 }, { "epoch": 0.39517333225715323, "grad_norm": 0.44103212021769717, "learning_rate": 1.4727953237417684e-05, "loss": 0.3743, "step": 1224 }, { "epoch": 0.3954961862867751, "grad_norm": 0.4655645422781623, "learning_rate": 1.4727240289291548e-05, "loss": 0.3543, "step": 1225 }, { "epoch": 0.395819040316397, "grad_norm": 0.4208147917483429, "learning_rate": 1.4726526425484021e-05, "loss": 0.3381, "step": 1226 }, { "epoch": 0.3961418943460188, "grad_norm": 0.48796104364428455, "learning_rate": 1.4725811646085548e-05, "loss": 0.4146, "step": 1227 }, { "epoch": 0.39646474837564066, "grad_norm": 0.6117827009294753, "learning_rate": 1.472509595118669e-05, "loss": 0.4176, "step": 1228 }, { "epoch": 0.3967876024052625, "grad_norm": 0.37874868893722097, "learning_rate": 1.4724379340878123e-05, "loss": 0.4445, "step": 1229 }, { "epoch": 0.39711045643488435, "grad_norm": 0.6781839954552541, "learning_rate": 1.4723661815250642e-05, "loss": 0.3454, "step": 1230 }, { "epoch": 0.39743331046450625, "grad_norm": 0.3962224365339185, "learning_rate": 1.4722943374395158e-05, "loss": 0.3516, "step": 1231 }, { "epoch": 0.3977561644941281, "grad_norm": 0.46099590136192403, "learning_rate": 1.4722224018402694e-05, "loss": 0.3338, "step": 1232 }, { "epoch": 0.39807901852374994, "grad_norm": 0.48790606503344364, "learning_rate": 1.4721503747364392e-05, "loss": 0.3393, "step": 1233 }, { "epoch": 0.3984018725533718, "grad_norm": 0.40800105593415054, "learning_rate": 1.4720782561371512e-05, "loss": 0.4407, "step": 1234 }, { "epoch": 0.3987247265829937, "grad_norm": 0.50476524625083, "learning_rate": 1.4720060460515424e-05, "loss": 0.3357, "step": 1235 }, { "epoch": 0.3990475806126155, "grad_norm": 0.40879485782685726, "learning_rate": 1.4719337444887617e-05, "loss": 0.4219, "step": 1236 }, { "epoch": 0.39937043464223737, "grad_norm": 0.4010103352632985, "learning_rate": 1.4718613514579699e-05, "loss": 0.3538, "step": 1237 }, { "epoch": 0.3996932886718592, "grad_norm": 0.47807486560395845, "learning_rate": 1.471788866968339e-05, "loss": 0.4003, "step": 1238 }, { "epoch": 0.4000161427014811, "grad_norm": 0.7415961025111301, "learning_rate": 1.4717162910290523e-05, "loss": 0.3371, "step": 1239 }, { "epoch": 0.40033899673110296, "grad_norm": 0.39909554418188287, "learning_rate": 1.4716436236493058e-05, "loss": 0.3819, "step": 1240 }, { "epoch": 0.4006618507607248, "grad_norm": 0.43028750841702174, "learning_rate": 1.4715708648383056e-05, "loss": 0.3939, "step": 1241 }, { "epoch": 0.40098470479034665, "grad_norm": 0.4557062629075772, "learning_rate": 1.4714980146052706e-05, "loss": 0.4419, "step": 1242 }, { "epoch": 0.40130755881996855, "grad_norm": 1.751873079723531, "learning_rate": 1.4714250729594306e-05, "loss": 0.4601, "step": 1243 }, { "epoch": 0.4016304128495904, "grad_norm": 0.4234811043930918, "learning_rate": 1.4713520399100275e-05, "loss": 0.351, "step": 1244 }, { "epoch": 0.40195326687921223, "grad_norm": 0.3788732384107695, "learning_rate": 1.4712789154663141e-05, "loss": 0.3576, "step": 1245 }, { "epoch": 0.4022761209088341, "grad_norm": 0.5688057911512775, "learning_rate": 1.4712056996375556e-05, "loss": 0.3451, "step": 1246 }, { "epoch": 0.402598974938456, "grad_norm": 0.8322251628727456, "learning_rate": 1.4711323924330281e-05, "loss": 0.3197, "step": 1247 }, { "epoch": 0.4029218289680778, "grad_norm": 0.5315718521179439, "learning_rate": 1.4710589938620193e-05, "loss": 0.3472, "step": 1248 }, { "epoch": 0.40324468299769967, "grad_norm": 0.39967000885342285, "learning_rate": 1.4709855039338292e-05, "loss": 0.4502, "step": 1249 }, { "epoch": 0.4035675370273215, "grad_norm": 0.4052581334473416, "learning_rate": 1.4709119226577686e-05, "loss": 0.3464, "step": 1250 }, { "epoch": 0.40389039105694335, "grad_norm": 0.5601278682739383, "learning_rate": 1.47083825004316e-05, "loss": 0.3983, "step": 1251 }, { "epoch": 0.40421324508656525, "grad_norm": 0.39932795286850786, "learning_rate": 1.4707644860993378e-05, "loss": 0.4023, "step": 1252 }, { "epoch": 0.4045360991161871, "grad_norm": 0.4829666660368531, "learning_rate": 1.4706906308356479e-05, "loss": 0.3078, "step": 1253 }, { "epoch": 0.40485895314580894, "grad_norm": 0.5234700929974225, "learning_rate": 1.4706166842614473e-05, "loss": 0.4249, "step": 1254 }, { "epoch": 0.4051818071754308, "grad_norm": 0.47032487050373206, "learning_rate": 1.4705426463861055e-05, "loss": 0.3441, "step": 1255 }, { "epoch": 0.4055046612050527, "grad_norm": 0.7501383683116989, "learning_rate": 1.4704685172190023e-05, "loss": 0.3681, "step": 1256 }, { "epoch": 0.40582751523467453, "grad_norm": 0.48512318997596143, "learning_rate": 1.4703942967695303e-05, "loss": 0.3497, "step": 1257 }, { "epoch": 0.4061503692642964, "grad_norm": 0.4457226098231372, "learning_rate": 1.470319985047093e-05, "loss": 0.3649, "step": 1258 }, { "epoch": 0.4064732232939182, "grad_norm": 0.4345390342437068, "learning_rate": 1.4702455820611053e-05, "loss": 0.3324, "step": 1259 }, { "epoch": 0.4067960773235401, "grad_norm": 0.5165482522204797, "learning_rate": 1.4701710878209942e-05, "loss": 0.403, "step": 1260 }, { "epoch": 0.40711893135316196, "grad_norm": 0.4767636411792692, "learning_rate": 1.4700965023361981e-05, "loss": 0.4458, "step": 1261 }, { "epoch": 0.4074417853827838, "grad_norm": 0.44345233236312054, "learning_rate": 1.4700218256161667e-05, "loss": 0.3458, "step": 1262 }, { "epoch": 0.40776463941240565, "grad_norm": 0.3801174769303794, "learning_rate": 1.4699470576703617e-05, "loss": 0.3182, "step": 1263 }, { "epoch": 0.40808749344202755, "grad_norm": 0.41546276731735804, "learning_rate": 1.469872198508256e-05, "loss": 0.3945, "step": 1264 }, { "epoch": 0.4084103474716494, "grad_norm": 0.45272167384121215, "learning_rate": 1.4697972481393338e-05, "loss": 0.3389, "step": 1265 }, { "epoch": 0.40873320150127124, "grad_norm": 0.6576745903245097, "learning_rate": 1.4697222065730916e-05, "loss": 0.3743, "step": 1266 }, { "epoch": 0.4090560555308931, "grad_norm": 0.36272036078014486, "learning_rate": 1.4696470738190369e-05, "loss": 0.3877, "step": 1267 }, { "epoch": 0.4093789095605149, "grad_norm": 0.4782326769760276, "learning_rate": 1.469571849886689e-05, "loss": 0.355, "step": 1268 }, { "epoch": 0.4097017635901368, "grad_norm": 0.4479175006361012, "learning_rate": 1.4694965347855786e-05, "loss": 0.3444, "step": 1269 }, { "epoch": 0.41002461761975867, "grad_norm": 0.38521524081697417, "learning_rate": 1.4694211285252481e-05, "loss": 0.3242, "step": 1270 }, { "epoch": 0.4103474716493805, "grad_norm": 0.4392922726232488, "learning_rate": 1.4693456311152512e-05, "loss": 0.3425, "step": 1271 }, { "epoch": 0.41067032567900236, "grad_norm": 0.43126243177718665, "learning_rate": 1.4692700425651536e-05, "loss": 0.3666, "step": 1272 }, { "epoch": 0.41099317970862426, "grad_norm": 0.4413031107071449, "learning_rate": 1.469194362884532e-05, "loss": 0.4146, "step": 1273 }, { "epoch": 0.4113160337382461, "grad_norm": 0.4812777587808371, "learning_rate": 1.469118592082975e-05, "loss": 0.321, "step": 1274 }, { "epoch": 0.41163888776786794, "grad_norm": 0.5411155736190102, "learning_rate": 1.4690427301700828e-05, "loss": 0.3494, "step": 1275 }, { "epoch": 0.4119617417974898, "grad_norm": 0.35379335127298256, "learning_rate": 1.4689667771554667e-05, "loss": 0.3178, "step": 1276 }, { "epoch": 0.4122845958271117, "grad_norm": 0.4518269583526312, "learning_rate": 1.46889073304875e-05, "loss": 0.448, "step": 1277 }, { "epoch": 0.41260744985673353, "grad_norm": 0.6060531829753097, "learning_rate": 1.4688145978595676e-05, "loss": 0.3448, "step": 1278 }, { "epoch": 0.4129303038863554, "grad_norm": 0.44059935846269704, "learning_rate": 1.4687383715975655e-05, "loss": 0.3635, "step": 1279 }, { "epoch": 0.4132531579159772, "grad_norm": 0.3973143132318481, "learning_rate": 1.4686620542724014e-05, "loss": 0.31, "step": 1280 }, { "epoch": 0.4135760119455991, "grad_norm": 0.4821478237268171, "learning_rate": 1.4685856458937446e-05, "loss": 0.4135, "step": 1281 }, { "epoch": 0.41389886597522096, "grad_norm": 0.39401996848226506, "learning_rate": 1.4685091464712761e-05, "loss": 0.3818, "step": 1282 }, { "epoch": 0.4142217200048428, "grad_norm": 0.514080934054679, "learning_rate": 1.4684325560146882e-05, "loss": 0.3636, "step": 1283 }, { "epoch": 0.41454457403446465, "grad_norm": 0.4939423309466608, "learning_rate": 1.4683558745336846e-05, "loss": 0.4007, "step": 1284 }, { "epoch": 0.41486742806408655, "grad_norm": 0.4526769925160539, "learning_rate": 1.4682791020379812e-05, "loss": 0.3993, "step": 1285 }, { "epoch": 0.4151902820937084, "grad_norm": 0.44198089395746026, "learning_rate": 1.4682022385373046e-05, "loss": 0.3482, "step": 1286 }, { "epoch": 0.41551313612333024, "grad_norm": 0.42570329802847895, "learning_rate": 1.4681252840413934e-05, "loss": 0.3808, "step": 1287 }, { "epoch": 0.4158359901529521, "grad_norm": 0.3661973265985823, "learning_rate": 1.4680482385599976e-05, "loss": 0.3662, "step": 1288 }, { "epoch": 0.41615884418257393, "grad_norm": 0.888868259323883, "learning_rate": 1.4679711021028787e-05, "loss": 0.3986, "step": 1289 }, { "epoch": 0.41648169821219583, "grad_norm": 0.5211936821184545, "learning_rate": 1.46789387467981e-05, "loss": 0.4031, "step": 1290 }, { "epoch": 0.41680455224181767, "grad_norm": 0.5101909490072315, "learning_rate": 1.4678165563005762e-05, "loss": 0.3921, "step": 1291 }, { "epoch": 0.4171274062714395, "grad_norm": 0.5051940333092272, "learning_rate": 1.4677391469749727e-05, "loss": 0.3901, "step": 1292 }, { "epoch": 0.41745026030106136, "grad_norm": 0.5575565043250325, "learning_rate": 1.4676616467128079e-05, "loss": 0.3004, "step": 1293 }, { "epoch": 0.41777311433068326, "grad_norm": 0.49659597636192165, "learning_rate": 1.4675840555239008e-05, "loss": 0.3848, "step": 1294 }, { "epoch": 0.4180959683603051, "grad_norm": 0.4172496928641187, "learning_rate": 1.467506373418082e-05, "loss": 0.4545, "step": 1295 }, { "epoch": 0.41841882238992695, "grad_norm": 0.4444890196947441, "learning_rate": 1.4674286004051937e-05, "loss": 0.3783, "step": 1296 }, { "epoch": 0.4187416764195488, "grad_norm": 0.4345746615512859, "learning_rate": 1.4673507364950898e-05, "loss": 0.3539, "step": 1297 }, { "epoch": 0.4190645304491707, "grad_norm": 0.5264904924844704, "learning_rate": 1.4672727816976354e-05, "loss": 0.4318, "step": 1298 }, { "epoch": 0.41938738447879254, "grad_norm": 0.45842335367085635, "learning_rate": 1.4671947360227073e-05, "loss": 0.3489, "step": 1299 }, { "epoch": 0.4197102385084144, "grad_norm": 0.4150523547389566, "learning_rate": 1.4671165994801936e-05, "loss": 0.3596, "step": 1300 }, { "epoch": 0.4200330925380362, "grad_norm": 0.9341911236829274, "learning_rate": 1.4670383720799945e-05, "loss": 0.3856, "step": 1301 }, { "epoch": 0.4203559465676581, "grad_norm": 0.41175849410359183, "learning_rate": 1.4669600538320212e-05, "loss": 0.394, "step": 1302 }, { "epoch": 0.42067880059727997, "grad_norm": 0.3971297298496757, "learning_rate": 1.4668816447461961e-05, "loss": 0.3745, "step": 1303 }, { "epoch": 0.4210016546269018, "grad_norm": 1.3815871251346294, "learning_rate": 1.4668031448324538e-05, "loss": 0.4048, "step": 1304 }, { "epoch": 0.42132450865652366, "grad_norm": 0.4052918314597232, "learning_rate": 1.4667245541007403e-05, "loss": 0.3903, "step": 1305 }, { "epoch": 0.4216473626861455, "grad_norm": 0.5287172738725341, "learning_rate": 1.4666458725610128e-05, "loss": 0.3257, "step": 1306 }, { "epoch": 0.4219702167157674, "grad_norm": 0.3532304017844352, "learning_rate": 1.4665671002232401e-05, "loss": 0.3052, "step": 1307 }, { "epoch": 0.42229307074538924, "grad_norm": 0.44142463023953, "learning_rate": 1.4664882370974025e-05, "loss": 0.3205, "step": 1308 }, { "epoch": 0.4226159247750111, "grad_norm": 0.4454167329638632, "learning_rate": 1.466409283193492e-05, "loss": 0.3312, "step": 1309 }, { "epoch": 0.42293877880463293, "grad_norm": 0.42874295275421526, "learning_rate": 1.4663302385215119e-05, "loss": 0.3651, "step": 1310 }, { "epoch": 0.42326163283425483, "grad_norm": 0.4074184242968335, "learning_rate": 1.4662511030914769e-05, "loss": 0.3458, "step": 1311 }, { "epoch": 0.4235844868638767, "grad_norm": 0.4868203476313854, "learning_rate": 1.4661718769134136e-05, "loss": 0.2984, "step": 1312 }, { "epoch": 0.4239073408934985, "grad_norm": 0.3925638126734977, "learning_rate": 1.4660925599973598e-05, "loss": 0.3368, "step": 1313 }, { "epoch": 0.42423019492312036, "grad_norm": 0.4410647137643419, "learning_rate": 1.4660131523533645e-05, "loss": 0.3953, "step": 1314 }, { "epoch": 0.42455304895274226, "grad_norm": 0.42159666056562145, "learning_rate": 1.465933653991489e-05, "loss": 0.4071, "step": 1315 }, { "epoch": 0.4248759029823641, "grad_norm": 0.4615029606142835, "learning_rate": 1.4658540649218053e-05, "loss": 0.3231, "step": 1316 }, { "epoch": 0.42519875701198595, "grad_norm": 0.978805310295166, "learning_rate": 1.4657743851543975e-05, "loss": 0.3597, "step": 1317 }, { "epoch": 0.4255216110416078, "grad_norm": 0.4607913967392723, "learning_rate": 1.4656946146993607e-05, "loss": 0.3594, "step": 1318 }, { "epoch": 0.4258444650712297, "grad_norm": 0.4231495224093402, "learning_rate": 1.4656147535668017e-05, "loss": 0.3688, "step": 1319 }, { "epoch": 0.42616731910085154, "grad_norm": 0.42741528624421843, "learning_rate": 1.4655348017668392e-05, "loss": 0.369, "step": 1320 }, { "epoch": 0.4264901731304734, "grad_norm": 0.4105340004934761, "learning_rate": 1.4654547593096024e-05, "loss": 0.4004, "step": 1321 }, { "epoch": 0.4268130271600952, "grad_norm": 0.4181035249707088, "learning_rate": 1.4653746262052328e-05, "loss": 0.3374, "step": 1322 }, { "epoch": 0.4271358811897171, "grad_norm": 0.38492038287415187, "learning_rate": 1.4652944024638833e-05, "loss": 0.3896, "step": 1323 }, { "epoch": 0.42745873521933897, "grad_norm": 0.4845973795174071, "learning_rate": 1.4652140880957182e-05, "loss": 0.3583, "step": 1324 }, { "epoch": 0.4277815892489608, "grad_norm": 0.41311476701574573, "learning_rate": 1.4651336831109126e-05, "loss": 0.3297, "step": 1325 }, { "epoch": 0.42810444327858266, "grad_norm": 0.4245310187053993, "learning_rate": 1.4650531875196547e-05, "loss": 0.3438, "step": 1326 }, { "epoch": 0.4284272973082045, "grad_norm": 0.4853505071745879, "learning_rate": 1.4649726013321422e-05, "loss": 0.3796, "step": 1327 }, { "epoch": 0.4287501513378264, "grad_norm": 0.42282552689788305, "learning_rate": 1.4648919245585858e-05, "loss": 0.3617, "step": 1328 }, { "epoch": 0.42907300536744825, "grad_norm": 0.4199843467333367, "learning_rate": 1.4648111572092071e-05, "loss": 0.3706, "step": 1329 }, { "epoch": 0.4293958593970701, "grad_norm": 0.478382071972772, "learning_rate": 1.4647302992942392e-05, "loss": 0.3433, "step": 1330 }, { "epoch": 0.42971871342669193, "grad_norm": 0.43974062856350504, "learning_rate": 1.4646493508239264e-05, "loss": 0.3886, "step": 1331 }, { "epoch": 0.43004156745631383, "grad_norm": 0.5205402326584826, "learning_rate": 1.464568311808525e-05, "loss": 0.333, "step": 1332 }, { "epoch": 0.4303644214859357, "grad_norm": 0.497909475293942, "learning_rate": 1.4644871822583026e-05, "loss": 0.3668, "step": 1333 }, { "epoch": 0.4306872755155575, "grad_norm": 0.5252967272835688, "learning_rate": 1.4644059621835381e-05, "loss": 0.3533, "step": 1334 }, { "epoch": 0.43101012954517937, "grad_norm": 0.35063397735442825, "learning_rate": 1.464324651594522e-05, "loss": 0.3741, "step": 1335 }, { "epoch": 0.43133298357480127, "grad_norm": 0.37721411745477307, "learning_rate": 1.4642432505015563e-05, "loss": 0.3108, "step": 1336 }, { "epoch": 0.4316558376044231, "grad_norm": 0.4737431251652168, "learning_rate": 1.4641617589149544e-05, "loss": 0.3897, "step": 1337 }, { "epoch": 0.43197869163404495, "grad_norm": 0.43127914288191604, "learning_rate": 1.4640801768450407e-05, "loss": 0.4752, "step": 1338 }, { "epoch": 0.4323015456636668, "grad_norm": 0.45010632361104896, "learning_rate": 1.4639985043021524e-05, "loss": 0.3405, "step": 1339 }, { "epoch": 0.4326243996932887, "grad_norm": 0.4502327061326408, "learning_rate": 1.4639167412966367e-05, "loss": 0.362, "step": 1340 }, { "epoch": 0.43294725372291054, "grad_norm": 0.38975356768209335, "learning_rate": 1.463834887838853e-05, "loss": 0.3541, "step": 1341 }, { "epoch": 0.4332701077525324, "grad_norm": 0.3789892804155382, "learning_rate": 1.4637529439391719e-05, "loss": 0.333, "step": 1342 }, { "epoch": 0.43359296178215423, "grad_norm": 0.5092047854060866, "learning_rate": 1.4636709096079758e-05, "loss": 0.3755, "step": 1343 }, { "epoch": 0.4339158158117761, "grad_norm": 0.4171211919301384, "learning_rate": 1.4635887848556581e-05, "loss": 0.3175, "step": 1344 }, { "epoch": 0.434238669841398, "grad_norm": 0.40053091377992733, "learning_rate": 1.4635065696926241e-05, "loss": 0.3971, "step": 1345 }, { "epoch": 0.4345615238710198, "grad_norm": 0.41463786680623865, "learning_rate": 1.4634242641292904e-05, "loss": 0.4195, "step": 1346 }, { "epoch": 0.43488437790064166, "grad_norm": 0.4536162098799491, "learning_rate": 1.4633418681760848e-05, "loss": 0.3464, "step": 1347 }, { "epoch": 0.4352072319302635, "grad_norm": 0.4402462484145565, "learning_rate": 1.4632593818434469e-05, "loss": 0.3961, "step": 1348 }, { "epoch": 0.4355300859598854, "grad_norm": 0.48530008838969785, "learning_rate": 1.4631768051418273e-05, "loss": 0.3674, "step": 1349 }, { "epoch": 0.43585293998950725, "grad_norm": 0.4986301192749785, "learning_rate": 1.4630941380816887e-05, "loss": 0.4118, "step": 1350 }, { "epoch": 0.4361757940191291, "grad_norm": 0.5492983293680966, "learning_rate": 1.463011380673505e-05, "loss": 0.4689, "step": 1351 }, { "epoch": 0.43649864804875094, "grad_norm": 0.413977717030732, "learning_rate": 1.462928532927761e-05, "loss": 0.3498, "step": 1352 }, { "epoch": 0.43682150207837284, "grad_norm": 0.47161739630202576, "learning_rate": 1.4628455948549537e-05, "loss": 0.3146, "step": 1353 }, { "epoch": 0.4371443561079947, "grad_norm": 0.39807955912588416, "learning_rate": 1.4627625664655912e-05, "loss": 0.3802, "step": 1354 }, { "epoch": 0.4374672101376165, "grad_norm": 0.43769292346219574, "learning_rate": 1.462679447770193e-05, "loss": 0.3363, "step": 1355 }, { "epoch": 0.43779006416723837, "grad_norm": 0.36113555995077573, "learning_rate": 1.4625962387792903e-05, "loss": 0.3063, "step": 1356 }, { "epoch": 0.43811291819686027, "grad_norm": 0.4134175921605896, "learning_rate": 1.4625129395034255e-05, "loss": 0.3395, "step": 1357 }, { "epoch": 0.4384357722264821, "grad_norm": 0.5637750286900809, "learning_rate": 1.4624295499531522e-05, "loss": 0.4033, "step": 1358 }, { "epoch": 0.43875862625610396, "grad_norm": 0.46086658945073244, "learning_rate": 1.4623460701390362e-05, "loss": 0.412, "step": 1359 }, { "epoch": 0.4390814802857258, "grad_norm": 0.506221401996436, "learning_rate": 1.462262500071654e-05, "loss": 0.3328, "step": 1360 }, { "epoch": 0.4394043343153477, "grad_norm": 0.4532652603149888, "learning_rate": 1.462178839761594e-05, "loss": 0.3075, "step": 1361 }, { "epoch": 0.43972718834496954, "grad_norm": 0.422095884928424, "learning_rate": 1.4620950892194556e-05, "loss": 0.3593, "step": 1362 }, { "epoch": 0.4400500423745914, "grad_norm": 0.43574526783444933, "learning_rate": 1.46201124845585e-05, "loss": 0.3048, "step": 1363 }, { "epoch": 0.44037289640421323, "grad_norm": 0.4335486299067754, "learning_rate": 1.4619273174813999e-05, "loss": 0.2947, "step": 1364 }, { "epoch": 0.4406957504338351, "grad_norm": 0.624190981055566, "learning_rate": 1.4618432963067387e-05, "loss": 0.3648, "step": 1365 }, { "epoch": 0.441018604463457, "grad_norm": 0.4635493697955835, "learning_rate": 1.4617591849425125e-05, "loss": 0.4173, "step": 1366 }, { "epoch": 0.4413414584930788, "grad_norm": 0.3486338874863413, "learning_rate": 1.4616749833993775e-05, "loss": 0.3555, "step": 1367 }, { "epoch": 0.44166431252270066, "grad_norm": 0.4201880908355821, "learning_rate": 1.4615906916880023e-05, "loss": 0.3696, "step": 1368 }, { "epoch": 0.4419871665523225, "grad_norm": 0.43762027642862106, "learning_rate": 1.4615063098190662e-05, "loss": 0.3274, "step": 1369 }, { "epoch": 0.4423100205819444, "grad_norm": 0.39315733406287545, "learning_rate": 1.4614218378032606e-05, "loss": 0.2895, "step": 1370 }, { "epoch": 0.44263287461156625, "grad_norm": 1.594249192555923, "learning_rate": 1.4613372756512877e-05, "loss": 0.306, "step": 1371 }, { "epoch": 0.4429557286411881, "grad_norm": 0.5056940689335465, "learning_rate": 1.4612526233738617e-05, "loss": 0.3816, "step": 1372 }, { "epoch": 0.44327858267080994, "grad_norm": 0.37991172098965964, "learning_rate": 1.4611678809817077e-05, "loss": 0.4045, "step": 1373 }, { "epoch": 0.44360143670043184, "grad_norm": 0.5717774819221988, "learning_rate": 1.4610830484855624e-05, "loss": 0.3377, "step": 1374 }, { "epoch": 0.4439242907300537, "grad_norm": 0.4268006180638915, "learning_rate": 1.4609981258961742e-05, "loss": 0.4468, "step": 1375 }, { "epoch": 0.4442471447596755, "grad_norm": 0.39277471626477367, "learning_rate": 1.4609131132243026e-05, "loss": 0.3778, "step": 1376 }, { "epoch": 0.44456999878929737, "grad_norm": 0.4770579603026673, "learning_rate": 1.4608280104807184e-05, "loss": 0.3941, "step": 1377 }, { "epoch": 0.44489285281891927, "grad_norm": 0.3755112704196099, "learning_rate": 1.4607428176762044e-05, "loss": 0.3723, "step": 1378 }, { "epoch": 0.4452157068485411, "grad_norm": 0.5122556292349812, "learning_rate": 1.4606575348215538e-05, "loss": 0.341, "step": 1379 }, { "epoch": 0.44553856087816296, "grad_norm": 0.5557933790135956, "learning_rate": 1.4605721619275722e-05, "loss": 0.423, "step": 1380 }, { "epoch": 0.4458614149077848, "grad_norm": 0.37344322653190287, "learning_rate": 1.4604866990050766e-05, "loss": 0.4057, "step": 1381 }, { "epoch": 0.44618426893740665, "grad_norm": 0.4631654223145528, "learning_rate": 1.4604011460648942e-05, "loss": 0.3847, "step": 1382 }, { "epoch": 0.44650712296702855, "grad_norm": 0.7561579693830367, "learning_rate": 1.4603155031178652e-05, "loss": 0.3658, "step": 1383 }, { "epoch": 0.4468299769966504, "grad_norm": 0.44335255406910523, "learning_rate": 1.4602297701748398e-05, "loss": 0.3719, "step": 1384 }, { "epoch": 0.44715283102627223, "grad_norm": 0.4301171247164182, "learning_rate": 1.4601439472466808e-05, "loss": 0.3142, "step": 1385 }, { "epoch": 0.4474756850558941, "grad_norm": 0.4890676963149657, "learning_rate": 1.4600580343442614e-05, "loss": 0.4305, "step": 1386 }, { "epoch": 0.447798539085516, "grad_norm": 0.44213144176112673, "learning_rate": 1.459972031478467e-05, "loss": 0.3451, "step": 1387 }, { "epoch": 0.4481213931151378, "grad_norm": 0.4159356638400888, "learning_rate": 1.4598859386601939e-05, "loss": 0.3689, "step": 1388 }, { "epoch": 0.44844424714475967, "grad_norm": 0.3884008865959295, "learning_rate": 1.45979975590035e-05, "loss": 0.2803, "step": 1389 }, { "epoch": 0.4487671011743815, "grad_norm": 0.43783144201567525, "learning_rate": 1.4597134832098544e-05, "loss": 0.3728, "step": 1390 }, { "epoch": 0.4490899552040034, "grad_norm": 0.3327940964610978, "learning_rate": 1.4596271205996378e-05, "loss": 0.3644, "step": 1391 }, { "epoch": 0.44941280923362525, "grad_norm": 0.565715652980787, "learning_rate": 1.4595406680806422e-05, "loss": 0.4065, "step": 1392 }, { "epoch": 0.4497356632632471, "grad_norm": 0.4183763136741931, "learning_rate": 1.4594541256638213e-05, "loss": 0.4093, "step": 1393 }, { "epoch": 0.45005851729286894, "grad_norm": 0.5081248689883546, "learning_rate": 1.4593674933601394e-05, "loss": 0.3053, "step": 1394 }, { "epoch": 0.45038137132249084, "grad_norm": 0.6462135385699149, "learning_rate": 1.459280771180573e-05, "loss": 0.3351, "step": 1395 }, { "epoch": 0.4507042253521127, "grad_norm": 0.3772413076880965, "learning_rate": 1.4591939591361097e-05, "loss": 0.3855, "step": 1396 }, { "epoch": 0.45102707938173453, "grad_norm": 0.4522698559192069, "learning_rate": 1.4591070572377484e-05, "loss": 0.339, "step": 1397 }, { "epoch": 0.4513499334113564, "grad_norm": 0.4163776272903024, "learning_rate": 1.4590200654964992e-05, "loss": 0.393, "step": 1398 }, { "epoch": 0.4516727874409783, "grad_norm": 0.4256839588893144, "learning_rate": 1.4589329839233844e-05, "loss": 0.4072, "step": 1399 }, { "epoch": 0.4519956414706001, "grad_norm": 0.4892500340609571, "learning_rate": 1.4588458125294367e-05, "loss": 0.3626, "step": 1400 }, { "epoch": 0.45231849550022196, "grad_norm": 0.6744072543127138, "learning_rate": 1.4587585513257007e-05, "loss": 0.3682, "step": 1401 }, { "epoch": 0.4526413495298438, "grad_norm": 0.6824295205964991, "learning_rate": 1.4586712003232322e-05, "loss": 0.3416, "step": 1402 }, { "epoch": 0.45296420355946565, "grad_norm": 0.44514009589962694, "learning_rate": 1.4585837595330984e-05, "loss": 0.3666, "step": 1403 }, { "epoch": 0.45328705758908755, "grad_norm": 0.4306773589895992, "learning_rate": 1.4584962289663782e-05, "loss": 0.3546, "step": 1404 }, { "epoch": 0.4536099116187094, "grad_norm": 0.4894928622389755, "learning_rate": 1.4584086086341616e-05, "loss": 0.4216, "step": 1405 }, { "epoch": 0.45393276564833124, "grad_norm": 0.44863299257649497, "learning_rate": 1.4583208985475495e-05, "loss": 0.3884, "step": 1406 }, { "epoch": 0.4542556196779531, "grad_norm": 0.5143090714971315, "learning_rate": 1.4582330987176552e-05, "loss": 0.3543, "step": 1407 }, { "epoch": 0.454578473707575, "grad_norm": 0.5667963307997337, "learning_rate": 1.4581452091556025e-05, "loss": 0.3617, "step": 1408 }, { "epoch": 0.4549013277371968, "grad_norm": 0.47562764107773087, "learning_rate": 1.458057229872527e-05, "loss": 0.4017, "step": 1409 }, { "epoch": 0.45522418176681867, "grad_norm": 0.5315342923667485, "learning_rate": 1.4579691608795755e-05, "loss": 0.4672, "step": 1410 }, { "epoch": 0.4555470357964405, "grad_norm": 0.36135451060797513, "learning_rate": 1.4578810021879062e-05, "loss": 0.3727, "step": 1411 }, { "epoch": 0.4558698898260624, "grad_norm": 0.48025024287337725, "learning_rate": 1.4577927538086887e-05, "loss": 0.415, "step": 1412 }, { "epoch": 0.45619274385568426, "grad_norm": 0.38263038854939846, "learning_rate": 1.4577044157531042e-05, "loss": 0.3983, "step": 1413 }, { "epoch": 0.4565155978853061, "grad_norm": 0.40115555766446154, "learning_rate": 1.4576159880323447e-05, "loss": 0.3648, "step": 1414 }, { "epoch": 0.45683845191492795, "grad_norm": 0.4179299894830692, "learning_rate": 1.457527470657614e-05, "loss": 0.3971, "step": 1415 }, { "epoch": 0.45716130594454985, "grad_norm": 0.4822779639048555, "learning_rate": 1.4574388636401269e-05, "loss": 0.3885, "step": 1416 }, { "epoch": 0.4574841599741717, "grad_norm": 0.36700837240074574, "learning_rate": 1.45735016699111e-05, "loss": 0.3276, "step": 1417 }, { "epoch": 0.45780701400379353, "grad_norm": 0.8707575964224313, "learning_rate": 1.4572613807218012e-05, "loss": 0.3586, "step": 1418 }, { "epoch": 0.4581298680334154, "grad_norm": 1.5521345204474508, "learning_rate": 1.4571725048434494e-05, "loss": 0.3617, "step": 1419 }, { "epoch": 0.4584527220630373, "grad_norm": 0.4228198524958448, "learning_rate": 1.4570835393673148e-05, "loss": 0.4386, "step": 1420 }, { "epoch": 0.4587755760926591, "grad_norm": 0.457685272007366, "learning_rate": 1.45699448430467e-05, "loss": 0.3646, "step": 1421 }, { "epoch": 0.45909843012228096, "grad_norm": 0.5686400331673758, "learning_rate": 1.4569053396667971e-05, "loss": 0.3621, "step": 1422 }, { "epoch": 0.4594212841519028, "grad_norm": 0.4665163651240504, "learning_rate": 1.4568161054649914e-05, "loss": 0.4026, "step": 1423 }, { "epoch": 0.45974413818152465, "grad_norm": 0.7367560322248884, "learning_rate": 1.4567267817105584e-05, "loss": 0.3658, "step": 1424 }, { "epoch": 0.46006699221114655, "grad_norm": 0.7093007430006749, "learning_rate": 1.4566373684148154e-05, "loss": 0.3683, "step": 1425 }, { "epoch": 0.4603898462407684, "grad_norm": 0.509769945396212, "learning_rate": 1.4565478655890909e-05, "loss": 0.3805, "step": 1426 }, { "epoch": 0.46071270027039024, "grad_norm": 0.4739162179510099, "learning_rate": 1.4564582732447249e-05, "loss": 0.4115, "step": 1427 }, { "epoch": 0.4610355543000121, "grad_norm": 0.454200998902494, "learning_rate": 1.4563685913930686e-05, "loss": 0.3268, "step": 1428 }, { "epoch": 0.461358408329634, "grad_norm": 0.38496439825912626, "learning_rate": 1.4562788200454842e-05, "loss": 0.3667, "step": 1429 }, { "epoch": 0.46168126235925583, "grad_norm": 0.458416758866017, "learning_rate": 1.456188959213346e-05, "loss": 0.4025, "step": 1430 }, { "epoch": 0.4620041163888777, "grad_norm": 0.40386564859583646, "learning_rate": 1.4560990089080394e-05, "loss": 0.3433, "step": 1431 }, { "epoch": 0.4623269704184995, "grad_norm": 0.5137136071761582, "learning_rate": 1.4560089691409605e-05, "loss": 0.3658, "step": 1432 }, { "epoch": 0.4626498244481214, "grad_norm": 0.4448388367853036, "learning_rate": 1.4559188399235175e-05, "loss": 0.3535, "step": 1433 }, { "epoch": 0.46297267847774326, "grad_norm": 0.40492829578434836, "learning_rate": 1.4558286212671295e-05, "loss": 0.3814, "step": 1434 }, { "epoch": 0.4632955325073651, "grad_norm": 0.5018299400990762, "learning_rate": 1.4557383131832273e-05, "loss": 0.345, "step": 1435 }, { "epoch": 0.46361838653698695, "grad_norm": 0.6123231937741075, "learning_rate": 1.4556479156832524e-05, "loss": 0.3796, "step": 1436 }, { "epoch": 0.46394124056660885, "grad_norm": 0.6988839214350205, "learning_rate": 1.4555574287786585e-05, "loss": 0.3874, "step": 1437 }, { "epoch": 0.4642640945962307, "grad_norm": 0.43069819243363744, "learning_rate": 1.45546685248091e-05, "loss": 0.3503, "step": 1438 }, { "epoch": 0.46458694862585254, "grad_norm": 0.4957362063341534, "learning_rate": 1.4553761868014829e-05, "loss": 0.2904, "step": 1439 }, { "epoch": 0.4649098026554744, "grad_norm": 0.40379233690710836, "learning_rate": 1.4552854317518642e-05, "loss": 0.3709, "step": 1440 }, { "epoch": 0.4652326566850962, "grad_norm": 0.5999024332800594, "learning_rate": 1.4551945873435527e-05, "loss": 0.3685, "step": 1441 }, { "epoch": 0.4655555107147181, "grad_norm": 0.44319578031247076, "learning_rate": 1.4551036535880578e-05, "loss": 0.362, "step": 1442 }, { "epoch": 0.46587836474433997, "grad_norm": 0.4122808324839989, "learning_rate": 1.4550126304969011e-05, "loss": 0.3406, "step": 1443 }, { "epoch": 0.4662012187739618, "grad_norm": 0.42847872712934837, "learning_rate": 1.454921518081615e-05, "loss": 0.3666, "step": 1444 }, { "epoch": 0.46652407280358366, "grad_norm": 0.47184233665162545, "learning_rate": 1.4548303163537431e-05, "loss": 0.3679, "step": 1445 }, { "epoch": 0.46684692683320556, "grad_norm": 0.43601343858019115, "learning_rate": 1.4547390253248411e-05, "loss": 0.3479, "step": 1446 }, { "epoch": 0.4671697808628274, "grad_norm": 0.5345721419951207, "learning_rate": 1.4546476450064748e-05, "loss": 0.4082, "step": 1447 }, { "epoch": 0.46749263489244924, "grad_norm": 0.4135493119042995, "learning_rate": 1.4545561754102222e-05, "loss": 0.3921, "step": 1448 }, { "epoch": 0.4678154889220711, "grad_norm": 0.655262257011645, "learning_rate": 1.4544646165476726e-05, "loss": 0.314, "step": 1449 }, { "epoch": 0.468138342951693, "grad_norm": 0.4202307685792153, "learning_rate": 1.454372968430426e-05, "loss": 0.334, "step": 1450 }, { "epoch": 0.46846119698131483, "grad_norm": 0.35568320648595686, "learning_rate": 1.4542812310700945e-05, "loss": 0.3457, "step": 1451 }, { "epoch": 0.4687840510109367, "grad_norm": 0.3836537184006408, "learning_rate": 1.4541894044783007e-05, "loss": 0.373, "step": 1452 }, { "epoch": 0.4691069050405585, "grad_norm": 0.6233945730649088, "learning_rate": 1.4540974886666792e-05, "loss": 0.3389, "step": 1453 }, { "epoch": 0.4694297590701804, "grad_norm": 0.5136474519146378, "learning_rate": 1.4540054836468753e-05, "loss": 0.3334, "step": 1454 }, { "epoch": 0.46975261309980226, "grad_norm": 0.453811277498552, "learning_rate": 1.4539133894305461e-05, "loss": 0.3692, "step": 1455 }, { "epoch": 0.4700754671294241, "grad_norm": 0.4030588020898255, "learning_rate": 1.45382120602936e-05, "loss": 0.3542, "step": 1456 }, { "epoch": 0.47039832115904595, "grad_norm": 0.4931117220440194, "learning_rate": 1.4537289334549963e-05, "loss": 0.3332, "step": 1457 }, { "epoch": 0.47072117518866785, "grad_norm": 0.43771588529949756, "learning_rate": 1.4536365717191456e-05, "loss": 0.3933, "step": 1458 }, { "epoch": 0.4710440292182897, "grad_norm": 0.3895631849035648, "learning_rate": 1.4535441208335106e-05, "loss": 0.2907, "step": 1459 }, { "epoch": 0.47136688324791154, "grad_norm": 0.36885534120320096, "learning_rate": 1.4534515808098038e-05, "loss": 0.3917, "step": 1460 }, { "epoch": 0.4716897372775334, "grad_norm": 0.5671555426771253, "learning_rate": 1.453358951659751e-05, "loss": 0.4512, "step": 1461 }, { "epoch": 0.4720125913071552, "grad_norm": 0.7744676115556061, "learning_rate": 1.4532662333950873e-05, "loss": 0.3499, "step": 1462 }, { "epoch": 0.4723354453367771, "grad_norm": 0.3942859938724589, "learning_rate": 1.4531734260275603e-05, "loss": 0.3424, "step": 1463 }, { "epoch": 0.47265829936639897, "grad_norm": 0.455716116963306, "learning_rate": 1.4530805295689285e-05, "loss": 0.3908, "step": 1464 }, { "epoch": 0.4729811533960208, "grad_norm": 0.49533756081571045, "learning_rate": 1.452987544030962e-05, "loss": 0.3824, "step": 1465 }, { "epoch": 0.47330400742564266, "grad_norm": 0.5025822107137119, "learning_rate": 1.4528944694254415e-05, "loss": 0.374, "step": 1466 }, { "epoch": 0.47362686145526456, "grad_norm": 0.6937604248451777, "learning_rate": 1.4528013057641598e-05, "loss": 0.3868, "step": 1467 }, { "epoch": 0.4739497154848864, "grad_norm": 0.3954599869320527, "learning_rate": 1.4527080530589204e-05, "loss": 0.3467, "step": 1468 }, { "epoch": 0.47427256951450825, "grad_norm": 0.5148319492000791, "learning_rate": 1.4526147113215386e-05, "loss": 0.3439, "step": 1469 }, { "epoch": 0.4745954235441301, "grad_norm": 0.5158881878356725, "learning_rate": 1.4525212805638402e-05, "loss": 0.3831, "step": 1470 }, { "epoch": 0.474918277573752, "grad_norm": 0.7230825080173306, "learning_rate": 1.4524277607976629e-05, "loss": 0.3836, "step": 1471 }, { "epoch": 0.47524113160337383, "grad_norm": 0.487081134609079, "learning_rate": 1.4523341520348558e-05, "loss": 0.365, "step": 1472 }, { "epoch": 0.4755639856329957, "grad_norm": 0.4379332077699249, "learning_rate": 1.4522404542872788e-05, "loss": 0.3855, "step": 1473 }, { "epoch": 0.4758868396626175, "grad_norm": 0.42847051796816793, "learning_rate": 1.4521466675668033e-05, "loss": 0.4457, "step": 1474 }, { "epoch": 0.4762096936922394, "grad_norm": 0.4719373620243792, "learning_rate": 1.452052791885312e-05, "loss": 0.397, "step": 1475 }, { "epoch": 0.47653254772186127, "grad_norm": 0.6371074329899215, "learning_rate": 1.4519588272546986e-05, "loss": 0.4461, "step": 1476 }, { "epoch": 0.4768554017514831, "grad_norm": 0.494554461187585, "learning_rate": 1.4518647736868685e-05, "loss": 0.4092, "step": 1477 }, { "epoch": 0.47717825578110495, "grad_norm": 0.6214518746913741, "learning_rate": 1.4517706311937379e-05, "loss": 0.388, "step": 1478 }, { "epoch": 0.4775011098107268, "grad_norm": 0.40548735724971435, "learning_rate": 1.451676399787235e-05, "loss": 0.3769, "step": 1479 }, { "epoch": 0.4778239638403487, "grad_norm": 0.4153003129845518, "learning_rate": 1.4515820794792984e-05, "loss": 0.4335, "step": 1480 }, { "epoch": 0.47814681786997054, "grad_norm": 0.5069142784382916, "learning_rate": 1.4514876702818785e-05, "loss": 0.3256, "step": 1481 }, { "epoch": 0.4784696718995924, "grad_norm": 0.552807542374448, "learning_rate": 1.4513931722069367e-05, "loss": 0.327, "step": 1482 }, { "epoch": 0.47879252592921423, "grad_norm": 0.45364415876921826, "learning_rate": 1.4512985852664458e-05, "loss": 0.4111, "step": 1483 }, { "epoch": 0.47911537995883613, "grad_norm": 0.4220928135229913, "learning_rate": 1.45120390947239e-05, "loss": 0.318, "step": 1484 }, { "epoch": 0.479438233988458, "grad_norm": 0.41857795947673754, "learning_rate": 1.4511091448367646e-05, "loss": 0.3969, "step": 1485 }, { "epoch": 0.4797610880180798, "grad_norm": 0.4807781474719529, "learning_rate": 1.451014291371576e-05, "loss": 0.3753, "step": 1486 }, { "epoch": 0.48008394204770166, "grad_norm": 0.5696241508957919, "learning_rate": 1.4509193490888418e-05, "loss": 0.4016, "step": 1487 }, { "epoch": 0.48040679607732356, "grad_norm": 0.3635201556320272, "learning_rate": 1.4508243180005913e-05, "loss": 0.3378, "step": 1488 }, { "epoch": 0.4807296501069454, "grad_norm": 0.39623100513201376, "learning_rate": 1.4507291981188652e-05, "loss": 0.3416, "step": 1489 }, { "epoch": 0.48105250413656725, "grad_norm": 0.5820164793379732, "learning_rate": 1.4506339894557142e-05, "loss": 0.3088, "step": 1490 }, { "epoch": 0.4813753581661891, "grad_norm": 0.4192820945452042, "learning_rate": 1.4505386920232018e-05, "loss": 0.4456, "step": 1491 }, { "epoch": 0.481698212195811, "grad_norm": 0.5182301625669299, "learning_rate": 1.4504433058334019e-05, "loss": 0.3195, "step": 1492 }, { "epoch": 0.48202106622543284, "grad_norm": 0.48672472436785474, "learning_rate": 1.4503478308983996e-05, "loss": 0.4096, "step": 1493 }, { "epoch": 0.4823439202550547, "grad_norm": 0.480176895815714, "learning_rate": 1.4502522672302917e-05, "loss": 0.3691, "step": 1494 }, { "epoch": 0.4826667742846765, "grad_norm": 0.5272057776535578, "learning_rate": 1.4501566148411857e-05, "loss": 0.3118, "step": 1495 }, { "epoch": 0.4829896283142984, "grad_norm": 0.3829232095068228, "learning_rate": 1.4500608737432007e-05, "loss": 0.3279, "step": 1496 }, { "epoch": 0.48331248234392027, "grad_norm": 0.39724625156848775, "learning_rate": 1.4499650439484672e-05, "loss": 0.3744, "step": 1497 }, { "epoch": 0.4836353363735421, "grad_norm": 0.46563835570500706, "learning_rate": 1.4498691254691265e-05, "loss": 0.3216, "step": 1498 }, { "epoch": 0.48395819040316396, "grad_norm": 0.48507612183881554, "learning_rate": 1.4497731183173314e-05, "loss": 0.3658, "step": 1499 }, { "epoch": 0.4842810444327858, "grad_norm": 0.4165002351161967, "learning_rate": 1.4496770225052457e-05, "loss": 0.3866, "step": 1500 }, { "epoch": 0.4846038984624077, "grad_norm": 0.4628479712436437, "learning_rate": 1.4495808380450448e-05, "loss": 0.3522, "step": 1501 }, { "epoch": 0.48492675249202954, "grad_norm": 0.46704345766890426, "learning_rate": 1.4494845649489153e-05, "loss": 0.3711, "step": 1502 }, { "epoch": 0.4852496065216514, "grad_norm": 0.4323195482108475, "learning_rate": 1.4493882032290545e-05, "loss": 0.3089, "step": 1503 }, { "epoch": 0.48557246055127323, "grad_norm": 0.4214883712690677, "learning_rate": 1.4492917528976715e-05, "loss": 0.3871, "step": 1504 }, { "epoch": 0.48589531458089513, "grad_norm": 0.47756710049923723, "learning_rate": 1.4491952139669865e-05, "loss": 0.3028, "step": 1505 }, { "epoch": 0.486218168610517, "grad_norm": 0.500195126580906, "learning_rate": 1.4490985864492305e-05, "loss": 0.4177, "step": 1506 }, { "epoch": 0.4865410226401388, "grad_norm": 0.4736018498177815, "learning_rate": 1.4490018703566465e-05, "loss": 0.3666, "step": 1507 }, { "epoch": 0.48686387666976066, "grad_norm": 0.41284374999436607, "learning_rate": 1.4489050657014881e-05, "loss": 0.3242, "step": 1508 }, { "epoch": 0.48718673069938256, "grad_norm": 0.4559307747843257, "learning_rate": 1.4488081724960203e-05, "loss": 0.395, "step": 1509 }, { "epoch": 0.4875095847290044, "grad_norm": 0.4581701015116004, "learning_rate": 1.4487111907525193e-05, "loss": 0.3318, "step": 1510 }, { "epoch": 0.48783243875862625, "grad_norm": 0.3342427374356882, "learning_rate": 1.4486141204832727e-05, "loss": 0.3693, "step": 1511 }, { "epoch": 0.4881552927882481, "grad_norm": 0.4012408353025965, "learning_rate": 1.4485169617005792e-05, "loss": 0.4099, "step": 1512 }, { "epoch": 0.48847814681787, "grad_norm": 0.46989132563788943, "learning_rate": 1.4484197144167482e-05, "loss": 0.4588, "step": 1513 }, { "epoch": 0.48880100084749184, "grad_norm": 0.3428684610804384, "learning_rate": 1.4483223786441016e-05, "loss": 0.3306, "step": 1514 }, { "epoch": 0.4891238548771137, "grad_norm": 0.4253583659077329, "learning_rate": 1.4482249543949708e-05, "loss": 0.3134, "step": 1515 }, { "epoch": 0.48944670890673553, "grad_norm": 0.4187148811509565, "learning_rate": 1.4481274416817e-05, "loss": 0.3212, "step": 1516 }, { "epoch": 0.48976956293635737, "grad_norm": 0.4697236591183174, "learning_rate": 1.4480298405166437e-05, "loss": 0.3977, "step": 1517 }, { "epoch": 0.49009241696597927, "grad_norm": 0.4299214886452003, "learning_rate": 1.447932150912168e-05, "loss": 0.3912, "step": 1518 }, { "epoch": 0.4904152709956011, "grad_norm": 0.3964990956438535, "learning_rate": 1.4478343728806498e-05, "loss": 0.3475, "step": 1519 }, { "epoch": 0.49073812502522296, "grad_norm": 0.435663222824711, "learning_rate": 1.4477365064344775e-05, "loss": 0.3614, "step": 1520 }, { "epoch": 0.4910609790548448, "grad_norm": 0.39188711029790146, "learning_rate": 1.4476385515860508e-05, "loss": 0.309, "step": 1521 }, { "epoch": 0.4913838330844667, "grad_norm": 0.4784982722286075, "learning_rate": 1.4475405083477803e-05, "loss": 0.3889, "step": 1522 }, { "epoch": 0.49170668711408855, "grad_norm": 0.4317867338251005, "learning_rate": 1.4474423767320878e-05, "loss": 0.4008, "step": 1523 }, { "epoch": 0.4920295411437104, "grad_norm": 0.40589179636902417, "learning_rate": 1.4473441567514068e-05, "loss": 0.3632, "step": 1524 }, { "epoch": 0.49235239517333224, "grad_norm": 0.3698050254601002, "learning_rate": 1.4472458484181813e-05, "loss": 0.3706, "step": 1525 }, { "epoch": 0.49267524920295414, "grad_norm": 0.4228401662441022, "learning_rate": 1.447147451744867e-05, "loss": 0.387, "step": 1526 }, { "epoch": 0.492998103232576, "grad_norm": 0.4334978403864301, "learning_rate": 1.4470489667439306e-05, "loss": 0.3905, "step": 1527 }, { "epoch": 0.4933209572621978, "grad_norm": 0.47136091608880987, "learning_rate": 1.4469503934278499e-05, "loss": 0.4079, "step": 1528 }, { "epoch": 0.49364381129181967, "grad_norm": 0.4189689712974094, "learning_rate": 1.4468517318091143e-05, "loss": 0.3804, "step": 1529 }, { "epoch": 0.49396666532144157, "grad_norm": 0.3688083833049135, "learning_rate": 1.4467529819002238e-05, "loss": 0.2851, "step": 1530 }, { "epoch": 0.4942895193510634, "grad_norm": 0.4478373103388335, "learning_rate": 1.44665414371369e-05, "loss": 0.3305, "step": 1531 }, { "epoch": 0.49461237338068526, "grad_norm": 0.8935391253359227, "learning_rate": 1.4465552172620356e-05, "loss": 0.397, "step": 1532 }, { "epoch": 0.4949352274103071, "grad_norm": 0.5940704497516508, "learning_rate": 1.4464562025577943e-05, "loss": 0.3487, "step": 1533 }, { "epoch": 0.495258081439929, "grad_norm": 0.4411424004704336, "learning_rate": 1.4463570996135111e-05, "loss": 0.4082, "step": 1534 }, { "epoch": 0.49558093546955084, "grad_norm": 0.5130952571503121, "learning_rate": 1.4462579084417425e-05, "loss": 0.3646, "step": 1535 }, { "epoch": 0.4959037894991727, "grad_norm": 0.4026842795884311, "learning_rate": 1.4461586290550557e-05, "loss": 0.3327, "step": 1536 }, { "epoch": 0.49622664352879453, "grad_norm": 0.4366575029104958, "learning_rate": 1.4460592614660292e-05, "loss": 0.3849, "step": 1537 }, { "epoch": 0.4965494975584164, "grad_norm": 0.499856561226089, "learning_rate": 1.4459598056872529e-05, "loss": 0.281, "step": 1538 }, { "epoch": 0.4968723515880383, "grad_norm": 0.4195695545530234, "learning_rate": 1.4458602617313273e-05, "loss": 0.4243, "step": 1539 }, { "epoch": 0.4971952056176601, "grad_norm": 0.37353302100668756, "learning_rate": 1.4457606296108653e-05, "loss": 0.3261, "step": 1540 }, { "epoch": 0.49751805964728196, "grad_norm": 0.4258674136645591, "learning_rate": 1.4456609093384892e-05, "loss": 0.3478, "step": 1541 }, { "epoch": 0.4978409136769038, "grad_norm": 0.40367804093929627, "learning_rate": 1.445561100926834e-05, "loss": 0.3576, "step": 1542 }, { "epoch": 0.4981637677065257, "grad_norm": 0.5398518166178663, "learning_rate": 1.4454612043885453e-05, "loss": 0.3647, "step": 1543 }, { "epoch": 0.49848662173614755, "grad_norm": 0.4111479562779954, "learning_rate": 1.4453612197362798e-05, "loss": 0.403, "step": 1544 }, { "epoch": 0.4988094757657694, "grad_norm": 0.451693906363503, "learning_rate": 1.445261146982705e-05, "loss": 0.3796, "step": 1545 }, { "epoch": 0.49913232979539124, "grad_norm": 0.44482702824436876, "learning_rate": 1.4451609861405004e-05, "loss": 0.3908, "step": 1546 }, { "epoch": 0.49945518382501314, "grad_norm": 0.42291138510143544, "learning_rate": 1.4450607372223563e-05, "loss": 0.4153, "step": 1547 }, { "epoch": 0.499778037854635, "grad_norm": 0.47612979462826577, "learning_rate": 1.4449604002409738e-05, "loss": 0.3452, "step": 1548 }, { "epoch": 0.5001008918842569, "grad_norm": 0.5022209587301101, "learning_rate": 1.4448599752090657e-05, "loss": 0.4023, "step": 1549 }, { "epoch": 0.5004237459138787, "grad_norm": 0.46894272573592793, "learning_rate": 1.4447594621393555e-05, "loss": 0.3636, "step": 1550 }, { "epoch": 0.5007465999435006, "grad_norm": 0.5233843504798762, "learning_rate": 1.4446588610445783e-05, "loss": 0.4702, "step": 1551 }, { "epoch": 0.5010694539731224, "grad_norm": 0.4272015098780948, "learning_rate": 1.4445581719374802e-05, "loss": 0.4033, "step": 1552 }, { "epoch": 0.5013923080027443, "grad_norm": 0.3762457469008636, "learning_rate": 1.4444573948308179e-05, "loss": 0.3029, "step": 1553 }, { "epoch": 0.5017151620323661, "grad_norm": 0.40075352822767907, "learning_rate": 1.44435652973736e-05, "loss": 0.3805, "step": 1554 }, { "epoch": 0.502038016061988, "grad_norm": 0.49368566630829264, "learning_rate": 1.444255576669886e-05, "loss": 0.3652, "step": 1555 }, { "epoch": 0.5023608700916098, "grad_norm": 0.5397994485791299, "learning_rate": 1.4441545356411864e-05, "loss": 0.325, "step": 1556 }, { "epoch": 0.5026837241212316, "grad_norm": 0.41977834414459037, "learning_rate": 1.4440534066640634e-05, "loss": 0.3569, "step": 1557 }, { "epoch": 0.5030065781508536, "grad_norm": 0.6009535142337842, "learning_rate": 1.443952189751329e-05, "loss": 0.3209, "step": 1558 }, { "epoch": 0.5033294321804754, "grad_norm": 0.46782767234326195, "learning_rate": 1.4438508849158081e-05, "loss": 0.4201, "step": 1559 }, { "epoch": 0.5036522862100973, "grad_norm": 0.7548494053964795, "learning_rate": 1.4437494921703357e-05, "loss": 0.3611, "step": 1560 }, { "epoch": 0.5039751402397191, "grad_norm": 0.582793880292644, "learning_rate": 1.443648011527758e-05, "loss": 0.3645, "step": 1561 }, { "epoch": 0.504297994269341, "grad_norm": 0.5182975059455805, "learning_rate": 1.443546443000932e-05, "loss": 0.3629, "step": 1562 }, { "epoch": 0.5046208482989628, "grad_norm": 0.48516721460571, "learning_rate": 1.4434447866027272e-05, "loss": 0.4032, "step": 1563 }, { "epoch": 0.5049437023285847, "grad_norm": 0.47552099901906386, "learning_rate": 1.4433430423460227e-05, "loss": 0.3983, "step": 1564 }, { "epoch": 0.5052665563582065, "grad_norm": 0.39926962681461814, "learning_rate": 1.4432412102437095e-05, "loss": 0.3501, "step": 1565 }, { "epoch": 0.5055894103878285, "grad_norm": 0.4166510626332838, "learning_rate": 1.4431392903086896e-05, "loss": 0.3434, "step": 1566 }, { "epoch": 0.5059122644174503, "grad_norm": 0.4779447092752385, "learning_rate": 1.4430372825538763e-05, "loss": 0.3681, "step": 1567 }, { "epoch": 0.5062351184470721, "grad_norm": 0.4012361480045925, "learning_rate": 1.4429351869921936e-05, "loss": 0.359, "step": 1568 }, { "epoch": 0.506557972476694, "grad_norm": 0.4908641980990743, "learning_rate": 1.4428330036365766e-05, "loss": 0.3439, "step": 1569 }, { "epoch": 0.5068808265063158, "grad_norm": 0.6540160124177294, "learning_rate": 1.4427307324999726e-05, "loss": 0.3728, "step": 1570 }, { "epoch": 0.5072036805359377, "grad_norm": 0.4384719058874475, "learning_rate": 1.4426283735953387e-05, "loss": 0.3773, "step": 1571 }, { "epoch": 0.5075265345655595, "grad_norm": 0.4320238467744891, "learning_rate": 1.4425259269356435e-05, "loss": 0.3042, "step": 1572 }, { "epoch": 0.5078493885951814, "grad_norm": 0.3978743166794824, "learning_rate": 1.442423392533867e-05, "loss": 0.38, "step": 1573 }, { "epoch": 0.5081722426248032, "grad_norm": 0.31456615993008447, "learning_rate": 1.4423207704030001e-05, "loss": 0.3348, "step": 1574 }, { "epoch": 0.5084950966544252, "grad_norm": 0.4138115288738691, "learning_rate": 1.4422180605560452e-05, "loss": 0.4277, "step": 1575 }, { "epoch": 0.508817950684047, "grad_norm": 0.37277566296969844, "learning_rate": 1.442115263006015e-05, "loss": 0.4063, "step": 1576 }, { "epoch": 0.5091408047136688, "grad_norm": 1.2033949506803263, "learning_rate": 1.4420123777659344e-05, "loss": 0.3372, "step": 1577 }, { "epoch": 0.5094636587432907, "grad_norm": 0.42261565996154365, "learning_rate": 1.4419094048488382e-05, "loss": 0.4136, "step": 1578 }, { "epoch": 0.5097865127729125, "grad_norm": 0.44093846934702474, "learning_rate": 1.4418063442677733e-05, "loss": 0.3841, "step": 1579 }, { "epoch": 0.5101093668025344, "grad_norm": 0.39104928749682205, "learning_rate": 1.4417031960357974e-05, "loss": 0.3617, "step": 1580 }, { "epoch": 0.5104322208321562, "grad_norm": 0.5278402461363656, "learning_rate": 1.441599960165979e-05, "loss": 0.3673, "step": 1581 }, { "epoch": 0.5107550748617781, "grad_norm": 0.5089490180428297, "learning_rate": 1.4414966366713982e-05, "loss": 0.3645, "step": 1582 }, { "epoch": 0.5110779288914, "grad_norm": 0.5114735637313986, "learning_rate": 1.4413932255651456e-05, "loss": 0.3823, "step": 1583 }, { "epoch": 0.5114007829210219, "grad_norm": 0.4642915256089158, "learning_rate": 1.4412897268603234e-05, "loss": 0.3592, "step": 1584 }, { "epoch": 0.5117236369506437, "grad_norm": 0.430617746418858, "learning_rate": 1.4411861405700448e-05, "loss": 0.4059, "step": 1585 }, { "epoch": 0.5120464909802656, "grad_norm": 0.49137793565383653, "learning_rate": 1.441082466707434e-05, "loss": 0.3242, "step": 1586 }, { "epoch": 0.5123693450098874, "grad_norm": 0.6354054686815104, "learning_rate": 1.4409787052856264e-05, "loss": 0.3207, "step": 1587 }, { "epoch": 0.5126921990395092, "grad_norm": 0.5847910150314269, "learning_rate": 1.4408748563177685e-05, "loss": 0.3295, "step": 1588 }, { "epoch": 0.5130150530691311, "grad_norm": 0.4571492162514275, "learning_rate": 1.4407709198170176e-05, "loss": 0.379, "step": 1589 }, { "epoch": 0.5133379070987529, "grad_norm": 0.4331261143725809, "learning_rate": 1.4406668957965423e-05, "loss": 0.3873, "step": 1590 }, { "epoch": 0.5136607611283749, "grad_norm": 0.4412151969530291, "learning_rate": 1.4405627842695222e-05, "loss": 0.4071, "step": 1591 }, { "epoch": 0.5139836151579967, "grad_norm": 0.5023938182142434, "learning_rate": 1.4404585852491488e-05, "loss": 0.3712, "step": 1592 }, { "epoch": 0.5143064691876186, "grad_norm": 0.3627003375546108, "learning_rate": 1.4403542987486231e-05, "loss": 0.317, "step": 1593 }, { "epoch": 0.5146293232172404, "grad_norm": 0.5625726561451334, "learning_rate": 1.4402499247811586e-05, "loss": 0.3959, "step": 1594 }, { "epoch": 0.5149521772468623, "grad_norm": 0.47274156596880185, "learning_rate": 1.4401454633599791e-05, "loss": 0.3418, "step": 1595 }, { "epoch": 0.5152750312764841, "grad_norm": 0.5121862564710254, "learning_rate": 1.4400409144983199e-05, "loss": 0.3623, "step": 1596 }, { "epoch": 0.515597885306106, "grad_norm": 0.3747021558708433, "learning_rate": 1.439936278209427e-05, "loss": 0.3216, "step": 1597 }, { "epoch": 0.5159207393357278, "grad_norm": 0.4370629657857402, "learning_rate": 1.4398315545065576e-05, "loss": 0.3819, "step": 1598 }, { "epoch": 0.5162435933653496, "grad_norm": 0.502805352631067, "learning_rate": 1.4397267434029802e-05, "loss": 0.3591, "step": 1599 }, { "epoch": 0.5165664473949716, "grad_norm": 0.4904127982659012, "learning_rate": 1.4396218449119746e-05, "loss": 0.3299, "step": 1600 }, { "epoch": 0.5168893014245934, "grad_norm": 0.7682749282541071, "learning_rate": 1.4395168590468305e-05, "loss": 0.3221, "step": 1601 }, { "epoch": 0.5172121554542153, "grad_norm": 0.48721679086605746, "learning_rate": 1.43941178582085e-05, "loss": 0.2997, "step": 1602 }, { "epoch": 0.5175350094838371, "grad_norm": 0.36575864265212915, "learning_rate": 1.4393066252473459e-05, "loss": 0.3094, "step": 1603 }, { "epoch": 0.517857863513459, "grad_norm": 0.472413822022977, "learning_rate": 1.439201377339641e-05, "loss": 0.3595, "step": 1604 }, { "epoch": 0.5181807175430808, "grad_norm": 0.5475006639694409, "learning_rate": 1.4390960421110711e-05, "loss": 0.316, "step": 1605 }, { "epoch": 0.5185035715727027, "grad_norm": 0.4793625426086169, "learning_rate": 1.4389906195749816e-05, "loss": 0.3478, "step": 1606 }, { "epoch": 0.5188264256023245, "grad_norm": 0.5647043386958797, "learning_rate": 1.4388851097447293e-05, "loss": 0.3319, "step": 1607 }, { "epoch": 0.5191492796319465, "grad_norm": 0.5705616809922412, "learning_rate": 1.4387795126336824e-05, "loss": 0.3304, "step": 1608 }, { "epoch": 0.5194721336615683, "grad_norm": 0.3786824044738472, "learning_rate": 1.4386738282552196e-05, "loss": 0.3851, "step": 1609 }, { "epoch": 0.5197949876911901, "grad_norm": 0.367363905390778, "learning_rate": 1.438568056622731e-05, "loss": 0.3194, "step": 1610 }, { "epoch": 0.520117841720812, "grad_norm": 0.6281607306447525, "learning_rate": 1.4384621977496182e-05, "loss": 0.4148, "step": 1611 }, { "epoch": 0.5204406957504338, "grad_norm": 0.4734703334408823, "learning_rate": 1.4383562516492927e-05, "loss": 0.3949, "step": 1612 }, { "epoch": 0.5207635497800557, "grad_norm": 0.4386946332449736, "learning_rate": 1.4382502183351783e-05, "loss": 0.3797, "step": 1613 }, { "epoch": 0.5210864038096775, "grad_norm": 0.3843375678957885, "learning_rate": 1.4381440978207089e-05, "loss": 0.366, "step": 1614 }, { "epoch": 0.5214092578392994, "grad_norm": 0.9811814770745791, "learning_rate": 1.4380378901193298e-05, "loss": 0.3738, "step": 1615 }, { "epoch": 0.5217321118689212, "grad_norm": 0.43981618882188833, "learning_rate": 1.4379315952444975e-05, "loss": 0.3164, "step": 1616 }, { "epoch": 0.5220549658985432, "grad_norm": 0.43363862627746363, "learning_rate": 1.4378252132096797e-05, "loss": 0.4015, "step": 1617 }, { "epoch": 0.522377819928165, "grad_norm": 0.5471832074996964, "learning_rate": 1.4377187440283543e-05, "loss": 0.3672, "step": 1618 }, { "epoch": 0.5227006739577869, "grad_norm": 0.3985408205757699, "learning_rate": 1.4376121877140112e-05, "loss": 0.3593, "step": 1619 }, { "epoch": 0.5230235279874087, "grad_norm": 0.5704794127498953, "learning_rate": 1.4375055442801509e-05, "loss": 0.3391, "step": 1620 }, { "epoch": 0.5233463820170305, "grad_norm": 0.39125694371927294, "learning_rate": 1.4373988137402848e-05, "loss": 0.3583, "step": 1621 }, { "epoch": 0.5236692360466524, "grad_norm": 0.4322177184067198, "learning_rate": 1.4372919961079355e-05, "loss": 0.3753, "step": 1622 }, { "epoch": 0.5239920900762742, "grad_norm": 0.4732932907438303, "learning_rate": 1.437185091396637e-05, "loss": 0.3603, "step": 1623 }, { "epoch": 0.5243149441058961, "grad_norm": 0.3593478623674238, "learning_rate": 1.4370780996199335e-05, "loss": 0.3562, "step": 1624 }, { "epoch": 0.524637798135518, "grad_norm": 0.41885115748738316, "learning_rate": 1.436971020791381e-05, "loss": 0.3284, "step": 1625 }, { "epoch": 0.5249606521651399, "grad_norm": 0.600861211086527, "learning_rate": 1.4368638549245462e-05, "loss": 0.4066, "step": 1626 }, { "epoch": 0.5252835061947617, "grad_norm": 0.48347384023279205, "learning_rate": 1.436756602033007e-05, "loss": 0.355, "step": 1627 }, { "epoch": 0.5256063602243836, "grad_norm": 0.4991155065051953, "learning_rate": 1.4366492621303517e-05, "loss": 0.3343, "step": 1628 }, { "epoch": 0.5259292142540054, "grad_norm": 0.5499548600301265, "learning_rate": 1.4365418352301808e-05, "loss": 0.364, "step": 1629 }, { "epoch": 0.5262520682836273, "grad_norm": 0.7925493967370553, "learning_rate": 1.4364343213461046e-05, "loss": 0.3621, "step": 1630 }, { "epoch": 0.5265749223132491, "grad_norm": 0.4082999105314196, "learning_rate": 1.4363267204917453e-05, "loss": 0.3048, "step": 1631 }, { "epoch": 0.5268977763428709, "grad_norm": 0.6679515057398174, "learning_rate": 1.4362190326807355e-05, "loss": 0.3736, "step": 1632 }, { "epoch": 0.5272206303724928, "grad_norm": 0.39842397348118586, "learning_rate": 1.4361112579267193e-05, "loss": 0.3462, "step": 1633 }, { "epoch": 0.5275434844021147, "grad_norm": 0.40356514464168053, "learning_rate": 1.4360033962433514e-05, "loss": 0.3493, "step": 1634 }, { "epoch": 0.5278663384317366, "grad_norm": 1.0046429513277941, "learning_rate": 1.435895447644298e-05, "loss": 0.3366, "step": 1635 }, { "epoch": 0.5281891924613584, "grad_norm": 0.39046598779405767, "learning_rate": 1.4357874121432362e-05, "loss": 0.3683, "step": 1636 }, { "epoch": 0.5285120464909803, "grad_norm": 0.3821367119212212, "learning_rate": 1.4356792897538533e-05, "loss": 0.3017, "step": 1637 }, { "epoch": 0.5288349005206021, "grad_norm": 0.741861731858265, "learning_rate": 1.4355710804898487e-05, "loss": 0.3703, "step": 1638 }, { "epoch": 0.529157754550224, "grad_norm": 0.4704738936996858, "learning_rate": 1.4354627843649324e-05, "loss": 0.3581, "step": 1639 }, { "epoch": 0.5294806085798458, "grad_norm": 0.4829195217145884, "learning_rate": 1.435354401392825e-05, "loss": 0.3191, "step": 1640 }, { "epoch": 0.5298034626094676, "grad_norm": 0.41953449938421583, "learning_rate": 1.435245931587259e-05, "loss": 0.3937, "step": 1641 }, { "epoch": 0.5301263166390896, "grad_norm": 0.4704271054543495, "learning_rate": 1.4351373749619771e-05, "loss": 0.4432, "step": 1642 }, { "epoch": 0.5304491706687114, "grad_norm": 0.539276677782367, "learning_rate": 1.4350287315307332e-05, "loss": 0.4366, "step": 1643 }, { "epoch": 0.5307720246983333, "grad_norm": 0.45962695370258255, "learning_rate": 1.4349200013072923e-05, "loss": 0.3784, "step": 1644 }, { "epoch": 0.5310948787279551, "grad_norm": 0.4497480290305102, "learning_rate": 1.4348111843054306e-05, "loss": 0.3027, "step": 1645 }, { "epoch": 0.531417732757577, "grad_norm": 0.4105793577352083, "learning_rate": 1.4347022805389347e-05, "loss": 0.3255, "step": 1646 }, { "epoch": 0.5317405867871988, "grad_norm": 0.4446782509685553, "learning_rate": 1.4345932900216029e-05, "loss": 0.3409, "step": 1647 }, { "epoch": 0.5320634408168207, "grad_norm": 0.3688048333904954, "learning_rate": 1.4344842127672438e-05, "loss": 0.3879, "step": 1648 }, { "epoch": 0.5323862948464425, "grad_norm": 0.4742345894671136, "learning_rate": 1.4343750487896778e-05, "loss": 0.3551, "step": 1649 }, { "epoch": 0.5327091488760645, "grad_norm": 0.4936850172831148, "learning_rate": 1.4342657981027352e-05, "loss": 0.3709, "step": 1650 }, { "epoch": 0.5330320029056863, "grad_norm": 0.4902970327058587, "learning_rate": 1.4341564607202587e-05, "loss": 0.3577, "step": 1651 }, { "epoch": 0.5333548569353082, "grad_norm": 0.4794831695980391, "learning_rate": 1.4340470366561003e-05, "loss": 0.4021, "step": 1652 }, { "epoch": 0.53367771096493, "grad_norm": 0.49371209879261724, "learning_rate": 1.4339375259241247e-05, "loss": 0.3716, "step": 1653 }, { "epoch": 0.5340005649945518, "grad_norm": 0.4599859224302241, "learning_rate": 1.4338279285382065e-05, "loss": 0.3429, "step": 1654 }, { "epoch": 0.5343234190241737, "grad_norm": 0.4220152035932215, "learning_rate": 1.4337182445122312e-05, "loss": 0.3833, "step": 1655 }, { "epoch": 0.5346462730537955, "grad_norm": 0.40517441315088604, "learning_rate": 1.4336084738600957e-05, "loss": 0.354, "step": 1656 }, { "epoch": 0.5349691270834174, "grad_norm": 0.405803185357693, "learning_rate": 1.4334986165957083e-05, "loss": 0.3269, "step": 1657 }, { "epoch": 0.5352919811130392, "grad_norm": 0.3569296162085836, "learning_rate": 1.4333886727329874e-05, "loss": 0.3186, "step": 1658 }, { "epoch": 0.5356148351426612, "grad_norm": 0.53038048241853, "learning_rate": 1.4332786422858625e-05, "loss": 0.4312, "step": 1659 }, { "epoch": 0.535937689172283, "grad_norm": 0.45195830890667943, "learning_rate": 1.4331685252682746e-05, "loss": 0.4288, "step": 1660 }, { "epoch": 0.5362605432019049, "grad_norm": 0.37903929200020897, "learning_rate": 1.4330583216941753e-05, "loss": 0.3482, "step": 1661 }, { "epoch": 0.5365833972315267, "grad_norm": 0.48638812635978773, "learning_rate": 1.4329480315775274e-05, "loss": 0.4051, "step": 1662 }, { "epoch": 0.5369062512611485, "grad_norm": 0.48875046749913775, "learning_rate": 1.4328376549323044e-05, "loss": 0.3554, "step": 1663 }, { "epoch": 0.5372291052907704, "grad_norm": 0.478151506634884, "learning_rate": 1.4327271917724907e-05, "loss": 0.3557, "step": 1664 }, { "epoch": 0.5375519593203922, "grad_norm": 0.4752336576688321, "learning_rate": 1.4326166421120821e-05, "loss": 0.3851, "step": 1665 }, { "epoch": 0.5378748133500141, "grad_norm": 0.4366720989152335, "learning_rate": 1.4325060059650849e-05, "loss": 0.3848, "step": 1666 }, { "epoch": 0.538197667379636, "grad_norm": 0.6142448911371554, "learning_rate": 1.4323952833455166e-05, "loss": 0.36, "step": 1667 }, { "epoch": 0.5385205214092579, "grad_norm": 0.4775630273984633, "learning_rate": 1.4322844742674056e-05, "loss": 0.4023, "step": 1668 }, { "epoch": 0.5388433754388797, "grad_norm": 0.5621682571800727, "learning_rate": 1.4321735787447913e-05, "loss": 0.3641, "step": 1669 }, { "epoch": 0.5391662294685016, "grad_norm": 0.47958618137581766, "learning_rate": 1.4320625967917239e-05, "loss": 0.2661, "step": 1670 }, { "epoch": 0.5394890834981234, "grad_norm": 0.5241928476375013, "learning_rate": 1.431951528422265e-05, "loss": 0.3465, "step": 1671 }, { "epoch": 0.5398119375277453, "grad_norm": 0.5261201611799695, "learning_rate": 1.4318403736504863e-05, "loss": 0.3744, "step": 1672 }, { "epoch": 0.5401347915573671, "grad_norm": 0.4084111028348659, "learning_rate": 1.4317291324904713e-05, "loss": 0.3504, "step": 1673 }, { "epoch": 0.5404576455869889, "grad_norm": 0.4925061681980886, "learning_rate": 1.431617804956314e-05, "loss": 0.305, "step": 1674 }, { "epoch": 0.5407804996166108, "grad_norm": 0.47429370742781923, "learning_rate": 1.4315063910621195e-05, "loss": 0.4227, "step": 1675 }, { "epoch": 0.5411033536462327, "grad_norm": 0.4217949335659584, "learning_rate": 1.4313948908220036e-05, "loss": 0.3697, "step": 1676 }, { "epoch": 0.5414262076758546, "grad_norm": 0.5354730149703977, "learning_rate": 1.4312833042500935e-05, "loss": 0.3801, "step": 1677 }, { "epoch": 0.5417490617054764, "grad_norm": 0.42539894346735596, "learning_rate": 1.431171631360527e-05, "loss": 0.3496, "step": 1678 }, { "epoch": 0.5420719157350983, "grad_norm": 0.5232090044945452, "learning_rate": 1.4310598721674529e-05, "loss": 0.2791, "step": 1679 }, { "epoch": 0.5423947697647201, "grad_norm": 0.3785919351544836, "learning_rate": 1.4309480266850306e-05, "loss": 0.3363, "step": 1680 }, { "epoch": 0.542717623794342, "grad_norm": 0.46202517683369987, "learning_rate": 1.4308360949274309e-05, "loss": 0.3805, "step": 1681 }, { "epoch": 0.5430404778239638, "grad_norm": 0.4574602745839506, "learning_rate": 1.430724076908836e-05, "loss": 0.351, "step": 1682 }, { "epoch": 0.5433633318535857, "grad_norm": 0.45349076425544393, "learning_rate": 1.4306119726434377e-05, "loss": 0.3492, "step": 1683 }, { "epoch": 0.5436861858832076, "grad_norm": 0.45203275053975667, "learning_rate": 1.4304997821454398e-05, "loss": 0.3254, "step": 1684 }, { "epoch": 0.5440090399128295, "grad_norm": 0.3513040683250184, "learning_rate": 1.4303875054290565e-05, "loss": 0.3361, "step": 1685 }, { "epoch": 0.5443318939424513, "grad_norm": 0.3765439654160628, "learning_rate": 1.4302751425085134e-05, "loss": 0.3298, "step": 1686 }, { "epoch": 0.5446547479720731, "grad_norm": 0.44830027209723555, "learning_rate": 1.4301626933980464e-05, "loss": 0.393, "step": 1687 }, { "epoch": 0.544977602001695, "grad_norm": 0.43389287228668866, "learning_rate": 1.430050158111903e-05, "loss": 0.3479, "step": 1688 }, { "epoch": 0.5453004560313168, "grad_norm": 0.4483846278170194, "learning_rate": 1.4299375366643408e-05, "loss": 0.3367, "step": 1689 }, { "epoch": 0.5456233100609387, "grad_norm": 0.45478856612754986, "learning_rate": 1.4298248290696291e-05, "loss": 0.3832, "step": 1690 }, { "epoch": 0.5459461640905605, "grad_norm": 0.4801013497526643, "learning_rate": 1.4297120353420478e-05, "loss": 0.3569, "step": 1691 }, { "epoch": 0.5462690181201824, "grad_norm": 0.36671600015889483, "learning_rate": 1.4295991554958876e-05, "loss": 0.2636, "step": 1692 }, { "epoch": 0.5465918721498043, "grad_norm": 0.4897110082707557, "learning_rate": 1.4294861895454501e-05, "loss": 0.3697, "step": 1693 }, { "epoch": 0.5469147261794262, "grad_norm": 0.3989692578052694, "learning_rate": 1.4293731375050483e-05, "loss": 0.4064, "step": 1694 }, { "epoch": 0.547237580209048, "grad_norm": 0.48355483435418006, "learning_rate": 1.4292599993890053e-05, "loss": 0.3294, "step": 1695 }, { "epoch": 0.5475604342386698, "grad_norm": 0.43190192270908073, "learning_rate": 1.4291467752116561e-05, "loss": 0.3513, "step": 1696 }, { "epoch": 0.5478832882682917, "grad_norm": 0.4540990000907308, "learning_rate": 1.4290334649873454e-05, "loss": 0.436, "step": 1697 }, { "epoch": 0.5482061422979135, "grad_norm": 0.37161600256273836, "learning_rate": 1.42892006873043e-05, "loss": 0.3344, "step": 1698 }, { "epoch": 0.5485289963275354, "grad_norm": 0.4604439134790531, "learning_rate": 1.4288065864552767e-05, "loss": 0.355, "step": 1699 }, { "epoch": 0.5488518503571572, "grad_norm": 0.44464954886907515, "learning_rate": 1.4286930181762638e-05, "loss": 0.3769, "step": 1700 }, { "epoch": 0.5491747043867792, "grad_norm": 0.41090013038185486, "learning_rate": 1.4285793639077803e-05, "loss": 0.3544, "step": 1701 }, { "epoch": 0.549497558416401, "grad_norm": 0.5571666018192948, "learning_rate": 1.4284656236642254e-05, "loss": 0.4144, "step": 1702 }, { "epoch": 0.5498204124460229, "grad_norm": 0.34108489517941043, "learning_rate": 1.4283517974600105e-05, "loss": 0.3123, "step": 1703 }, { "epoch": 0.5501432664756447, "grad_norm": 0.44970009634432734, "learning_rate": 1.4282378853095569e-05, "loss": 0.3675, "step": 1704 }, { "epoch": 0.5504661205052666, "grad_norm": 0.41338617188813603, "learning_rate": 1.4281238872272974e-05, "loss": 0.3334, "step": 1705 }, { "epoch": 0.5507889745348884, "grad_norm": 0.43716557562442565, "learning_rate": 1.4280098032276751e-05, "loss": 0.3804, "step": 1706 }, { "epoch": 0.5511118285645102, "grad_norm": 0.5039151260450317, "learning_rate": 1.4278956333251445e-05, "loss": 0.369, "step": 1707 }, { "epoch": 0.5514346825941321, "grad_norm": 0.3991589843249707, "learning_rate": 1.4277813775341706e-05, "loss": 0.3533, "step": 1708 }, { "epoch": 0.5517575366237539, "grad_norm": 0.37950792350345103, "learning_rate": 1.4276670358692295e-05, "loss": 0.3424, "step": 1709 }, { "epoch": 0.5520803906533759, "grad_norm": 0.43952278451178717, "learning_rate": 1.4275526083448082e-05, "loss": 0.3711, "step": 1710 }, { "epoch": 0.5524032446829977, "grad_norm": 0.43355745114677685, "learning_rate": 1.4274380949754044e-05, "loss": 0.3533, "step": 1711 }, { "epoch": 0.5527260987126196, "grad_norm": 0.44282124488640656, "learning_rate": 1.4273234957755268e-05, "loss": 0.3498, "step": 1712 }, { "epoch": 0.5530489527422414, "grad_norm": 0.4361867884139567, "learning_rate": 1.4272088107596953e-05, "loss": 0.3779, "step": 1713 }, { "epoch": 0.5533718067718633, "grad_norm": 0.4130531740153131, "learning_rate": 1.4270940399424399e-05, "loss": 0.3261, "step": 1714 }, { "epoch": 0.5536946608014851, "grad_norm": 0.3856557439832635, "learning_rate": 1.426979183338302e-05, "loss": 0.3102, "step": 1715 }, { "epoch": 0.554017514831107, "grad_norm": 0.4150636129150073, "learning_rate": 1.426864240961834e-05, "loss": 0.349, "step": 1716 }, { "epoch": 0.5543403688607288, "grad_norm": 0.5056612733579322, "learning_rate": 1.4267492128275986e-05, "loss": 0.3228, "step": 1717 }, { "epoch": 0.5546632228903507, "grad_norm": 0.4175720007302167, "learning_rate": 1.4266340989501697e-05, "loss": 0.3719, "step": 1718 }, { "epoch": 0.5549860769199726, "grad_norm": 0.3839764576135057, "learning_rate": 1.4265188993441324e-05, "loss": 0.3856, "step": 1719 }, { "epoch": 0.5553089309495944, "grad_norm": 0.43420397028288754, "learning_rate": 1.4264036140240824e-05, "loss": 0.299, "step": 1720 }, { "epoch": 0.5556317849792163, "grad_norm": 0.583038469123561, "learning_rate": 1.4262882430046257e-05, "loss": 0.4039, "step": 1721 }, { "epoch": 0.5559546390088381, "grad_norm": 0.42336212753056085, "learning_rate": 1.42617278630038e-05, "loss": 0.3784, "step": 1722 }, { "epoch": 0.55627749303846, "grad_norm": 0.530448918479209, "learning_rate": 1.4260572439259737e-05, "loss": 0.3687, "step": 1723 }, { "epoch": 0.5566003470680818, "grad_norm": 0.45396875647218526, "learning_rate": 1.425941615896045e-05, "loss": 0.3729, "step": 1724 }, { "epoch": 0.5569232010977037, "grad_norm": 0.46367401866065827, "learning_rate": 1.4258259022252448e-05, "loss": 0.33, "step": 1725 }, { "epoch": 0.5572460551273256, "grad_norm": 0.5019653503629066, "learning_rate": 1.4257101029282333e-05, "loss": 0.3801, "step": 1726 }, { "epoch": 0.5575689091569475, "grad_norm": 0.4307751062860576, "learning_rate": 1.4255942180196824e-05, "loss": 0.3394, "step": 1727 }, { "epoch": 0.5578917631865693, "grad_norm": 0.48511440969497777, "learning_rate": 1.4254782475142742e-05, "loss": 0.3523, "step": 1728 }, { "epoch": 0.5582146172161911, "grad_norm": 0.44295816759511314, "learning_rate": 1.4253621914267024e-05, "loss": 0.3459, "step": 1729 }, { "epoch": 0.558537471245813, "grad_norm": 0.44629627696938856, "learning_rate": 1.425246049771671e-05, "loss": 0.3353, "step": 1730 }, { "epoch": 0.5588603252754348, "grad_norm": 0.4563204534523394, "learning_rate": 1.425129822563895e-05, "loss": 0.3744, "step": 1731 }, { "epoch": 0.5591831793050567, "grad_norm": 0.6603243633557287, "learning_rate": 1.4250135098181e-05, "loss": 0.4245, "step": 1732 }, { "epoch": 0.5595060333346785, "grad_norm": 0.44927143956894855, "learning_rate": 1.424897111549023e-05, "loss": 0.3606, "step": 1733 }, { "epoch": 0.5598288873643004, "grad_norm": 0.4978661312850878, "learning_rate": 1.4247806277714113e-05, "loss": 0.3315, "step": 1734 }, { "epoch": 0.5601517413939223, "grad_norm": 0.6592022327275922, "learning_rate": 1.4246640585000233e-05, "loss": 0.4024, "step": 1735 }, { "epoch": 0.5604745954235442, "grad_norm": 0.3405883317742189, "learning_rate": 1.4245474037496282e-05, "loss": 0.3596, "step": 1736 }, { "epoch": 0.560797449453166, "grad_norm": 0.6759845137869717, "learning_rate": 1.4244306635350059e-05, "loss": 0.3425, "step": 1737 }, { "epoch": 0.5611203034827879, "grad_norm": 0.4862934081447777, "learning_rate": 1.424313837870947e-05, "loss": 0.3664, "step": 1738 }, { "epoch": 0.5614431575124097, "grad_norm": 0.4293805794446071, "learning_rate": 1.4241969267722541e-05, "loss": 0.3902, "step": 1739 }, { "epoch": 0.5617660115420315, "grad_norm": 0.40995972856614127, "learning_rate": 1.4240799302537385e-05, "loss": 0.3577, "step": 1740 }, { "epoch": 0.5620888655716534, "grad_norm": 0.4481516622765057, "learning_rate": 1.4239628483302241e-05, "loss": 0.3449, "step": 1741 }, { "epoch": 0.5624117196012752, "grad_norm": 0.516063735973987, "learning_rate": 1.423845681016545e-05, "loss": 0.3387, "step": 1742 }, { "epoch": 0.5627345736308972, "grad_norm": 0.5259753899927205, "learning_rate": 1.423728428327546e-05, "loss": 0.3667, "step": 1743 }, { "epoch": 0.563057427660519, "grad_norm": 0.42000837009521785, "learning_rate": 1.4236110902780829e-05, "loss": 0.3732, "step": 1744 }, { "epoch": 0.5633802816901409, "grad_norm": 0.47663284692717106, "learning_rate": 1.4234936668830222e-05, "loss": 0.3254, "step": 1745 }, { "epoch": 0.5637031357197627, "grad_norm": 0.3840826107446079, "learning_rate": 1.4233761581572414e-05, "loss": 0.3157, "step": 1746 }, { "epoch": 0.5640259897493846, "grad_norm": 0.4217347962243966, "learning_rate": 1.4232585641156288e-05, "loss": 0.3343, "step": 1747 }, { "epoch": 0.5643488437790064, "grad_norm": 0.3955165694658186, "learning_rate": 1.4231408847730832e-05, "loss": 0.3289, "step": 1748 }, { "epoch": 0.5646716978086282, "grad_norm": 0.49312213361922513, "learning_rate": 1.4230231201445144e-05, "loss": 0.3766, "step": 1749 }, { "epoch": 0.5649945518382501, "grad_norm": 0.3776557180297843, "learning_rate": 1.422905270244843e-05, "loss": 0.4424, "step": 1750 }, { "epoch": 0.5653174058678719, "grad_norm": 0.4189288219635042, "learning_rate": 1.4227873350890006e-05, "loss": 0.4007, "step": 1751 }, { "epoch": 0.5656402598974939, "grad_norm": 0.42622995729291235, "learning_rate": 1.4226693146919294e-05, "loss": 0.3318, "step": 1752 }, { "epoch": 0.5659631139271157, "grad_norm": 0.37384508117843335, "learning_rate": 1.4225512090685823e-05, "loss": 0.3755, "step": 1753 }, { "epoch": 0.5662859679567376, "grad_norm": 0.40708793234645885, "learning_rate": 1.4224330182339233e-05, "loss": 0.3542, "step": 1754 }, { "epoch": 0.5666088219863594, "grad_norm": 0.5289805218172859, "learning_rate": 1.4223147422029266e-05, "loss": 0.3971, "step": 1755 }, { "epoch": 0.5669316760159813, "grad_norm": 0.3853277509600375, "learning_rate": 1.4221963809905781e-05, "loss": 0.3361, "step": 1756 }, { "epoch": 0.5672545300456031, "grad_norm": 0.3722258452892635, "learning_rate": 1.4220779346118738e-05, "loss": 0.3582, "step": 1757 }, { "epoch": 0.567577384075225, "grad_norm": 0.43171712453217853, "learning_rate": 1.4219594030818207e-05, "loss": 0.4539, "step": 1758 }, { "epoch": 0.5679002381048468, "grad_norm": 0.4215721873733276, "learning_rate": 1.4218407864154363e-05, "loss": 0.3538, "step": 1759 }, { "epoch": 0.5682230921344688, "grad_norm": 0.4164885639092685, "learning_rate": 1.4217220846277499e-05, "loss": 0.3237, "step": 1760 }, { "epoch": 0.5685459461640906, "grad_norm": 0.4105078067239255, "learning_rate": 1.4216032977338e-05, "loss": 0.3622, "step": 1761 }, { "epoch": 0.5688688001937124, "grad_norm": 0.4403972735804253, "learning_rate": 1.4214844257486373e-05, "loss": 0.3048, "step": 1762 }, { "epoch": 0.5691916542233343, "grad_norm": 0.44301684461519447, "learning_rate": 1.4213654686873225e-05, "loss": 0.3464, "step": 1763 }, { "epoch": 0.5695145082529561, "grad_norm": 0.4412738668464136, "learning_rate": 1.4212464265649273e-05, "loss": 0.3511, "step": 1764 }, { "epoch": 0.569837362282578, "grad_norm": 0.5156846589414786, "learning_rate": 1.4211272993965344e-05, "loss": 0.3618, "step": 1765 }, { "epoch": 0.5701602163121998, "grad_norm": 0.5067710746145705, "learning_rate": 1.4210080871972367e-05, "loss": 0.3501, "step": 1766 }, { "epoch": 0.5704830703418217, "grad_norm": 0.44115236961395926, "learning_rate": 1.4208887899821384e-05, "loss": 0.3964, "step": 1767 }, { "epoch": 0.5708059243714435, "grad_norm": 0.43063236006587174, "learning_rate": 1.4207694077663544e-05, "loss": 0.3501, "step": 1768 }, { "epoch": 0.5711287784010655, "grad_norm": 0.37185245968788216, "learning_rate": 1.42064994056501e-05, "loss": 0.293, "step": 1769 }, { "epoch": 0.5714516324306873, "grad_norm": 0.6776748591613874, "learning_rate": 1.4205303883932417e-05, "loss": 0.3735, "step": 1770 }, { "epoch": 0.5717744864603092, "grad_norm": 0.519150492874917, "learning_rate": 1.4204107512661966e-05, "loss": 0.4008, "step": 1771 }, { "epoch": 0.572097340489931, "grad_norm": 0.5763744424052207, "learning_rate": 1.4202910291990324e-05, "loss": 0.2906, "step": 1772 }, { "epoch": 0.5724201945195528, "grad_norm": 0.48904794997976836, "learning_rate": 1.420171222206918e-05, "loss": 0.3286, "step": 1773 }, { "epoch": 0.5727430485491747, "grad_norm": 0.5381283122444106, "learning_rate": 1.4200513303050326e-05, "loss": 0.4191, "step": 1774 }, { "epoch": 0.5730659025787965, "grad_norm": 0.5087745801336473, "learning_rate": 1.4199313535085663e-05, "loss": 0.2759, "step": 1775 }, { "epoch": 0.5733887566084184, "grad_norm": 0.435646632278564, "learning_rate": 1.41981129183272e-05, "loss": 0.3335, "step": 1776 }, { "epoch": 0.5737116106380403, "grad_norm": 0.4420366814317685, "learning_rate": 1.4196911452927056e-05, "loss": 0.3549, "step": 1777 }, { "epoch": 0.5740344646676622, "grad_norm": 0.47621875402955527, "learning_rate": 1.4195709139037455e-05, "loss": 0.4229, "step": 1778 }, { "epoch": 0.574357318697284, "grad_norm": 0.9306818980425506, "learning_rate": 1.4194505976810722e-05, "loss": 0.3383, "step": 1779 }, { "epoch": 0.5746801727269059, "grad_norm": 0.4080793122220823, "learning_rate": 1.4193301966399303e-05, "loss": 0.3844, "step": 1780 }, { "epoch": 0.5750030267565277, "grad_norm": 0.6459091981749964, "learning_rate": 1.4192097107955741e-05, "loss": 0.4072, "step": 1781 }, { "epoch": 0.5753258807861495, "grad_norm": 0.41552927554898395, "learning_rate": 1.419089140163269e-05, "loss": 0.3805, "step": 1782 }, { "epoch": 0.5756487348157714, "grad_norm": 0.44943037405300634, "learning_rate": 1.4189684847582916e-05, "loss": 0.3717, "step": 1783 }, { "epoch": 0.5759715888453932, "grad_norm": 0.44096624791946526, "learning_rate": 1.4188477445959281e-05, "loss": 0.3235, "step": 1784 }, { "epoch": 0.5762944428750152, "grad_norm": 0.36048434204998375, "learning_rate": 1.4187269196914766e-05, "loss": 0.3418, "step": 1785 }, { "epoch": 0.576617296904637, "grad_norm": 0.42620073638112915, "learning_rate": 1.4186060100602452e-05, "loss": 0.365, "step": 1786 }, { "epoch": 0.5769401509342589, "grad_norm": 0.3835066916476059, "learning_rate": 1.418485015717553e-05, "loss": 0.3965, "step": 1787 }, { "epoch": 0.5772630049638807, "grad_norm": 0.35110455132687884, "learning_rate": 1.41836393667873e-05, "loss": 0.359, "step": 1788 }, { "epoch": 0.5775858589935026, "grad_norm": 0.5084091849799208, "learning_rate": 1.4182427729591166e-05, "loss": 0.3452, "step": 1789 }, { "epoch": 0.5779087130231244, "grad_norm": 0.3906031779039764, "learning_rate": 1.4181215245740638e-05, "loss": 0.3684, "step": 1790 }, { "epoch": 0.5782315670527463, "grad_norm": 0.5575946295840669, "learning_rate": 1.418000191538934e-05, "loss": 0.3948, "step": 1791 }, { "epoch": 0.5785544210823681, "grad_norm": 0.5928204257762274, "learning_rate": 1.4178787738691001e-05, "loss": 0.3613, "step": 1792 }, { "epoch": 0.5788772751119899, "grad_norm": 0.3846572529228678, "learning_rate": 1.417757271579945e-05, "loss": 0.3901, "step": 1793 }, { "epoch": 0.5792001291416119, "grad_norm": 0.44570922804951485, "learning_rate": 1.4176356846868633e-05, "loss": 0.3992, "step": 1794 }, { "epoch": 0.5795229831712337, "grad_norm": 0.5218174331139511, "learning_rate": 1.4175140132052596e-05, "loss": 0.3766, "step": 1795 }, { "epoch": 0.5798458372008556, "grad_norm": 0.6505980062905442, "learning_rate": 1.4173922571505498e-05, "loss": 0.3743, "step": 1796 }, { "epoch": 0.5801686912304774, "grad_norm": 0.4296577872625817, "learning_rate": 1.41727041653816e-05, "loss": 0.3489, "step": 1797 }, { "epoch": 0.5804915452600993, "grad_norm": 0.41917535681636997, "learning_rate": 1.4171484913835274e-05, "loss": 0.3297, "step": 1798 }, { "epoch": 0.5808143992897211, "grad_norm": 0.46166256223588076, "learning_rate": 1.4170264817020997e-05, "loss": 0.3725, "step": 1799 }, { "epoch": 0.581137253319343, "grad_norm": 0.4293627578155449, "learning_rate": 1.416904387509335e-05, "loss": 0.3795, "step": 1800 }, { "epoch": 0.5814601073489648, "grad_norm": 0.41951315706235065, "learning_rate": 1.4167822088207032e-05, "loss": 0.387, "step": 1801 }, { "epoch": 0.5817829613785868, "grad_norm": 0.3766578753202241, "learning_rate": 1.4166599456516835e-05, "loss": 0.3523, "step": 1802 }, { "epoch": 0.5821058154082086, "grad_norm": 0.38326210934990224, "learning_rate": 1.4165375980177666e-05, "loss": 0.3067, "step": 1803 }, { "epoch": 0.5824286694378304, "grad_norm": 0.43503288042905897, "learning_rate": 1.4164151659344542e-05, "loss": 0.3665, "step": 1804 }, { "epoch": 0.5827515234674523, "grad_norm": 0.6551941596878981, "learning_rate": 1.4162926494172578e-05, "loss": 0.3326, "step": 1805 }, { "epoch": 0.5830743774970741, "grad_norm": 0.4124665678178372, "learning_rate": 1.4161700484817004e-05, "loss": 0.3443, "step": 1806 }, { "epoch": 0.583397231526696, "grad_norm": 0.3727025382675416, "learning_rate": 1.4160473631433149e-05, "loss": 0.3479, "step": 1807 }, { "epoch": 0.5837200855563178, "grad_norm": 0.4588617103401732, "learning_rate": 1.415924593417646e-05, "loss": 0.3467, "step": 1808 }, { "epoch": 0.5840429395859397, "grad_norm": 0.467929251871753, "learning_rate": 1.4158017393202481e-05, "loss": 0.3709, "step": 1809 }, { "epoch": 0.5843657936155615, "grad_norm": 0.441027300726436, "learning_rate": 1.4156788008666868e-05, "loss": 0.4284, "step": 1810 }, { "epoch": 0.5846886476451835, "grad_norm": 0.43076175818860196, "learning_rate": 1.415555778072538e-05, "loss": 0.3438, "step": 1811 }, { "epoch": 0.5850115016748053, "grad_norm": 0.5156532045011726, "learning_rate": 1.4154326709533886e-05, "loss": 0.4037, "step": 1812 }, { "epoch": 0.5853343557044272, "grad_norm": 1.2069124839972865, "learning_rate": 1.4153094795248362e-05, "loss": 0.3877, "step": 1813 }, { "epoch": 0.585657209734049, "grad_norm": 0.4503651898288058, "learning_rate": 1.415186203802489e-05, "loss": 0.406, "step": 1814 }, { "epoch": 0.5859800637636708, "grad_norm": 0.46272836546477286, "learning_rate": 1.4150628438019658e-05, "loss": 0.4075, "step": 1815 }, { "epoch": 0.5863029177932927, "grad_norm": 0.3595797001618569, "learning_rate": 1.4149393995388958e-05, "loss": 0.2979, "step": 1816 }, { "epoch": 0.5866257718229145, "grad_norm": 0.4170617000607737, "learning_rate": 1.4148158710289198e-05, "loss": 0.3308, "step": 1817 }, { "epoch": 0.5869486258525364, "grad_norm": 0.4050296851944654, "learning_rate": 1.4146922582876885e-05, "loss": 0.4214, "step": 1818 }, { "epoch": 0.5872714798821583, "grad_norm": 0.4931993480077737, "learning_rate": 1.4145685613308634e-05, "loss": 0.3769, "step": 1819 }, { "epoch": 0.5875943339117802, "grad_norm": 0.6091949049115247, "learning_rate": 1.4144447801741165e-05, "loss": 0.3378, "step": 1820 }, { "epoch": 0.587917187941402, "grad_norm": 0.32614114117284415, "learning_rate": 1.4143209148331314e-05, "loss": 0.3463, "step": 1821 }, { "epoch": 0.5882400419710239, "grad_norm": 0.42486173121905474, "learning_rate": 1.414196965323601e-05, "loss": 0.4415, "step": 1822 }, { "epoch": 0.5885628960006457, "grad_norm": 0.4137718059949814, "learning_rate": 1.4140729316612297e-05, "loss": 0.3998, "step": 1823 }, { "epoch": 0.5888857500302676, "grad_norm": 0.4209537734572589, "learning_rate": 1.4139488138617324e-05, "loss": 0.3249, "step": 1824 }, { "epoch": 0.5892086040598894, "grad_norm": 0.4841014178862226, "learning_rate": 1.413824611940835e-05, "loss": 0.2886, "step": 1825 }, { "epoch": 0.5895314580895112, "grad_norm": 0.3736007160642266, "learning_rate": 1.4137003259142729e-05, "loss": 0.402, "step": 1826 }, { "epoch": 0.5898543121191331, "grad_norm": 0.4735655006865349, "learning_rate": 1.4135759557977938e-05, "loss": 0.3431, "step": 1827 }, { "epoch": 0.590177166148755, "grad_norm": 0.467100358842273, "learning_rate": 1.4134515016071548e-05, "loss": 0.4199, "step": 1828 }, { "epoch": 0.5905000201783769, "grad_norm": 0.3602871801105927, "learning_rate": 1.4133269633581242e-05, "loss": 0.3656, "step": 1829 }, { "epoch": 0.5908228742079987, "grad_norm": 0.462146103685933, "learning_rate": 1.413202341066481e-05, "loss": 0.3597, "step": 1830 }, { "epoch": 0.5911457282376206, "grad_norm": 0.4136226353992495, "learning_rate": 1.4130776347480142e-05, "loss": 0.3444, "step": 1831 }, { "epoch": 0.5914685822672424, "grad_norm": 0.45704123600815505, "learning_rate": 1.4129528444185242e-05, "loss": 0.3516, "step": 1832 }, { "epoch": 0.5917914362968643, "grad_norm": 0.4869963399911004, "learning_rate": 1.4128279700938218e-05, "loss": 0.3484, "step": 1833 }, { "epoch": 0.5921142903264861, "grad_norm": 0.40793394301777935, "learning_rate": 1.4127030117897285e-05, "loss": 0.3462, "step": 1834 }, { "epoch": 0.592437144356108, "grad_norm": 0.4133710794766472, "learning_rate": 1.412577969522076e-05, "loss": 0.4076, "step": 1835 }, { "epoch": 0.5927599983857299, "grad_norm": 0.417850556900677, "learning_rate": 1.4124528433067073e-05, "loss": 0.3584, "step": 1836 }, { "epoch": 0.5930828524153517, "grad_norm": 0.4236235547699879, "learning_rate": 1.4123276331594756e-05, "loss": 0.4254, "step": 1837 }, { "epoch": 0.5934057064449736, "grad_norm": 0.4457808234362453, "learning_rate": 1.4122023390962447e-05, "loss": 0.3186, "step": 1838 }, { "epoch": 0.5937285604745954, "grad_norm": 0.41436626497206186, "learning_rate": 1.4120769611328896e-05, "loss": 0.4056, "step": 1839 }, { "epoch": 0.5940514145042173, "grad_norm": 0.401030537893304, "learning_rate": 1.411951499285295e-05, "loss": 0.3069, "step": 1840 }, { "epoch": 0.5943742685338391, "grad_norm": 0.4851068566364305, "learning_rate": 1.4118259535693573e-05, "loss": 0.3925, "step": 1841 }, { "epoch": 0.594697122563461, "grad_norm": 0.5084541221127, "learning_rate": 1.4117003240009825e-05, "loss": 0.3633, "step": 1842 }, { "epoch": 0.5950199765930828, "grad_norm": 0.40046032637861506, "learning_rate": 1.4115746105960878e-05, "loss": 0.3486, "step": 1843 }, { "epoch": 0.5953428306227047, "grad_norm": 0.49649714133257283, "learning_rate": 1.411448813370601e-05, "loss": 0.3685, "step": 1844 }, { "epoch": 0.5956656846523266, "grad_norm": 0.5386254187437844, "learning_rate": 1.4113229323404606e-05, "loss": 0.2846, "step": 1845 }, { "epoch": 0.5959885386819485, "grad_norm": 0.4919717121163362, "learning_rate": 1.4111969675216151e-05, "loss": 0.3799, "step": 1846 }, { "epoch": 0.5963113927115703, "grad_norm": 0.6569149293533763, "learning_rate": 1.4110709189300246e-05, "loss": 0.4039, "step": 1847 }, { "epoch": 0.5966342467411921, "grad_norm": 0.38322128895657065, "learning_rate": 1.4109447865816588e-05, "loss": 0.4182, "step": 1848 }, { "epoch": 0.596957100770814, "grad_norm": 0.592590432045332, "learning_rate": 1.410818570492499e-05, "loss": 0.3626, "step": 1849 }, { "epoch": 0.5972799548004358, "grad_norm": 0.44867527604837526, "learning_rate": 1.4106922706785361e-05, "loss": 0.3832, "step": 1850 }, { "epoch": 0.5976028088300577, "grad_norm": 0.6015445624122377, "learning_rate": 1.4105658871557722e-05, "loss": 0.3738, "step": 1851 }, { "epoch": 0.5979256628596795, "grad_norm": 0.4579872229620051, "learning_rate": 1.4104394199402203e-05, "loss": 0.3316, "step": 1852 }, { "epoch": 0.5982485168893015, "grad_norm": 0.4673993037793287, "learning_rate": 1.4103128690479033e-05, "loss": 0.3012, "step": 1853 }, { "epoch": 0.5985713709189233, "grad_norm": 1.0420882895900656, "learning_rate": 1.4101862344948551e-05, "loss": 0.3869, "step": 1854 }, { "epoch": 0.5988942249485452, "grad_norm": 0.4831770547414068, "learning_rate": 1.4100595162971199e-05, "loss": 0.3181, "step": 1855 }, { "epoch": 0.599217078978167, "grad_norm": 0.39505145474177256, "learning_rate": 1.4099327144707533e-05, "loss": 0.3968, "step": 1856 }, { "epoch": 0.5995399330077889, "grad_norm": 0.4586390392357517, "learning_rate": 1.4098058290318203e-05, "loss": 0.3698, "step": 1857 }, { "epoch": 0.5998627870374107, "grad_norm": 0.4291008228496865, "learning_rate": 1.4096788599963974e-05, "loss": 0.3563, "step": 1858 }, { "epoch": 0.6001856410670325, "grad_norm": 0.4815563871173544, "learning_rate": 1.4095518073805716e-05, "loss": 0.3393, "step": 1859 }, { "epoch": 0.6005084950966544, "grad_norm": 0.6174195347713055, "learning_rate": 1.40942467120044e-05, "loss": 0.3357, "step": 1860 }, { "epoch": 0.6008313491262763, "grad_norm": 0.3751067624333864, "learning_rate": 1.4092974514721103e-05, "loss": 0.378, "step": 1861 }, { "epoch": 0.6011542031558982, "grad_norm": 0.391667360000119, "learning_rate": 1.4091701482117017e-05, "loss": 0.3461, "step": 1862 }, { "epoch": 0.60147705718552, "grad_norm": 0.33212201700546906, "learning_rate": 1.409042761435343e-05, "loss": 0.3398, "step": 1863 }, { "epoch": 0.6017999112151419, "grad_norm": 0.37962042309498323, "learning_rate": 1.408915291159174e-05, "loss": 0.2834, "step": 1864 }, { "epoch": 0.6021227652447637, "grad_norm": 0.4298701460739722, "learning_rate": 1.408787737399345e-05, "loss": 0.4309, "step": 1865 }, { "epoch": 0.6024456192743856, "grad_norm": 0.5728191748542546, "learning_rate": 1.4086601001720169e-05, "loss": 0.3483, "step": 1866 }, { "epoch": 0.6027684733040074, "grad_norm": 0.430272927042395, "learning_rate": 1.408532379493361e-05, "loss": 0.3301, "step": 1867 }, { "epoch": 0.6030913273336292, "grad_norm": 0.5315378088894903, "learning_rate": 1.4084045753795596e-05, "loss": 0.4119, "step": 1868 }, { "epoch": 0.6034141813632511, "grad_norm": 0.45404200634852426, "learning_rate": 1.4082766878468051e-05, "loss": 0.3294, "step": 1869 }, { "epoch": 0.603737035392873, "grad_norm": 0.48552004035353974, "learning_rate": 1.4081487169113008e-05, "loss": 0.3614, "step": 1870 }, { "epoch": 0.6040598894224949, "grad_norm": 0.35360894584364166, "learning_rate": 1.4080206625892605e-05, "loss": 0.3118, "step": 1871 }, { "epoch": 0.6043827434521167, "grad_norm": 0.4417110098972817, "learning_rate": 1.4078925248969084e-05, "loss": 0.3363, "step": 1872 }, { "epoch": 0.6047055974817386, "grad_norm": 0.4907413363395902, "learning_rate": 1.407764303850479e-05, "loss": 0.3524, "step": 1873 }, { "epoch": 0.6050284515113604, "grad_norm": 0.5079482034748564, "learning_rate": 1.4076359994662183e-05, "loss": 0.3747, "step": 1874 }, { "epoch": 0.6053513055409823, "grad_norm": 0.43572388592291367, "learning_rate": 1.4075076117603822e-05, "loss": 0.3356, "step": 1875 }, { "epoch": 0.6056741595706041, "grad_norm": 0.4113237847629259, "learning_rate": 1.4073791407492371e-05, "loss": 0.3321, "step": 1876 }, { "epoch": 0.605997013600226, "grad_norm": 0.7021366983383085, "learning_rate": 1.4072505864490601e-05, "loss": 0.3285, "step": 1877 }, { "epoch": 0.6063198676298479, "grad_norm": 0.4271956857718999, "learning_rate": 1.4071219488761389e-05, "loss": 0.3367, "step": 1878 }, { "epoch": 0.6066427216594698, "grad_norm": 0.3975799652620438, "learning_rate": 1.4069932280467718e-05, "loss": 0.4034, "step": 1879 }, { "epoch": 0.6069655756890916, "grad_norm": 0.4052133295958764, "learning_rate": 1.4068644239772672e-05, "loss": 0.3159, "step": 1880 }, { "epoch": 0.6072884297187134, "grad_norm": 0.43395411229516695, "learning_rate": 1.4067355366839446e-05, "loss": 0.3163, "step": 1881 }, { "epoch": 0.6076112837483353, "grad_norm": 0.3928388460631131, "learning_rate": 1.4066065661831341e-05, "loss": 0.3167, "step": 1882 }, { "epoch": 0.6079341377779571, "grad_norm": 0.5038061565874861, "learning_rate": 1.4064775124911759e-05, "loss": 0.3932, "step": 1883 }, { "epoch": 0.608256991807579, "grad_norm": 0.3888042653680441, "learning_rate": 1.4063483756244207e-05, "loss": 0.4068, "step": 1884 }, { "epoch": 0.6085798458372008, "grad_norm": 0.7750635573662372, "learning_rate": 1.4062191555992302e-05, "loss": 0.4513, "step": 1885 }, { "epoch": 0.6089026998668227, "grad_norm": 0.47569957174549454, "learning_rate": 1.4060898524319765e-05, "loss": 0.3386, "step": 1886 }, { "epoch": 0.6092255538964446, "grad_norm": 0.42582372242906513, "learning_rate": 1.4059604661390418e-05, "loss": 0.376, "step": 1887 }, { "epoch": 0.6095484079260665, "grad_norm": 0.5842484408908171, "learning_rate": 1.4058309967368194e-05, "loss": 0.3254, "step": 1888 }, { "epoch": 0.6098712619556883, "grad_norm": 0.4886185572248802, "learning_rate": 1.405701444241713e-05, "loss": 0.3397, "step": 1889 }, { "epoch": 0.6101941159853101, "grad_norm": 0.4444044631500749, "learning_rate": 1.4055718086701362e-05, "loss": 0.3484, "step": 1890 }, { "epoch": 0.610516970014932, "grad_norm": 0.3789294953717436, "learning_rate": 1.4054420900385144e-05, "loss": 0.3873, "step": 1891 }, { "epoch": 0.6108398240445538, "grad_norm": 0.3948747480772556, "learning_rate": 1.405312288363282e-05, "loss": 0.3264, "step": 1892 }, { "epoch": 0.6111626780741757, "grad_norm": 0.4559237958902669, "learning_rate": 1.4051824036608856e-05, "loss": 0.3731, "step": 1893 }, { "epoch": 0.6114855321037975, "grad_norm": 0.5592934451790066, "learning_rate": 1.4050524359477806e-05, "loss": 0.4424, "step": 1894 }, { "epoch": 0.6118083861334195, "grad_norm": 0.45225504193201327, "learning_rate": 1.4049223852404341e-05, "loss": 0.3, "step": 1895 }, { "epoch": 0.6121312401630413, "grad_norm": 0.44378757970983046, "learning_rate": 1.4047922515553232e-05, "loss": 0.3867, "step": 1896 }, { "epoch": 0.6124540941926632, "grad_norm": 0.42626763899743897, "learning_rate": 1.4046620349089356e-05, "loss": 0.3876, "step": 1897 }, { "epoch": 0.612776948222285, "grad_norm": 0.42487272984302754, "learning_rate": 1.40453173531777e-05, "loss": 0.3921, "step": 1898 }, { "epoch": 0.6130998022519069, "grad_norm": 0.3792191035491863, "learning_rate": 1.4044013527983346e-05, "loss": 0.3251, "step": 1899 }, { "epoch": 0.6134226562815287, "grad_norm": 0.3984897946576864, "learning_rate": 1.4042708873671492e-05, "loss": 0.3302, "step": 1900 }, { "epoch": 0.6137455103111505, "grad_norm": 0.49471559866275017, "learning_rate": 1.4041403390407434e-05, "loss": 0.3705, "step": 1901 }, { "epoch": 0.6140683643407724, "grad_norm": 0.5781140291793466, "learning_rate": 1.4040097078356573e-05, "loss": 0.4031, "step": 1902 }, { "epoch": 0.6143912183703942, "grad_norm": 0.41028352086365577, "learning_rate": 1.4038789937684417e-05, "loss": 0.4311, "step": 1903 }, { "epoch": 0.6147140724000162, "grad_norm": 0.46200388553566246, "learning_rate": 1.4037481968556582e-05, "loss": 0.4133, "step": 1904 }, { "epoch": 0.615036926429638, "grad_norm": 0.49340319065880683, "learning_rate": 1.4036173171138786e-05, "loss": 0.3481, "step": 1905 }, { "epoch": 0.6153597804592599, "grad_norm": 0.5334832570748534, "learning_rate": 1.4034863545596846e-05, "loss": 0.3672, "step": 1906 }, { "epoch": 0.6156826344888817, "grad_norm": 0.40605382901340464, "learning_rate": 1.4033553092096697e-05, "loss": 0.346, "step": 1907 }, { "epoch": 0.6160054885185036, "grad_norm": 0.45518820283717326, "learning_rate": 1.4032241810804367e-05, "loss": 0.3536, "step": 1908 }, { "epoch": 0.6163283425481254, "grad_norm": 0.5063683497413286, "learning_rate": 1.4030929701885995e-05, "loss": 0.3999, "step": 1909 }, { "epoch": 0.6166511965777473, "grad_norm": 0.4908940885640825, "learning_rate": 1.4029616765507823e-05, "loss": 0.3604, "step": 1910 }, { "epoch": 0.6169740506073691, "grad_norm": 0.418104286168778, "learning_rate": 1.4028303001836199e-05, "loss": 0.3107, "step": 1911 }, { "epoch": 0.617296904636991, "grad_norm": 0.5254583737899582, "learning_rate": 1.4026988411037576e-05, "loss": 0.3921, "step": 1912 }, { "epoch": 0.6176197586666129, "grad_norm": 0.3502275267183345, "learning_rate": 1.402567299327851e-05, "loss": 0.3057, "step": 1913 }, { "epoch": 0.6179426126962347, "grad_norm": 0.6086189447347491, "learning_rate": 1.402435674872566e-05, "loss": 0.3239, "step": 1914 }, { "epoch": 0.6182654667258566, "grad_norm": 0.5803937655608467, "learning_rate": 1.4023039677545794e-05, "loss": 0.3592, "step": 1915 }, { "epoch": 0.6185883207554784, "grad_norm": 0.40065821125543405, "learning_rate": 1.4021721779905785e-05, "loss": 0.3644, "step": 1916 }, { "epoch": 0.6189111747851003, "grad_norm": 0.4481526055724962, "learning_rate": 1.402040305597261e-05, "loss": 0.3995, "step": 1917 }, { "epoch": 0.6192340288147221, "grad_norm": 0.5003034622155641, "learning_rate": 1.4019083505913343e-05, "loss": 0.3089, "step": 1918 }, { "epoch": 0.619556882844344, "grad_norm": 0.35156410460564025, "learning_rate": 1.4017763129895174e-05, "loss": 0.319, "step": 1919 }, { "epoch": 0.6198797368739658, "grad_norm": 0.3933144931349339, "learning_rate": 1.4016441928085392e-05, "loss": 0.302, "step": 1920 }, { "epoch": 0.6202025909035878, "grad_norm": 0.46229583901530624, "learning_rate": 1.4015119900651393e-05, "loss": 0.3101, "step": 1921 }, { "epoch": 0.6205254449332096, "grad_norm": 0.3834623623067952, "learning_rate": 1.401379704776067e-05, "loss": 0.2935, "step": 1922 }, { "epoch": 0.6208482989628314, "grad_norm": 0.4372781660812456, "learning_rate": 1.4012473369580835e-05, "loss": 0.4016, "step": 1923 }, { "epoch": 0.6211711529924533, "grad_norm": 0.3582465526509069, "learning_rate": 1.4011148866279592e-05, "loss": 0.2813, "step": 1924 }, { "epoch": 0.6214940070220751, "grad_norm": 0.5149932384259976, "learning_rate": 1.4009823538024754e-05, "loss": 0.3436, "step": 1925 }, { "epoch": 0.621816861051697, "grad_norm": 0.3731123348384239, "learning_rate": 1.4008497384984238e-05, "loss": 0.4239, "step": 1926 }, { "epoch": 0.6221397150813188, "grad_norm": 0.5418159901448661, "learning_rate": 1.4007170407326067e-05, "loss": 0.3961, "step": 1927 }, { "epoch": 0.6224625691109407, "grad_norm": 0.4052227437700352, "learning_rate": 1.4005842605218365e-05, "loss": 0.4097, "step": 1928 }, { "epoch": 0.6227854231405626, "grad_norm": 0.5362314149912742, "learning_rate": 1.4004513978829365e-05, "loss": 0.3613, "step": 1929 }, { "epoch": 0.6231082771701845, "grad_norm": 0.48001956230131915, "learning_rate": 1.4003184528327398e-05, "loss": 0.3618, "step": 1930 }, { "epoch": 0.6234311311998063, "grad_norm": 0.4344456362375171, "learning_rate": 1.400185425388091e-05, "loss": 0.4161, "step": 1931 }, { "epoch": 0.6237539852294282, "grad_norm": 0.43730944781164965, "learning_rate": 1.400052315565844e-05, "loss": 0.3493, "step": 1932 }, { "epoch": 0.62407683925905, "grad_norm": 0.5329545781772764, "learning_rate": 1.399919123382864e-05, "loss": 0.374, "step": 1933 }, { "epoch": 0.6243996932886718, "grad_norm": 0.3967043841873316, "learning_rate": 1.3997858488560258e-05, "loss": 0.3709, "step": 1934 }, { "epoch": 0.6247225473182937, "grad_norm": 0.5305154400380359, "learning_rate": 1.3996524920022154e-05, "loss": 0.3968, "step": 1935 }, { "epoch": 0.6250454013479155, "grad_norm": 0.4597616711430405, "learning_rate": 1.3995190528383292e-05, "loss": 0.3606, "step": 1936 }, { "epoch": 0.6253682553775375, "grad_norm": 0.40573444057406377, "learning_rate": 1.399385531381273e-05, "loss": 0.4223, "step": 1937 }, { "epoch": 0.6256911094071593, "grad_norm": 0.4472711817122073, "learning_rate": 1.3992519276479644e-05, "loss": 0.3393, "step": 1938 }, { "epoch": 0.6260139634367812, "grad_norm": 0.42880559374516536, "learning_rate": 1.3991182416553307e-05, "loss": 0.3756, "step": 1939 }, { "epoch": 0.626336817466403, "grad_norm": 0.4053715178011058, "learning_rate": 1.3989844734203097e-05, "loss": 0.3772, "step": 1940 }, { "epoch": 0.6266596714960249, "grad_norm": 0.44968189067972375, "learning_rate": 1.3988506229598495e-05, "loss": 0.3517, "step": 1941 }, { "epoch": 0.6269825255256467, "grad_norm": 0.4099628813354539, "learning_rate": 1.3987166902909091e-05, "loss": 0.3367, "step": 1942 }, { "epoch": 0.6273053795552685, "grad_norm": 0.3998670062314393, "learning_rate": 1.3985826754304573e-05, "loss": 0.3936, "step": 1943 }, { "epoch": 0.6276282335848904, "grad_norm": 0.426818305308399, "learning_rate": 1.3984485783954734e-05, "loss": 0.3034, "step": 1944 }, { "epoch": 0.6279510876145122, "grad_norm": 0.3644433209316346, "learning_rate": 1.398314399202948e-05, "loss": 0.3094, "step": 1945 }, { "epoch": 0.6282739416441342, "grad_norm": 0.5268512529150392, "learning_rate": 1.3981801378698807e-05, "loss": 0.392, "step": 1946 }, { "epoch": 0.628596795673756, "grad_norm": 0.4122330984344811, "learning_rate": 1.3980457944132825e-05, "loss": 0.2789, "step": 1947 }, { "epoch": 0.6289196497033779, "grad_norm": 0.4355321000703817, "learning_rate": 1.397911368850175e-05, "loss": 0.3738, "step": 1948 }, { "epoch": 0.6292425037329997, "grad_norm": 0.41251623204310756, "learning_rate": 1.3977768611975888e-05, "loss": 0.3502, "step": 1949 }, { "epoch": 0.6295653577626216, "grad_norm": 0.34829679123835755, "learning_rate": 1.3976422714725665e-05, "loss": 0.3051, "step": 1950 }, { "epoch": 0.6298882117922434, "grad_norm": 0.5643747092080513, "learning_rate": 1.3975075996921603e-05, "loss": 0.3268, "step": 1951 }, { "epoch": 0.6302110658218653, "grad_norm": 0.38669838344978313, "learning_rate": 1.397372845873433e-05, "loss": 0.3729, "step": 1952 }, { "epoch": 0.6305339198514871, "grad_norm": 0.4582049941309255, "learning_rate": 1.3972380100334573e-05, "loss": 0.3816, "step": 1953 }, { "epoch": 0.6308567738811091, "grad_norm": 0.3954125406267551, "learning_rate": 1.3971030921893173e-05, "loss": 0.4638, "step": 1954 }, { "epoch": 0.6311796279107309, "grad_norm": 0.38977984824879053, "learning_rate": 1.3969680923581064e-05, "loss": 0.3707, "step": 1955 }, { "epoch": 0.6315024819403527, "grad_norm": 0.5027892395674285, "learning_rate": 1.3968330105569294e-05, "loss": 0.3746, "step": 1956 }, { "epoch": 0.6318253359699746, "grad_norm": 0.5964255835933675, "learning_rate": 1.3966978468029005e-05, "loss": 0.3666, "step": 1957 }, { "epoch": 0.6321481899995964, "grad_norm": 0.44396842363072286, "learning_rate": 1.3965626011131451e-05, "loss": 0.3611, "step": 1958 }, { "epoch": 0.6324710440292183, "grad_norm": 0.46371909494650976, "learning_rate": 1.3964272735047985e-05, "loss": 0.3772, "step": 1959 }, { "epoch": 0.6327938980588401, "grad_norm": 0.3964556984838059, "learning_rate": 1.3962918639950064e-05, "loss": 0.3291, "step": 1960 }, { "epoch": 0.633116752088462, "grad_norm": 0.3677516059229403, "learning_rate": 1.3961563726009252e-05, "loss": 0.3094, "step": 1961 }, { "epoch": 0.6334396061180838, "grad_norm": 0.3841575861994609, "learning_rate": 1.3960207993397213e-05, "loss": 0.3493, "step": 1962 }, { "epoch": 0.6337624601477058, "grad_norm": 0.38061827667664366, "learning_rate": 1.3958851442285718e-05, "loss": 0.3458, "step": 1963 }, { "epoch": 0.6340853141773276, "grad_norm": 0.3538421605343147, "learning_rate": 1.3957494072846641e-05, "loss": 0.3323, "step": 1964 }, { "epoch": 0.6344081682069495, "grad_norm": 0.3365585035707792, "learning_rate": 1.3956135885251956e-05, "loss": 0.3381, "step": 1965 }, { "epoch": 0.6347310222365713, "grad_norm": 0.4350745141314514, "learning_rate": 1.3954776879673748e-05, "loss": 0.3499, "step": 1966 }, { "epoch": 0.6350538762661931, "grad_norm": 0.39078875518342154, "learning_rate": 1.3953417056284193e-05, "loss": 0.3646, "step": 1967 }, { "epoch": 0.635376730295815, "grad_norm": 0.4566909634430266, "learning_rate": 1.3952056415255588e-05, "loss": 0.3632, "step": 1968 }, { "epoch": 0.6356995843254368, "grad_norm": 0.4199949715736395, "learning_rate": 1.3950694956760318e-05, "loss": 0.3981, "step": 1969 }, { "epoch": 0.6360224383550587, "grad_norm": 0.4344319882677067, "learning_rate": 1.394933268097088e-05, "loss": 0.3414, "step": 1970 }, { "epoch": 0.6363452923846806, "grad_norm": 0.43891055548289154, "learning_rate": 1.3947969588059875e-05, "loss": 0.4004, "step": 1971 }, { "epoch": 0.6366681464143025, "grad_norm": 0.43879478057771126, "learning_rate": 1.39466056782e-05, "loss": 0.403, "step": 1972 }, { "epoch": 0.6369910004439243, "grad_norm": 0.3240012121135603, "learning_rate": 1.3945240951564065e-05, "loss": 0.3717, "step": 1973 }, { "epoch": 0.6373138544735462, "grad_norm": 0.4739815213452463, "learning_rate": 1.3943875408324977e-05, "loss": 0.4088, "step": 1974 }, { "epoch": 0.637636708503168, "grad_norm": 0.45773181754689773, "learning_rate": 1.3942509048655746e-05, "loss": 0.4478, "step": 1975 }, { "epoch": 0.6379595625327898, "grad_norm": 0.45971268780691676, "learning_rate": 1.3941141872729492e-05, "loss": 0.3246, "step": 1976 }, { "epoch": 0.6382824165624117, "grad_norm": 0.4099841505683666, "learning_rate": 1.3939773880719433e-05, "loss": 0.3691, "step": 1977 }, { "epoch": 0.6386052705920335, "grad_norm": 0.4821511926138039, "learning_rate": 1.3938405072798891e-05, "loss": 0.3596, "step": 1978 }, { "epoch": 0.6389281246216554, "grad_norm": 0.4197633984791537, "learning_rate": 1.3937035449141293e-05, "loss": 0.3615, "step": 1979 }, { "epoch": 0.6392509786512773, "grad_norm": 0.41596087243054425, "learning_rate": 1.3935665009920169e-05, "loss": 0.3342, "step": 1980 }, { "epoch": 0.6395738326808992, "grad_norm": 0.45757541186814327, "learning_rate": 1.393429375530915e-05, "loss": 0.3415, "step": 1981 }, { "epoch": 0.639896686710521, "grad_norm": 0.38421085931739557, "learning_rate": 1.3932921685481971e-05, "loss": 0.4331, "step": 1982 }, { "epoch": 0.6402195407401429, "grad_norm": 0.7794306348088293, "learning_rate": 1.3931548800612474e-05, "loss": 0.3347, "step": 1983 }, { "epoch": 0.6405423947697647, "grad_norm": 0.4064395006052289, "learning_rate": 1.3930175100874601e-05, "loss": 0.3862, "step": 1984 }, { "epoch": 0.6408652487993866, "grad_norm": 0.38018566436161616, "learning_rate": 1.3928800586442397e-05, "loss": 0.3335, "step": 1985 }, { "epoch": 0.6411881028290084, "grad_norm": 0.4513034795388955, "learning_rate": 1.3927425257490011e-05, "loss": 0.3914, "step": 1986 }, { "epoch": 0.6415109568586302, "grad_norm": 0.6744788997720049, "learning_rate": 1.3926049114191699e-05, "loss": 0.3663, "step": 1987 }, { "epoch": 0.6418338108882522, "grad_norm": 0.45845372682327895, "learning_rate": 1.3924672156721811e-05, "loss": 0.3144, "step": 1988 }, { "epoch": 0.642156664917874, "grad_norm": 0.42536989351469123, "learning_rate": 1.3923294385254809e-05, "loss": 0.3639, "step": 1989 }, { "epoch": 0.6424795189474959, "grad_norm": 0.3779983141021004, "learning_rate": 1.3921915799965253e-05, "loss": 0.3207, "step": 1990 }, { "epoch": 0.6428023729771177, "grad_norm": 0.4714791467475131, "learning_rate": 1.3920536401027808e-05, "loss": 0.3335, "step": 1991 }, { "epoch": 0.6431252270067396, "grad_norm": 0.43878980011626073, "learning_rate": 1.3919156188617246e-05, "loss": 0.3053, "step": 1992 }, { "epoch": 0.6434480810363614, "grad_norm": 0.3917618867295952, "learning_rate": 1.3917775162908431e-05, "loss": 0.3857, "step": 1993 }, { "epoch": 0.6437709350659833, "grad_norm": 0.36835478365132135, "learning_rate": 1.3916393324076342e-05, "loss": 0.3105, "step": 1994 }, { "epoch": 0.6440937890956051, "grad_norm": 0.47247553399478615, "learning_rate": 1.3915010672296053e-05, "loss": 0.3965, "step": 1995 }, { "epoch": 0.6444166431252271, "grad_norm": 0.5112401307863783, "learning_rate": 1.3913627207742749e-05, "loss": 0.3708, "step": 1996 }, { "epoch": 0.6447394971548489, "grad_norm": 0.4620982967819112, "learning_rate": 1.3912242930591708e-05, "loss": 0.39, "step": 1997 }, { "epoch": 0.6450623511844708, "grad_norm": 0.47594390648902996, "learning_rate": 1.3910857841018317e-05, "loss": 0.4152, "step": 1998 }, { "epoch": 0.6453852052140926, "grad_norm": 0.42369943874363497, "learning_rate": 1.3909471939198069e-05, "loss": 0.3481, "step": 1999 }, { "epoch": 0.6457080592437144, "grad_norm": 0.4299939540283288, "learning_rate": 1.390808522530655e-05, "loss": 0.4026, "step": 2000 }, { "epoch": 0.6460309132733363, "grad_norm": 0.4742814074593114, "learning_rate": 1.3906697699519458e-05, "loss": 0.3821, "step": 2001 }, { "epoch": 0.6463537673029581, "grad_norm": 0.5157376700381704, "learning_rate": 1.3905309362012592e-05, "loss": 0.3147, "step": 2002 }, { "epoch": 0.64667662133258, "grad_norm": 0.44849015333765735, "learning_rate": 1.3903920212961846e-05, "loss": 0.3517, "step": 2003 }, { "epoch": 0.6469994753622018, "grad_norm": 0.4014424975331948, "learning_rate": 1.3902530252543232e-05, "loss": 0.3222, "step": 2004 }, { "epoch": 0.6473223293918238, "grad_norm": 0.4608593744435627, "learning_rate": 1.3901139480932849e-05, "loss": 0.3569, "step": 2005 }, { "epoch": 0.6476451834214456, "grad_norm": 0.4620873166940478, "learning_rate": 1.389974789830691e-05, "loss": 0.3005, "step": 2006 }, { "epoch": 0.6479680374510675, "grad_norm": 0.3843453410425647, "learning_rate": 1.3898355504841726e-05, "loss": 0.3224, "step": 2007 }, { "epoch": 0.6482908914806893, "grad_norm": 0.45560359671824585, "learning_rate": 1.3896962300713708e-05, "loss": 0.3488, "step": 2008 }, { "epoch": 0.6486137455103111, "grad_norm": 0.4856977214923467, "learning_rate": 1.3895568286099376e-05, "loss": 0.3154, "step": 2009 }, { "epoch": 0.648936599539933, "grad_norm": 0.4362210865021625, "learning_rate": 1.3894173461175353e-05, "loss": 0.4126, "step": 2010 }, { "epoch": 0.6492594535695548, "grad_norm": 0.3402519524822603, "learning_rate": 1.3892777826118356e-05, "loss": 0.2941, "step": 2011 }, { "epoch": 0.6495823075991767, "grad_norm": 0.4045886937886224, "learning_rate": 1.3891381381105209e-05, "loss": 0.3059, "step": 2012 }, { "epoch": 0.6499051616287986, "grad_norm": 0.36759546547840904, "learning_rate": 1.3889984126312847e-05, "loss": 0.3715, "step": 2013 }, { "epoch": 0.6502280156584205, "grad_norm": 0.3655281098920954, "learning_rate": 1.3888586061918291e-05, "loss": 0.3284, "step": 2014 }, { "epoch": 0.6505508696880423, "grad_norm": 0.44039980252027366, "learning_rate": 1.3887187188098683e-05, "loss": 0.3172, "step": 2015 }, { "epoch": 0.6508737237176642, "grad_norm": 0.39361741990287685, "learning_rate": 1.3885787505031253e-05, "loss": 0.3747, "step": 2016 }, { "epoch": 0.651196577747286, "grad_norm": 0.35358900291837125, "learning_rate": 1.3884387012893337e-05, "loss": 0.3554, "step": 2017 }, { "epoch": 0.6515194317769079, "grad_norm": 0.5388306687634798, "learning_rate": 1.388298571186238e-05, "loss": 0.3601, "step": 2018 }, { "epoch": 0.6518422858065297, "grad_norm": 0.43036774987302, "learning_rate": 1.3881583602115923e-05, "loss": 0.3848, "step": 2019 }, { "epoch": 0.6521651398361515, "grad_norm": 0.38572073287606556, "learning_rate": 1.3880180683831614e-05, "loss": 0.3398, "step": 2020 }, { "epoch": 0.6524879938657734, "grad_norm": 0.3735839335231854, "learning_rate": 1.3878776957187194e-05, "loss": 0.3626, "step": 2021 }, { "epoch": 0.6528108478953953, "grad_norm": 0.3547345542524078, "learning_rate": 1.3877372422360521e-05, "loss": 0.3383, "step": 2022 }, { "epoch": 0.6531337019250172, "grad_norm": 0.4283662269188073, "learning_rate": 1.3875967079529545e-05, "loss": 0.4246, "step": 2023 }, { "epoch": 0.653456555954639, "grad_norm": 0.4267148068861489, "learning_rate": 1.3874560928872317e-05, "loss": 0.3661, "step": 2024 }, { "epoch": 0.6537794099842609, "grad_norm": 0.46045511901253144, "learning_rate": 1.3873153970567e-05, "loss": 0.3677, "step": 2025 }, { "epoch": 0.6541022640138827, "grad_norm": 0.43262145798897517, "learning_rate": 1.3871746204791849e-05, "loss": 0.3371, "step": 2026 }, { "epoch": 0.6544251180435046, "grad_norm": 0.38425135529838367, "learning_rate": 1.387033763172523e-05, "loss": 0.3615, "step": 2027 }, { "epoch": 0.6547479720731264, "grad_norm": 0.5532681679490276, "learning_rate": 1.3868928251545604e-05, "loss": 0.3362, "step": 2028 }, { "epoch": 0.6550708261027482, "grad_norm": 0.46105214507630754, "learning_rate": 1.3867518064431541e-05, "loss": 0.3434, "step": 2029 }, { "epoch": 0.6553936801323702, "grad_norm": 0.41723916219784507, "learning_rate": 1.3866107070561707e-05, "loss": 0.4527, "step": 2030 }, { "epoch": 0.655716534161992, "grad_norm": 0.47180806631120364, "learning_rate": 1.3864695270114872e-05, "loss": 0.4044, "step": 2031 }, { "epoch": 0.6560393881916139, "grad_norm": 0.39284806307623477, "learning_rate": 1.3863282663269914e-05, "loss": 0.3547, "step": 2032 }, { "epoch": 0.6563622422212357, "grad_norm": 0.4660931686460172, "learning_rate": 1.3861869250205805e-05, "loss": 0.3673, "step": 2033 }, { "epoch": 0.6566850962508576, "grad_norm": 0.4780459515316684, "learning_rate": 1.386045503110162e-05, "loss": 0.324, "step": 2034 }, { "epoch": 0.6570079502804794, "grad_norm": 0.38670985489652104, "learning_rate": 1.3859040006136547e-05, "loss": 0.3186, "step": 2035 }, { "epoch": 0.6573308043101013, "grad_norm": 0.5669855063424386, "learning_rate": 1.3857624175489856e-05, "loss": 0.4001, "step": 2036 }, { "epoch": 0.6576536583397231, "grad_norm": 0.4717099710178998, "learning_rate": 1.385620753934094e-05, "loss": 0.3577, "step": 2037 }, { "epoch": 0.657976512369345, "grad_norm": 0.37125140441464016, "learning_rate": 1.3854790097869281e-05, "loss": 0.3189, "step": 2038 }, { "epoch": 0.6582993663989669, "grad_norm": 0.3867091968837398, "learning_rate": 1.3853371851254469e-05, "loss": 0.3609, "step": 2039 }, { "epoch": 0.6586222204285888, "grad_norm": 0.3932725837802478, "learning_rate": 1.3851952799676192e-05, "loss": 0.342, "step": 2040 }, { "epoch": 0.6589450744582106, "grad_norm": 0.43135668883377454, "learning_rate": 1.385053294331424e-05, "loss": 0.3775, "step": 2041 }, { "epoch": 0.6592679284878324, "grad_norm": 0.48781710440163234, "learning_rate": 1.3849112282348513e-05, "loss": 0.351, "step": 2042 }, { "epoch": 0.6595907825174543, "grad_norm": 0.5233447330495021, "learning_rate": 1.3847690816959e-05, "loss": 0.3631, "step": 2043 }, { "epoch": 0.6599136365470761, "grad_norm": 0.416559945080969, "learning_rate": 1.3846268547325805e-05, "loss": 0.3859, "step": 2044 }, { "epoch": 0.660236490576698, "grad_norm": 0.42908935255643665, "learning_rate": 1.3844845473629121e-05, "loss": 0.4009, "step": 2045 }, { "epoch": 0.6605593446063198, "grad_norm": 2.6292944785743226, "learning_rate": 1.3843421596049254e-05, "loss": 0.4086, "step": 2046 }, { "epoch": 0.6608821986359418, "grad_norm": 0.5675524851270454, "learning_rate": 1.3841996914766606e-05, "loss": 0.3849, "step": 2047 }, { "epoch": 0.6612050526655636, "grad_norm": 0.6346307941062032, "learning_rate": 1.3840571429961684e-05, "loss": 0.4054, "step": 2048 }, { "epoch": 0.6615279066951855, "grad_norm": 0.505252969483882, "learning_rate": 1.3839145141815092e-05, "loss": 0.3331, "step": 2049 }, { "epoch": 0.6618507607248073, "grad_norm": 0.4138274138175555, "learning_rate": 1.383771805050754e-05, "loss": 0.4296, "step": 2050 }, { "epoch": 0.6621736147544292, "grad_norm": 0.4471528618765477, "learning_rate": 1.383629015621984e-05, "loss": 0.3633, "step": 2051 }, { "epoch": 0.662496468784051, "grad_norm": 0.4283458375717146, "learning_rate": 1.38348614591329e-05, "loss": 0.3684, "step": 2052 }, { "epoch": 0.6628193228136728, "grad_norm": 0.39332600956950753, "learning_rate": 1.383343195942774e-05, "loss": 0.3272, "step": 2053 }, { "epoch": 0.6631421768432947, "grad_norm": 0.40418118599739034, "learning_rate": 1.383200165728547e-05, "loss": 0.3761, "step": 2054 }, { "epoch": 0.6634650308729165, "grad_norm": 0.48321206762283353, "learning_rate": 1.383057055288731e-05, "loss": 0.3299, "step": 2055 }, { "epoch": 0.6637878849025385, "grad_norm": 0.4208013499865497, "learning_rate": 1.3829138646414578e-05, "loss": 0.3784, "step": 2056 }, { "epoch": 0.6641107389321603, "grad_norm": 0.4903658346207932, "learning_rate": 1.3827705938048698e-05, "loss": 0.2712, "step": 2057 }, { "epoch": 0.6644335929617822, "grad_norm": 0.542050232276336, "learning_rate": 1.3826272427971186e-05, "loss": 0.3368, "step": 2058 }, { "epoch": 0.664756446991404, "grad_norm": 0.5071844281161776, "learning_rate": 1.3824838116363673e-05, "loss": 0.4059, "step": 2059 }, { "epoch": 0.6650793010210259, "grad_norm": 0.5548224976658793, "learning_rate": 1.3823403003407876e-05, "loss": 0.3216, "step": 2060 }, { "epoch": 0.6654021550506477, "grad_norm": 0.45018974019522606, "learning_rate": 1.382196708928563e-05, "loss": 0.4403, "step": 2061 }, { "epoch": 0.6657250090802695, "grad_norm": 0.4217706455926426, "learning_rate": 1.3820530374178859e-05, "loss": 0.366, "step": 2062 }, { "epoch": 0.6660478631098914, "grad_norm": 0.3827220080784297, "learning_rate": 1.3819092858269592e-05, "loss": 0.3672, "step": 2063 }, { "epoch": 0.6663707171395133, "grad_norm": 0.510425937182645, "learning_rate": 1.3817654541739963e-05, "loss": 0.3296, "step": 2064 }, { "epoch": 0.6666935711691352, "grad_norm": 0.46481254588276877, "learning_rate": 1.3816215424772203e-05, "loss": 0.3724, "step": 2065 }, { "epoch": 0.667016425198757, "grad_norm": 0.36697251549134074, "learning_rate": 1.3814775507548646e-05, "loss": 0.3558, "step": 2066 }, { "epoch": 0.6673392792283789, "grad_norm": 0.41059390478975677, "learning_rate": 1.381333479025173e-05, "loss": 0.3967, "step": 2067 }, { "epoch": 0.6676621332580007, "grad_norm": 0.47194719108871686, "learning_rate": 1.381189327306399e-05, "loss": 0.3503, "step": 2068 }, { "epoch": 0.6679849872876226, "grad_norm": 0.36464186374675167, "learning_rate": 1.3810450956168063e-05, "loss": 0.2906, "step": 2069 }, { "epoch": 0.6683078413172444, "grad_norm": 0.47293189689298015, "learning_rate": 1.3809007839746689e-05, "loss": 0.3145, "step": 2070 }, { "epoch": 0.6686306953468663, "grad_norm": 0.45647376645186344, "learning_rate": 1.3807563923982711e-05, "loss": 0.352, "step": 2071 }, { "epoch": 0.6689535493764882, "grad_norm": 0.3702218785625269, "learning_rate": 1.380611920905907e-05, "loss": 0.377, "step": 2072 }, { "epoch": 0.66927640340611, "grad_norm": 0.5273287062137975, "learning_rate": 1.3804673695158808e-05, "loss": 0.3848, "step": 2073 }, { "epoch": 0.6695992574357319, "grad_norm": 0.47943699811077434, "learning_rate": 1.3803227382465072e-05, "loss": 0.3348, "step": 2074 }, { "epoch": 0.6699221114653537, "grad_norm": 0.49402403927437555, "learning_rate": 1.3801780271161107e-05, "loss": 0.315, "step": 2075 }, { "epoch": 0.6702449654949756, "grad_norm": 0.4588481289203886, "learning_rate": 1.380033236143026e-05, "loss": 0.3809, "step": 2076 }, { "epoch": 0.6705678195245974, "grad_norm": 0.3876593760292256, "learning_rate": 1.3798883653455977e-05, "loss": 0.3951, "step": 2077 }, { "epoch": 0.6708906735542193, "grad_norm": 0.39254334300997423, "learning_rate": 1.3797434147421811e-05, "loss": 0.3992, "step": 2078 }, { "epoch": 0.6712135275838411, "grad_norm": 0.4504071164217793, "learning_rate": 1.379598384351141e-05, "loss": 0.4038, "step": 2079 }, { "epoch": 0.671536381613463, "grad_norm": 0.9408251671622202, "learning_rate": 1.3794532741908524e-05, "loss": 0.371, "step": 2080 }, { "epoch": 0.6718592356430849, "grad_norm": 0.6058068830832638, "learning_rate": 1.379308084279701e-05, "loss": 0.3639, "step": 2081 }, { "epoch": 0.6721820896727068, "grad_norm": 0.4691975533580365, "learning_rate": 1.379162814636082e-05, "loss": 0.3134, "step": 2082 }, { "epoch": 0.6725049437023286, "grad_norm": 0.49317024156413547, "learning_rate": 1.3790174652784007e-05, "loss": 0.3399, "step": 2083 }, { "epoch": 0.6728277977319504, "grad_norm": 0.7172552339074321, "learning_rate": 1.3788720362250727e-05, "loss": 0.3366, "step": 2084 }, { "epoch": 0.6731506517615723, "grad_norm": 0.3960872520353485, "learning_rate": 1.3787265274945238e-05, "loss": 0.276, "step": 2085 }, { "epoch": 0.6734735057911941, "grad_norm": 0.4099292963156464, "learning_rate": 1.3785809391051895e-05, "loss": 0.3621, "step": 2086 }, { "epoch": 0.673796359820816, "grad_norm": 0.4049795512353467, "learning_rate": 1.378435271075516e-05, "loss": 0.3543, "step": 2087 }, { "epoch": 0.6741192138504378, "grad_norm": 0.48131459370510227, "learning_rate": 1.378289523423959e-05, "loss": 0.3018, "step": 2088 }, { "epoch": 0.6744420678800598, "grad_norm": 0.3929909822157725, "learning_rate": 1.3781436961689848e-05, "loss": 0.3637, "step": 2089 }, { "epoch": 0.6747649219096816, "grad_norm": 0.42669707170015203, "learning_rate": 1.3779977893290691e-05, "loss": 0.406, "step": 2090 }, { "epoch": 0.6750877759393035, "grad_norm": 0.4934491584846051, "learning_rate": 1.3778518029226983e-05, "loss": 0.3899, "step": 2091 }, { "epoch": 0.6754106299689253, "grad_norm": 0.4429278538662389, "learning_rate": 1.3777057369683687e-05, "loss": 0.345, "step": 2092 }, { "epoch": 0.6757334839985472, "grad_norm": 0.4063178659569768, "learning_rate": 1.3775595914845868e-05, "loss": 0.3875, "step": 2093 }, { "epoch": 0.676056338028169, "grad_norm": 0.3718360509923182, "learning_rate": 1.3774133664898684e-05, "loss": 0.3188, "step": 2094 }, { "epoch": 0.6763791920577908, "grad_norm": 0.5112399471358764, "learning_rate": 1.3772670620027408e-05, "loss": 0.3288, "step": 2095 }, { "epoch": 0.6767020460874127, "grad_norm": 0.42993836715310557, "learning_rate": 1.3771206780417402e-05, "loss": 0.4231, "step": 2096 }, { "epoch": 0.6770249001170345, "grad_norm": 0.545219405528474, "learning_rate": 1.3769742146254134e-05, "loss": 0.316, "step": 2097 }, { "epoch": 0.6773477541466565, "grad_norm": 0.43316399326398247, "learning_rate": 1.3768276717723169e-05, "loss": 0.4016, "step": 2098 }, { "epoch": 0.6776706081762783, "grad_norm": 0.43246370245460636, "learning_rate": 1.3766810495010175e-05, "loss": 0.3736, "step": 2099 }, { "epoch": 0.6779934622059002, "grad_norm": 0.3942568549432553, "learning_rate": 1.3765343478300921e-05, "loss": 0.3301, "step": 2100 }, { "epoch": 0.678316316235522, "grad_norm": 0.43024802744208257, "learning_rate": 1.3763875667781277e-05, "loss": 0.3782, "step": 2101 }, { "epoch": 0.6786391702651439, "grad_norm": 0.5728380665176692, "learning_rate": 1.3762407063637212e-05, "loss": 0.4317, "step": 2102 }, { "epoch": 0.6789620242947657, "grad_norm": 0.4097206007456285, "learning_rate": 1.3760937666054797e-05, "loss": 0.4265, "step": 2103 }, { "epoch": 0.6792848783243876, "grad_norm": 0.49843749537113596, "learning_rate": 1.3759467475220197e-05, "loss": 0.4119, "step": 2104 }, { "epoch": 0.6796077323540094, "grad_norm": 0.3814868851711702, "learning_rate": 1.3757996491319693e-05, "loss": 0.3288, "step": 2105 }, { "epoch": 0.6799305863836314, "grad_norm": 0.4083089600351032, "learning_rate": 1.3756524714539649e-05, "loss": 0.3939, "step": 2106 }, { "epoch": 0.6802534404132532, "grad_norm": 0.46378866260813606, "learning_rate": 1.3755052145066539e-05, "loss": 0.3756, "step": 2107 }, { "epoch": 0.680576294442875, "grad_norm": 0.452697264891202, "learning_rate": 1.3753578783086937e-05, "loss": 0.3197, "step": 2108 }, { "epoch": 0.6808991484724969, "grad_norm": 0.5552801155518019, "learning_rate": 1.3752104628787515e-05, "loss": 0.3767, "step": 2109 }, { "epoch": 0.6812220025021187, "grad_norm": 0.40208963073545045, "learning_rate": 1.3750629682355046e-05, "loss": 0.2967, "step": 2110 }, { "epoch": 0.6815448565317406, "grad_norm": 0.44321119236503087, "learning_rate": 1.3749153943976404e-05, "loss": 0.4137, "step": 2111 }, { "epoch": 0.6818677105613624, "grad_norm": 1.1065939793943607, "learning_rate": 1.3747677413838562e-05, "loss": 0.347, "step": 2112 }, { "epoch": 0.6821905645909843, "grad_norm": 0.44479677369125853, "learning_rate": 1.3746200092128595e-05, "loss": 0.4103, "step": 2113 }, { "epoch": 0.6825134186206061, "grad_norm": 0.5091967847200313, "learning_rate": 1.3744721979033677e-05, "loss": 0.3564, "step": 2114 }, { "epoch": 0.6828362726502281, "grad_norm": 0.3947629254052183, "learning_rate": 1.3743243074741086e-05, "loss": 0.3526, "step": 2115 }, { "epoch": 0.6831591266798499, "grad_norm": 0.4389234370295678, "learning_rate": 1.3741763379438195e-05, "loss": 0.3807, "step": 2116 }, { "epoch": 0.6834819807094717, "grad_norm": 0.5786910548804532, "learning_rate": 1.3740282893312478e-05, "loss": 0.3089, "step": 2117 }, { "epoch": 0.6838048347390936, "grad_norm": 0.4240504679824216, "learning_rate": 1.373880161655151e-05, "loss": 0.3538, "step": 2118 }, { "epoch": 0.6841276887687154, "grad_norm": 0.46580396005254454, "learning_rate": 1.3737319549342972e-05, "loss": 0.3331, "step": 2119 }, { "epoch": 0.6844505427983373, "grad_norm": 0.520009586898797, "learning_rate": 1.3735836691874636e-05, "loss": 0.3452, "step": 2120 }, { "epoch": 0.6847733968279591, "grad_norm": 0.48766729916905116, "learning_rate": 1.3734353044334377e-05, "loss": 0.3583, "step": 2121 }, { "epoch": 0.685096250857581, "grad_norm": 0.4523243175010725, "learning_rate": 1.3732868606910173e-05, "loss": 0.3428, "step": 2122 }, { "epoch": 0.6854191048872029, "grad_norm": 0.37720253490365885, "learning_rate": 1.3731383379790099e-05, "loss": 0.3019, "step": 2123 }, { "epoch": 0.6857419589168248, "grad_norm": 0.4923300540483854, "learning_rate": 1.3729897363162332e-05, "loss": 0.3947, "step": 2124 }, { "epoch": 0.6860648129464466, "grad_norm": 0.48978402269412663, "learning_rate": 1.3728410557215149e-05, "loss": 0.3786, "step": 2125 }, { "epoch": 0.6863876669760685, "grad_norm": 0.45012368724679463, "learning_rate": 1.3726922962136924e-05, "loss": 0.4114, "step": 2126 }, { "epoch": 0.6867105210056903, "grad_norm": 0.5206913802611522, "learning_rate": 1.3725434578116136e-05, "loss": 0.4406, "step": 2127 }, { "epoch": 0.6870333750353121, "grad_norm": 0.5172118445273409, "learning_rate": 1.372394540534136e-05, "loss": 0.3528, "step": 2128 }, { "epoch": 0.687356229064934, "grad_norm": 0.533015146182524, "learning_rate": 1.3722455444001274e-05, "loss": 0.3969, "step": 2129 }, { "epoch": 0.6876790830945558, "grad_norm": 0.44278445003354255, "learning_rate": 1.372096469428465e-05, "loss": 0.4048, "step": 2130 }, { "epoch": 0.6880019371241777, "grad_norm": 0.487506297049939, "learning_rate": 1.3719473156380364e-05, "loss": 0.3765, "step": 2131 }, { "epoch": 0.6883247911537996, "grad_norm": 0.5620577701695018, "learning_rate": 1.3717980830477395e-05, "loss": 0.2695, "step": 2132 }, { "epoch": 0.6886476451834215, "grad_norm": 0.3352666850276332, "learning_rate": 1.3716487716764819e-05, "loss": 0.2946, "step": 2133 }, { "epoch": 0.6889704992130433, "grad_norm": 0.44430954925328564, "learning_rate": 1.3714993815431809e-05, "loss": 0.3921, "step": 2134 }, { "epoch": 0.6892933532426652, "grad_norm": 0.5160350667510828, "learning_rate": 1.3713499126667642e-05, "loss": 0.3756, "step": 2135 }, { "epoch": 0.689616207272287, "grad_norm": 0.34985800059136574, "learning_rate": 1.3712003650661692e-05, "loss": 0.3314, "step": 2136 }, { "epoch": 0.6899390613019089, "grad_norm": 0.45709808242118855, "learning_rate": 1.3710507387603431e-05, "loss": 0.3516, "step": 2137 }, { "epoch": 0.6902619153315307, "grad_norm": 0.3767899288159065, "learning_rate": 1.370901033768244e-05, "loss": 0.3763, "step": 2138 }, { "epoch": 0.6905847693611525, "grad_norm": 0.39609259884327935, "learning_rate": 1.3707512501088388e-05, "loss": 0.3364, "step": 2139 }, { "epoch": 0.6909076233907745, "grad_norm": 0.56189914941916, "learning_rate": 1.370601387801105e-05, "loss": 0.4519, "step": 2140 }, { "epoch": 0.6912304774203963, "grad_norm": 0.43721634534580905, "learning_rate": 1.3704514468640301e-05, "loss": 0.3869, "step": 2141 }, { "epoch": 0.6915533314500182, "grad_norm": 0.36061461264608363, "learning_rate": 1.3703014273166112e-05, "loss": 0.4079, "step": 2142 }, { "epoch": 0.69187618547964, "grad_norm": 0.4649417065398642, "learning_rate": 1.3701513291778555e-05, "loss": 0.3278, "step": 2143 }, { "epoch": 0.6921990395092619, "grad_norm": 0.3881308997551459, "learning_rate": 1.3700011524667808e-05, "loss": 0.3148, "step": 2144 }, { "epoch": 0.6925218935388837, "grad_norm": 0.4464087139370531, "learning_rate": 1.3698508972024136e-05, "loss": 0.3535, "step": 2145 }, { "epoch": 0.6928447475685056, "grad_norm": 0.41071762016144936, "learning_rate": 1.3697005634037913e-05, "loss": 0.4028, "step": 2146 }, { "epoch": 0.6931676015981274, "grad_norm": 0.4303801252654509, "learning_rate": 1.3695501510899612e-05, "loss": 0.3724, "step": 2147 }, { "epoch": 0.6934904556277494, "grad_norm": 0.4104296539029859, "learning_rate": 1.36939966027998e-05, "loss": 0.3776, "step": 2148 }, { "epoch": 0.6938133096573712, "grad_norm": 0.5147608506913913, "learning_rate": 1.3692490909929147e-05, "loss": 0.4173, "step": 2149 }, { "epoch": 0.694136163686993, "grad_norm": 0.4234451239785683, "learning_rate": 1.3690984432478422e-05, "loss": 0.3754, "step": 2150 }, { "epoch": 0.6944590177166149, "grad_norm": 0.40908189588394567, "learning_rate": 1.36894771706385e-05, "loss": 0.3424, "step": 2151 }, { "epoch": 0.6947818717462367, "grad_norm": 0.3867088224399239, "learning_rate": 1.368796912460034e-05, "loss": 0.3046, "step": 2152 }, { "epoch": 0.6951047257758586, "grad_norm": 0.44897769547354416, "learning_rate": 1.3686460294555015e-05, "loss": 0.3788, "step": 2153 }, { "epoch": 0.6954275798054804, "grad_norm": 0.4987321693200617, "learning_rate": 1.368495068069369e-05, "loss": 0.3525, "step": 2154 }, { "epoch": 0.6957504338351023, "grad_norm": 0.43234805521335884, "learning_rate": 1.3683440283207627e-05, "loss": 0.3037, "step": 2155 }, { "epoch": 0.6960732878647241, "grad_norm": 0.377650147379634, "learning_rate": 1.3681929102288201e-05, "loss": 0.3529, "step": 2156 }, { "epoch": 0.6963961418943461, "grad_norm": 0.5526917182987664, "learning_rate": 1.3680417138126866e-05, "loss": 0.3914, "step": 2157 }, { "epoch": 0.6967189959239679, "grad_norm": 0.5861658382308617, "learning_rate": 1.3678904390915193e-05, "loss": 0.3952, "step": 2158 }, { "epoch": 0.6970418499535898, "grad_norm": 0.44499502959585724, "learning_rate": 1.3677390860844843e-05, "loss": 0.4021, "step": 2159 }, { "epoch": 0.6973647039832116, "grad_norm": 0.40576997808535936, "learning_rate": 1.3675876548107578e-05, "loss": 0.3756, "step": 2160 }, { "epoch": 0.6976875580128334, "grad_norm": 0.38409926370350217, "learning_rate": 1.3674361452895258e-05, "loss": 0.3567, "step": 2161 }, { "epoch": 0.6980104120424553, "grad_norm": 0.5012215571741454, "learning_rate": 1.3672845575399842e-05, "loss": 0.3721, "step": 2162 }, { "epoch": 0.6983332660720771, "grad_norm": 0.7462973509172899, "learning_rate": 1.3671328915813395e-05, "loss": 0.3189, "step": 2163 }, { "epoch": 0.698656120101699, "grad_norm": 0.3632349222078559, "learning_rate": 1.3669811474328074e-05, "loss": 0.3565, "step": 2164 }, { "epoch": 0.6989789741313209, "grad_norm": 0.4600623212469123, "learning_rate": 1.3668293251136135e-05, "loss": 0.3303, "step": 2165 }, { "epoch": 0.6993018281609428, "grad_norm": 0.38873451880143134, "learning_rate": 1.3666774246429933e-05, "loss": 0.3414, "step": 2166 }, { "epoch": 0.6996246821905646, "grad_norm": 0.45055725687409154, "learning_rate": 1.3665254460401926e-05, "loss": 0.359, "step": 2167 }, { "epoch": 0.6999475362201865, "grad_norm": 0.3758004225558438, "learning_rate": 1.3663733893244672e-05, "loss": 0.3317, "step": 2168 }, { "epoch": 0.7002703902498083, "grad_norm": 0.3970318548883702, "learning_rate": 1.366221254515082e-05, "loss": 0.4156, "step": 2169 }, { "epoch": 0.7005932442794301, "grad_norm": 0.42607120121221015, "learning_rate": 1.3660690416313123e-05, "loss": 0.3611, "step": 2170 }, { "epoch": 0.700916098309052, "grad_norm": 0.442447536956529, "learning_rate": 1.3659167506924436e-05, "loss": 0.3614, "step": 2171 }, { "epoch": 0.7012389523386738, "grad_norm": 0.46044220905707767, "learning_rate": 1.3657643817177708e-05, "loss": 0.3297, "step": 2172 }, { "epoch": 0.7015618063682957, "grad_norm": 0.37707249463354586, "learning_rate": 1.3656119347265985e-05, "loss": 0.3056, "step": 2173 }, { "epoch": 0.7018846603979176, "grad_norm": 0.5724619050107114, "learning_rate": 1.365459409738242e-05, "loss": 0.3781, "step": 2174 }, { "epoch": 0.7022075144275395, "grad_norm": 0.5589655440308433, "learning_rate": 1.365306806772026e-05, "loss": 0.4175, "step": 2175 }, { "epoch": 0.7025303684571613, "grad_norm": 0.3661842974620014, "learning_rate": 1.3651541258472848e-05, "loss": 0.3689, "step": 2176 }, { "epoch": 0.7028532224867832, "grad_norm": 0.5110120119835265, "learning_rate": 1.3650013669833628e-05, "loss": 0.3791, "step": 2177 }, { "epoch": 0.703176076516405, "grad_norm": 0.40746203554652005, "learning_rate": 1.3648485301996147e-05, "loss": 0.4529, "step": 2178 }, { "epoch": 0.7034989305460269, "grad_norm": 0.40346860158164705, "learning_rate": 1.3646956155154043e-05, "loss": 0.3873, "step": 2179 }, { "epoch": 0.7038217845756487, "grad_norm": 0.44144780450833737, "learning_rate": 1.3645426229501062e-05, "loss": 0.3432, "step": 2180 }, { "epoch": 0.7041446386052705, "grad_norm": 0.46415169263478484, "learning_rate": 1.3643895525231039e-05, "loss": 0.4195, "step": 2181 }, { "epoch": 0.7044674926348925, "grad_norm": 0.46393456637574326, "learning_rate": 1.3642364042537913e-05, "loss": 0.4353, "step": 2182 }, { "epoch": 0.7047903466645143, "grad_norm": 0.4384378810507866, "learning_rate": 1.3640831781615722e-05, "loss": 0.3977, "step": 2183 }, { "epoch": 0.7051132006941362, "grad_norm": 0.4460840627742546, "learning_rate": 1.3639298742658601e-05, "loss": 0.3497, "step": 2184 }, { "epoch": 0.705436054723758, "grad_norm": 0.44407930716110783, "learning_rate": 1.3637764925860783e-05, "loss": 0.3484, "step": 2185 }, { "epoch": 0.7057589087533799, "grad_norm": 0.360852869776971, "learning_rate": 1.3636230331416601e-05, "loss": 0.3941, "step": 2186 }, { "epoch": 0.7060817627830017, "grad_norm": 0.4514286995769782, "learning_rate": 1.3634694959520487e-05, "loss": 0.3583, "step": 2187 }, { "epoch": 0.7064046168126236, "grad_norm": 0.43945007813650977, "learning_rate": 1.3633158810366968e-05, "loss": 0.3936, "step": 2188 }, { "epoch": 0.7067274708422454, "grad_norm": 0.3787302670393506, "learning_rate": 1.3631621884150676e-05, "loss": 0.4053, "step": 2189 }, { "epoch": 0.7070503248718673, "grad_norm": 0.36158767983764234, "learning_rate": 1.3630084181066333e-05, "loss": 0.3568, "step": 2190 }, { "epoch": 0.7073731789014892, "grad_norm": 0.4030404734956365, "learning_rate": 1.3628545701308767e-05, "loss": 0.3882, "step": 2191 }, { "epoch": 0.707696032931111, "grad_norm": 0.48251374663340446, "learning_rate": 1.36270064450729e-05, "loss": 0.3603, "step": 2192 }, { "epoch": 0.7080188869607329, "grad_norm": 0.332057316361229, "learning_rate": 1.3625466412553754e-05, "loss": 0.3146, "step": 2193 }, { "epoch": 0.7083417409903547, "grad_norm": 0.4244401830367042, "learning_rate": 1.3623925603946448e-05, "loss": 0.405, "step": 2194 }, { "epoch": 0.7086645950199766, "grad_norm": 0.3439782182697493, "learning_rate": 1.36223840194462e-05, "loss": 0.3092, "step": 2195 }, { "epoch": 0.7089874490495984, "grad_norm": 0.3765267929470224, "learning_rate": 1.362084165924833e-05, "loss": 0.3357, "step": 2196 }, { "epoch": 0.7093103030792203, "grad_norm": 0.5879262216457272, "learning_rate": 1.361929852354825e-05, "loss": 0.4555, "step": 2197 }, { "epoch": 0.7096331571088421, "grad_norm": 0.3574266883038289, "learning_rate": 1.3617754612541474e-05, "loss": 0.3581, "step": 2198 }, { "epoch": 0.7099560111384641, "grad_norm": 0.5281950466889196, "learning_rate": 1.3616209926423614e-05, "loss": 0.3721, "step": 2199 }, { "epoch": 0.7102788651680859, "grad_norm": 0.41084005228808407, "learning_rate": 1.3614664465390376e-05, "loss": 0.3564, "step": 2200 }, { "epoch": 0.7106017191977078, "grad_norm": 0.44001209599264807, "learning_rate": 1.3613118229637574e-05, "loss": 0.3451, "step": 2201 }, { "epoch": 0.7109245732273296, "grad_norm": 0.4173250778666645, "learning_rate": 1.3611571219361109e-05, "loss": 0.4089, "step": 2202 }, { "epoch": 0.7112474272569514, "grad_norm": 0.3684151884209428, "learning_rate": 1.3610023434756985e-05, "loss": 0.3313, "step": 2203 }, { "epoch": 0.7115702812865733, "grad_norm": 0.7685738896747606, "learning_rate": 1.3608474876021307e-05, "loss": 0.3487, "step": 2204 }, { "epoch": 0.7118931353161951, "grad_norm": 0.3786327589374576, "learning_rate": 1.3606925543350275e-05, "loss": 0.29, "step": 2205 }, { "epoch": 0.712215989345817, "grad_norm": 0.3497223317499396, "learning_rate": 1.3605375436940188e-05, "loss": 0.3008, "step": 2206 }, { "epoch": 0.7125388433754389, "grad_norm": 0.34217983433529403, "learning_rate": 1.3603824556987437e-05, "loss": 0.3444, "step": 2207 }, { "epoch": 0.7128616974050608, "grad_norm": 0.42610469694636494, "learning_rate": 1.3602272903688522e-05, "loss": 0.33, "step": 2208 }, { "epoch": 0.7131845514346826, "grad_norm": 0.42130935303405154, "learning_rate": 1.3600720477240033e-05, "loss": 0.3193, "step": 2209 }, { "epoch": 0.7135074054643045, "grad_norm": 0.432005842752674, "learning_rate": 1.3599167277838662e-05, "loss": 0.3575, "step": 2210 }, { "epoch": 0.7138302594939263, "grad_norm": 0.4841105201090997, "learning_rate": 1.3597613305681194e-05, "loss": 0.3613, "step": 2211 }, { "epoch": 0.7141531135235482, "grad_norm": 1.1688521731890669, "learning_rate": 1.3596058560964519e-05, "loss": 0.3377, "step": 2212 }, { "epoch": 0.71447596755317, "grad_norm": 0.3765939088818186, "learning_rate": 1.3594503043885617e-05, "loss": 0.4029, "step": 2213 }, { "epoch": 0.7147988215827918, "grad_norm": 0.4616195878563243, "learning_rate": 1.3592946754641572e-05, "loss": 0.3305, "step": 2214 }, { "epoch": 0.7151216756124137, "grad_norm": 0.438946740171313, "learning_rate": 1.3591389693429567e-05, "loss": 0.3667, "step": 2215 }, { "epoch": 0.7154445296420356, "grad_norm": 0.39969512652288347, "learning_rate": 1.3589831860446874e-05, "loss": 0.3268, "step": 2216 }, { "epoch": 0.7157673836716575, "grad_norm": 0.5468796237766911, "learning_rate": 1.3588273255890869e-05, "loss": 0.3984, "step": 2217 }, { "epoch": 0.7160902377012793, "grad_norm": 0.4838677481288768, "learning_rate": 1.358671387995903e-05, "loss": 0.3712, "step": 2218 }, { "epoch": 0.7164130917309012, "grad_norm": 0.4877647667097384, "learning_rate": 1.3585153732848921e-05, "loss": 0.3357, "step": 2219 }, { "epoch": 0.716735945760523, "grad_norm": 0.411744434483841, "learning_rate": 1.3583592814758215e-05, "loss": 0.3313, "step": 2220 }, { "epoch": 0.7170587997901449, "grad_norm": 0.4534479660617184, "learning_rate": 1.3582031125884676e-05, "loss": 0.394, "step": 2221 }, { "epoch": 0.7173816538197667, "grad_norm": 0.4518615105155592, "learning_rate": 1.358046866642617e-05, "loss": 0.2732, "step": 2222 }, { "epoch": 0.7177045078493886, "grad_norm": 0.3212581696780315, "learning_rate": 1.357890543658066e-05, "loss": 0.3158, "step": 2223 }, { "epoch": 0.7180273618790105, "grad_norm": 0.40861448709187115, "learning_rate": 1.3577341436546197e-05, "loss": 0.3125, "step": 2224 }, { "epoch": 0.7183502159086324, "grad_norm": 0.541996729394355, "learning_rate": 1.3575776666520945e-05, "loss": 0.3706, "step": 2225 }, { "epoch": 0.7186730699382542, "grad_norm": 0.38649616028478656, "learning_rate": 1.3574211126703157e-05, "loss": 0.3805, "step": 2226 }, { "epoch": 0.718995923967876, "grad_norm": 0.5637546204517752, "learning_rate": 1.3572644817291183e-05, "loss": 0.4223, "step": 2227 }, { "epoch": 0.7193187779974979, "grad_norm": 0.5503184601064447, "learning_rate": 1.3571077738483474e-05, "loss": 0.3778, "step": 2228 }, { "epoch": 0.7196416320271197, "grad_norm": 0.40341023635769846, "learning_rate": 1.3569509890478576e-05, "loss": 0.4133, "step": 2229 }, { "epoch": 0.7199644860567416, "grad_norm": 0.38424907966824495, "learning_rate": 1.3567941273475132e-05, "loss": 0.3187, "step": 2230 }, { "epoch": 0.7202873400863634, "grad_norm": 0.4690060035966923, "learning_rate": 1.3566371887671886e-05, "loss": 0.335, "step": 2231 }, { "epoch": 0.7206101941159853, "grad_norm": 0.3628287765194038, "learning_rate": 1.3564801733267675e-05, "loss": 0.3299, "step": 2232 }, { "epoch": 0.7209330481456072, "grad_norm": 0.44150782392897964, "learning_rate": 1.3563230810461435e-05, "loss": 0.4055, "step": 2233 }, { "epoch": 0.7212559021752291, "grad_norm": 0.369526001483847, "learning_rate": 1.3561659119452204e-05, "loss": 0.3421, "step": 2234 }, { "epoch": 0.7215787562048509, "grad_norm": 0.4277464459985018, "learning_rate": 1.3560086660439106e-05, "loss": 0.3513, "step": 2235 }, { "epoch": 0.7219016102344727, "grad_norm": 0.46026977617943543, "learning_rate": 1.3558513433621378e-05, "loss": 0.2847, "step": 2236 }, { "epoch": 0.7222244642640946, "grad_norm": 0.36604608036490355, "learning_rate": 1.3556939439198337e-05, "loss": 0.3608, "step": 2237 }, { "epoch": 0.7225473182937164, "grad_norm": 0.3696852540622641, "learning_rate": 1.3555364677369413e-05, "loss": 0.3422, "step": 2238 }, { "epoch": 0.7228701723233383, "grad_norm": 0.42848578538771764, "learning_rate": 1.3553789148334121e-05, "loss": 0.3081, "step": 2239 }, { "epoch": 0.7231930263529601, "grad_norm": 0.5807889535217203, "learning_rate": 1.3552212852292082e-05, "loss": 0.3645, "step": 2240 }, { "epoch": 0.7235158803825821, "grad_norm": 0.41588851015471073, "learning_rate": 1.3550635789443007e-05, "loss": 0.4616, "step": 2241 }, { "epoch": 0.7238387344122039, "grad_norm": 0.5079053187461854, "learning_rate": 1.354905795998671e-05, "loss": 0.3905, "step": 2242 }, { "epoch": 0.7241615884418258, "grad_norm": 0.5358710332631795, "learning_rate": 1.35474793641231e-05, "loss": 0.3553, "step": 2243 }, { "epoch": 0.7244844424714476, "grad_norm": 0.4683220504884421, "learning_rate": 1.3545900002052183e-05, "loss": 0.471, "step": 2244 }, { "epoch": 0.7248072965010695, "grad_norm": 0.4299769254909479, "learning_rate": 1.3544319873974062e-05, "loss": 0.3683, "step": 2245 }, { "epoch": 0.7251301505306913, "grad_norm": 0.44278795990883235, "learning_rate": 1.3542738980088935e-05, "loss": 0.3276, "step": 2246 }, { "epoch": 0.7254530045603131, "grad_norm": 0.3389512528918113, "learning_rate": 1.3541157320597101e-05, "loss": 0.3163, "step": 2247 }, { "epoch": 0.725775858589935, "grad_norm": 0.382226467368658, "learning_rate": 1.3539574895698955e-05, "loss": 0.3174, "step": 2248 }, { "epoch": 0.7260987126195568, "grad_norm": 0.44988213250298215, "learning_rate": 1.3537991705594986e-05, "loss": 0.3704, "step": 2249 }, { "epoch": 0.7264215666491788, "grad_norm": 0.3794136860046237, "learning_rate": 1.3536407750485781e-05, "loss": 0.3479, "step": 2250 }, { "epoch": 0.7267444206788006, "grad_norm": 0.4207178929382511, "learning_rate": 1.3534823030572029e-05, "loss": 0.3954, "step": 2251 }, { "epoch": 0.7270672747084225, "grad_norm": 0.43858551001954205, "learning_rate": 1.3533237546054506e-05, "loss": 0.372, "step": 2252 }, { "epoch": 0.7273901287380443, "grad_norm": 0.41622531752773595, "learning_rate": 1.3531651297134098e-05, "loss": 0.3687, "step": 2253 }, { "epoch": 0.7277129827676662, "grad_norm": 0.5180690052101263, "learning_rate": 1.3530064284011776e-05, "loss": 0.3219, "step": 2254 }, { "epoch": 0.728035836797288, "grad_norm": 0.44662522954068967, "learning_rate": 1.352847650688861e-05, "loss": 0.3504, "step": 2255 }, { "epoch": 0.7283586908269098, "grad_norm": 0.5014845834590322, "learning_rate": 1.3526887965965775e-05, "loss": 0.3244, "step": 2256 }, { "epoch": 0.7286815448565317, "grad_norm": 0.40089113693133943, "learning_rate": 1.3525298661444536e-05, "loss": 0.3249, "step": 2257 }, { "epoch": 0.7290043988861536, "grad_norm": 0.3789583452376241, "learning_rate": 1.3523708593526251e-05, "loss": 0.3633, "step": 2258 }, { "epoch": 0.7293272529157755, "grad_norm": 0.42320705786928786, "learning_rate": 1.3522117762412383e-05, "loss": 0.3135, "step": 2259 }, { "epoch": 0.7296501069453973, "grad_norm": 0.47426316508655997, "learning_rate": 1.3520526168304488e-05, "loss": 0.3794, "step": 2260 }, { "epoch": 0.7299729609750192, "grad_norm": 0.4703473902576996, "learning_rate": 1.3518933811404218e-05, "loss": 0.3697, "step": 2261 }, { "epoch": 0.730295815004641, "grad_norm": 0.4966876271396526, "learning_rate": 1.3517340691913321e-05, "loss": 0.3283, "step": 2262 }, { "epoch": 0.7306186690342629, "grad_norm": 0.4115680397014163, "learning_rate": 1.3515746810033648e-05, "loss": 0.3348, "step": 2263 }, { "epoch": 0.7309415230638847, "grad_norm": 0.4678957476784171, "learning_rate": 1.3514152165967138e-05, "loss": 0.3642, "step": 2264 }, { "epoch": 0.7312643770935066, "grad_norm": 0.43416774644860995, "learning_rate": 1.3512556759915829e-05, "loss": 0.4167, "step": 2265 }, { "epoch": 0.7315872311231284, "grad_norm": 0.3519442235141646, "learning_rate": 1.3510960592081858e-05, "loss": 0.3222, "step": 2266 }, { "epoch": 0.7319100851527504, "grad_norm": 0.4258465022459623, "learning_rate": 1.3509363662667458e-05, "loss": 0.429, "step": 2267 }, { "epoch": 0.7322329391823722, "grad_norm": 0.3504962061836021, "learning_rate": 1.3507765971874955e-05, "loss": 0.3404, "step": 2268 }, { "epoch": 0.732555793211994, "grad_norm": 0.4585860678425371, "learning_rate": 1.3506167519906781e-05, "loss": 0.2934, "step": 2269 }, { "epoch": 0.7328786472416159, "grad_norm": 0.5420730961169101, "learning_rate": 1.3504568306965448e-05, "loss": 0.3812, "step": 2270 }, { "epoch": 0.7332015012712377, "grad_norm": 0.3911263137409879, "learning_rate": 1.3502968333253583e-05, "loss": 0.3652, "step": 2271 }, { "epoch": 0.7335243553008596, "grad_norm": 0.4532433546217905, "learning_rate": 1.3501367598973898e-05, "loss": 0.3422, "step": 2272 }, { "epoch": 0.7338472093304814, "grad_norm": 0.5387875155939231, "learning_rate": 1.3499766104329202e-05, "loss": 0.4118, "step": 2273 }, { "epoch": 0.7341700633601033, "grad_norm": 0.42113310812860294, "learning_rate": 1.3498163849522399e-05, "loss": 0.3452, "step": 2274 }, { "epoch": 0.7344929173897252, "grad_norm": 0.4671226151139029, "learning_rate": 1.34965608347565e-05, "loss": 0.3287, "step": 2275 }, { "epoch": 0.7348157714193471, "grad_norm": 0.5192640057088386, "learning_rate": 1.34949570602346e-05, "loss": 0.357, "step": 2276 }, { "epoch": 0.7351386254489689, "grad_norm": 0.4275266187000478, "learning_rate": 1.3493352526159896e-05, "loss": 0.3717, "step": 2277 }, { "epoch": 0.7354614794785908, "grad_norm": 0.42411355960967856, "learning_rate": 1.349174723273568e-05, "loss": 0.3949, "step": 2278 }, { "epoch": 0.7357843335082126, "grad_norm": 0.4371296174104299, "learning_rate": 1.349014118016534e-05, "loss": 0.3678, "step": 2279 }, { "epoch": 0.7361071875378344, "grad_norm": 0.40928463539987553, "learning_rate": 1.3488534368652362e-05, "loss": 0.3639, "step": 2280 }, { "epoch": 0.7364300415674563, "grad_norm": 0.4599771319429729, "learning_rate": 1.3486926798400326e-05, "loss": 0.3869, "step": 2281 }, { "epoch": 0.7367528955970781, "grad_norm": 0.4358505581672679, "learning_rate": 1.348531846961291e-05, "loss": 0.424, "step": 2282 }, { "epoch": 0.7370757496267001, "grad_norm": 0.4681022375068936, "learning_rate": 1.3483709382493886e-05, "loss": 0.3977, "step": 2283 }, { "epoch": 0.7373986036563219, "grad_norm": 0.43251689519606856, "learning_rate": 1.3482099537247123e-05, "loss": 0.3334, "step": 2284 }, { "epoch": 0.7377214576859438, "grad_norm": 0.4604534014930727, "learning_rate": 1.3480488934076587e-05, "loss": 0.3618, "step": 2285 }, { "epoch": 0.7380443117155656, "grad_norm": 0.3858941770224727, "learning_rate": 1.3478877573186338e-05, "loss": 0.3994, "step": 2286 }, { "epoch": 0.7383671657451875, "grad_norm": 0.39224387372401465, "learning_rate": 1.3477265454780532e-05, "loss": 0.3705, "step": 2287 }, { "epoch": 0.7386900197748093, "grad_norm": 0.3667684512854825, "learning_rate": 1.3475652579063428e-05, "loss": 0.3609, "step": 2288 }, { "epoch": 0.7390128738044311, "grad_norm": 0.4509186433910399, "learning_rate": 1.347403894623937e-05, "loss": 0.2661, "step": 2289 }, { "epoch": 0.739335727834053, "grad_norm": 0.4733562884187101, "learning_rate": 1.3472424556512802e-05, "loss": 0.3816, "step": 2290 }, { "epoch": 0.7396585818636748, "grad_norm": 0.44791798209990497, "learning_rate": 1.3470809410088268e-05, "loss": 0.3359, "step": 2291 }, { "epoch": 0.7399814358932968, "grad_norm": 0.4161717779234738, "learning_rate": 1.3469193507170405e-05, "loss": 0.3584, "step": 2292 }, { "epoch": 0.7403042899229186, "grad_norm": 0.403147710650227, "learning_rate": 1.3467576847963946e-05, "loss": 0.3541, "step": 2293 }, { "epoch": 0.7406271439525405, "grad_norm": 0.37865920180332596, "learning_rate": 1.3465959432673715e-05, "loss": 0.3526, "step": 2294 }, { "epoch": 0.7409499979821623, "grad_norm": 0.5248701445884635, "learning_rate": 1.346434126150464e-05, "loss": 0.3428, "step": 2295 }, { "epoch": 0.7412728520117842, "grad_norm": 0.48866219187433296, "learning_rate": 1.3462722334661743e-05, "loss": 0.3912, "step": 2296 }, { "epoch": 0.741595706041406, "grad_norm": 0.5117276787967114, "learning_rate": 1.3461102652350137e-05, "loss": 0.3315, "step": 2297 }, { "epoch": 0.7419185600710279, "grad_norm": 0.3682270788079679, "learning_rate": 1.345948221477503e-05, "loss": 0.3188, "step": 2298 }, { "epoch": 0.7422414141006497, "grad_norm": 0.37297688377153937, "learning_rate": 1.3457861022141738e-05, "loss": 0.3511, "step": 2299 }, { "epoch": 0.7425642681302717, "grad_norm": 0.460084870838882, "learning_rate": 1.3456239074655656e-05, "loss": 0.3279, "step": 2300 }, { "epoch": 0.7428871221598935, "grad_norm": 0.49429795290509976, "learning_rate": 1.3454616372522287e-05, "loss": 0.3084, "step": 2301 }, { "epoch": 0.7432099761895153, "grad_norm": 0.40951944920383754, "learning_rate": 1.3452992915947224e-05, "loss": 0.408, "step": 2302 }, { "epoch": 0.7435328302191372, "grad_norm": 0.473035719234996, "learning_rate": 1.3451368705136155e-05, "loss": 0.3587, "step": 2303 }, { "epoch": 0.743855684248759, "grad_norm": 0.37293013183470314, "learning_rate": 1.3449743740294867e-05, "loss": 0.3568, "step": 2304 }, { "epoch": 0.7441785382783809, "grad_norm": 0.5325834269674525, "learning_rate": 1.344811802162924e-05, "loss": 0.3766, "step": 2305 }, { "epoch": 0.7445013923080027, "grad_norm": 0.40327022141335267, "learning_rate": 1.344649154934525e-05, "loss": 0.3679, "step": 2306 }, { "epoch": 0.7448242463376246, "grad_norm": 0.4400709045689687, "learning_rate": 1.344486432364897e-05, "loss": 0.4033, "step": 2307 }, { "epoch": 0.7451471003672464, "grad_norm": 0.3719962296384474, "learning_rate": 1.3443236344746569e-05, "loss": 0.3567, "step": 2308 }, { "epoch": 0.7454699543968684, "grad_norm": 0.5597516160514153, "learning_rate": 1.3441607612844305e-05, "loss": 0.3761, "step": 2309 }, { "epoch": 0.7457928084264902, "grad_norm": 0.40431966142644665, "learning_rate": 1.343997812814854e-05, "loss": 0.4218, "step": 2310 }, { "epoch": 0.746115662456112, "grad_norm": 0.379980891731131, "learning_rate": 1.3438347890865725e-05, "loss": 0.3575, "step": 2311 }, { "epoch": 0.7464385164857339, "grad_norm": 0.48028703108957105, "learning_rate": 1.343671690120241e-05, "loss": 0.311, "step": 2312 }, { "epoch": 0.7467613705153557, "grad_norm": 0.4452006066413407, "learning_rate": 1.343508515936524e-05, "loss": 0.434, "step": 2313 }, { "epoch": 0.7470842245449776, "grad_norm": 0.4254032486925149, "learning_rate": 1.3433452665560951e-05, "loss": 0.3162, "step": 2314 }, { "epoch": 0.7474070785745994, "grad_norm": 0.395595200792581, "learning_rate": 1.3431819419996384e-05, "loss": 0.3761, "step": 2315 }, { "epoch": 0.7477299326042213, "grad_norm": 0.4369086854905039, "learning_rate": 1.3430185422878461e-05, "loss": 0.3857, "step": 2316 }, { "epoch": 0.7480527866338432, "grad_norm": 0.37470126167838874, "learning_rate": 1.3428550674414211e-05, "loss": 0.3701, "step": 2317 }, { "epoch": 0.7483756406634651, "grad_norm": 0.44195378424309134, "learning_rate": 1.3426915174810757e-05, "loss": 0.3705, "step": 2318 }, { "epoch": 0.7486984946930869, "grad_norm": 0.3933334181034501, "learning_rate": 1.3425278924275313e-05, "loss": 0.3363, "step": 2319 }, { "epoch": 0.7490213487227088, "grad_norm": 0.377689921040645, "learning_rate": 1.3423641923015188e-05, "loss": 0.3453, "step": 2320 }, { "epoch": 0.7493442027523306, "grad_norm": 0.5073355503367185, "learning_rate": 1.3422004171237789e-05, "loss": 0.377, "step": 2321 }, { "epoch": 0.7496670567819524, "grad_norm": 0.6790380422069837, "learning_rate": 1.3420365669150614e-05, "loss": 0.3489, "step": 2322 }, { "epoch": 0.7499899108115743, "grad_norm": 0.45288119864689913, "learning_rate": 1.3418726416961264e-05, "loss": 0.3235, "step": 2323 }, { "epoch": 0.7503127648411961, "grad_norm": 0.43248732090035213, "learning_rate": 1.3417086414877428e-05, "loss": 0.2986, "step": 2324 }, { "epoch": 0.750635618870818, "grad_norm": 0.36512665358511337, "learning_rate": 1.341544566310689e-05, "loss": 0.3636, "step": 2325 }, { "epoch": 0.7509584729004399, "grad_norm": 0.3810403309812954, "learning_rate": 1.3413804161857533e-05, "loss": 0.3525, "step": 2326 }, { "epoch": 0.7512813269300618, "grad_norm": 0.43120246150879443, "learning_rate": 1.3412161911337332e-05, "loss": 0.3703, "step": 2327 }, { "epoch": 0.7516041809596836, "grad_norm": 0.42917812640845393, "learning_rate": 1.341051891175436e-05, "loss": 0.3176, "step": 2328 }, { "epoch": 0.7519270349893055, "grad_norm": 0.522141893186491, "learning_rate": 1.340887516331678e-05, "loss": 0.3239, "step": 2329 }, { "epoch": 0.7522498890189273, "grad_norm": 0.42604832104777335, "learning_rate": 1.3407230666232853e-05, "loss": 0.3888, "step": 2330 }, { "epoch": 0.7525727430485492, "grad_norm": 0.44761497997550914, "learning_rate": 1.3405585420710938e-05, "loss": 0.3577, "step": 2331 }, { "epoch": 0.752895597078171, "grad_norm": 0.473093720521497, "learning_rate": 1.340393942695948e-05, "loss": 0.3275, "step": 2332 }, { "epoch": 0.7532184511077928, "grad_norm": 0.40241246997391356, "learning_rate": 1.3402292685187028e-05, "loss": 0.3686, "step": 2333 }, { "epoch": 0.7535413051374148, "grad_norm": 0.46245919877278263, "learning_rate": 1.3400645195602222e-05, "loss": 0.3519, "step": 2334 }, { "epoch": 0.7538641591670366, "grad_norm": 0.4615558131642691, "learning_rate": 1.3398996958413793e-05, "loss": 0.4004, "step": 2335 }, { "epoch": 0.7541870131966585, "grad_norm": 0.3972811698978138, "learning_rate": 1.3397347973830574e-05, "loss": 0.3742, "step": 2336 }, { "epoch": 0.7545098672262803, "grad_norm": 0.5105204242100457, "learning_rate": 1.3395698242061487e-05, "loss": 0.355, "step": 2337 }, { "epoch": 0.7548327212559022, "grad_norm": 0.4472254226932797, "learning_rate": 1.3394047763315553e-05, "loss": 0.4001, "step": 2338 }, { "epoch": 0.755155575285524, "grad_norm": 0.5138705281366189, "learning_rate": 1.3392396537801882e-05, "loss": 0.3823, "step": 2339 }, { "epoch": 0.7554784293151459, "grad_norm": 0.42030094206566054, "learning_rate": 1.3390744565729686e-05, "loss": 0.3948, "step": 2340 }, { "epoch": 0.7558012833447677, "grad_norm": 0.31267837204992444, "learning_rate": 1.3389091847308264e-05, "loss": 0.3591, "step": 2341 }, { "epoch": 0.7561241373743895, "grad_norm": 0.4583174133595387, "learning_rate": 1.3387438382747015e-05, "loss": 0.3323, "step": 2342 }, { "epoch": 0.7564469914040115, "grad_norm": 0.5127269254534922, "learning_rate": 1.338578417225543e-05, "loss": 0.3488, "step": 2343 }, { "epoch": 0.7567698454336333, "grad_norm": 0.35613038055369295, "learning_rate": 1.3384129216043097e-05, "loss": 0.3616, "step": 2344 }, { "epoch": 0.7570926994632552, "grad_norm": 1.4217562302018372, "learning_rate": 1.3382473514319694e-05, "loss": 0.3615, "step": 2345 }, { "epoch": 0.757415553492877, "grad_norm": 0.4559847961662435, "learning_rate": 1.3380817067294998e-05, "loss": 0.3533, "step": 2346 }, { "epoch": 0.7577384075224989, "grad_norm": 0.4894579106387931, "learning_rate": 1.3379159875178878e-05, "loss": 0.329, "step": 2347 }, { "epoch": 0.7580612615521207, "grad_norm": 0.4392469048025786, "learning_rate": 1.3377501938181298e-05, "loss": 0.4008, "step": 2348 }, { "epoch": 0.7583841155817426, "grad_norm": 0.4270487078917668, "learning_rate": 1.3375843256512314e-05, "loss": 0.3563, "step": 2349 }, { "epoch": 0.7587069696113644, "grad_norm": 0.3575682568842295, "learning_rate": 1.3374183830382085e-05, "loss": 0.3537, "step": 2350 }, { "epoch": 0.7590298236409864, "grad_norm": 0.38483095392067923, "learning_rate": 1.3372523660000851e-05, "loss": 0.3499, "step": 2351 }, { "epoch": 0.7593526776706082, "grad_norm": 0.5051939811206482, "learning_rate": 1.3370862745578958e-05, "loss": 0.4371, "step": 2352 }, { "epoch": 0.7596755317002301, "grad_norm": 0.379838150963619, "learning_rate": 1.3369201087326839e-05, "loss": 0.402, "step": 2353 }, { "epoch": 0.7599983857298519, "grad_norm": 0.3865224080958584, "learning_rate": 1.3367538685455024e-05, "loss": 0.3028, "step": 2354 }, { "epoch": 0.7603212397594737, "grad_norm": 0.43691861150979616, "learning_rate": 1.3365875540174139e-05, "loss": 0.3754, "step": 2355 }, { "epoch": 0.7606440937890956, "grad_norm": 0.44801399196944325, "learning_rate": 1.33642116516949e-05, "loss": 0.3709, "step": 2356 }, { "epoch": 0.7609669478187174, "grad_norm": 0.38725710145614567, "learning_rate": 1.3362547020228119e-05, "loss": 0.3539, "step": 2357 }, { "epoch": 0.7612898018483393, "grad_norm": 0.44601052935611823, "learning_rate": 1.3360881645984706e-05, "loss": 0.3945, "step": 2358 }, { "epoch": 0.7616126558779612, "grad_norm": 0.32914393540519216, "learning_rate": 1.3359215529175658e-05, "loss": 0.3272, "step": 2359 }, { "epoch": 0.7619355099075831, "grad_norm": 0.40727287797821077, "learning_rate": 1.3357548670012072e-05, "loss": 0.3411, "step": 2360 }, { "epoch": 0.7622583639372049, "grad_norm": 0.38594846905356667, "learning_rate": 1.3355881068705134e-05, "loss": 0.366, "step": 2361 }, { "epoch": 0.7625812179668268, "grad_norm": 0.4804524432680785, "learning_rate": 1.3354212725466131e-05, "loss": 0.3632, "step": 2362 }, { "epoch": 0.7629040719964486, "grad_norm": 0.48986585978759783, "learning_rate": 1.3352543640506434e-05, "loss": 0.3764, "step": 2363 }, { "epoch": 0.7632269260260705, "grad_norm": 0.5438097332483844, "learning_rate": 1.335087381403752e-05, "loss": 0.3496, "step": 2364 }, { "epoch": 0.7635497800556923, "grad_norm": 0.3941105696189333, "learning_rate": 1.334920324627095e-05, "loss": 0.3628, "step": 2365 }, { "epoch": 0.7638726340853141, "grad_norm": 0.5730352095554206, "learning_rate": 1.3347531937418385e-05, "loss": 0.3634, "step": 2366 }, { "epoch": 0.764195488114936, "grad_norm": 0.510995280610263, "learning_rate": 1.3345859887691574e-05, "loss": 0.3642, "step": 2367 }, { "epoch": 0.7645183421445579, "grad_norm": 0.585026793187949, "learning_rate": 1.3344187097302367e-05, "loss": 0.3794, "step": 2368 }, { "epoch": 0.7648411961741798, "grad_norm": 0.455643486599697, "learning_rate": 1.3342513566462703e-05, "loss": 0.3278, "step": 2369 }, { "epoch": 0.7651640502038016, "grad_norm": 0.391840167634878, "learning_rate": 1.3340839295384615e-05, "loss": 0.4254, "step": 2370 }, { "epoch": 0.7654869042334235, "grad_norm": 0.4142219735272355, "learning_rate": 1.333916428428023e-05, "loss": 0.352, "step": 2371 }, { "epoch": 0.7658097582630453, "grad_norm": 0.4507446929556927, "learning_rate": 1.3337488533361772e-05, "loss": 0.4294, "step": 2372 }, { "epoch": 0.7661326122926672, "grad_norm": 0.5349845537386553, "learning_rate": 1.3335812042841556e-05, "loss": 0.4843, "step": 2373 }, { "epoch": 0.766455466322289, "grad_norm": 1.3753864513398923, "learning_rate": 1.3334134812931992e-05, "loss": 0.3984, "step": 2374 }, { "epoch": 0.7667783203519108, "grad_norm": 0.36764796159012314, "learning_rate": 1.333245684384558e-05, "loss": 0.3818, "step": 2375 }, { "epoch": 0.7671011743815328, "grad_norm": 0.4192000203494573, "learning_rate": 1.3330778135794917e-05, "loss": 0.3393, "step": 2376 }, { "epoch": 0.7674240284111546, "grad_norm": 0.45979374050095373, "learning_rate": 1.3329098688992693e-05, "loss": 0.3838, "step": 2377 }, { "epoch": 0.7677468824407765, "grad_norm": 0.40126383501556373, "learning_rate": 1.3327418503651696e-05, "loss": 0.3532, "step": 2378 }, { "epoch": 0.7680697364703983, "grad_norm": 0.4437544105909978, "learning_rate": 1.3325737579984797e-05, "loss": 0.3252, "step": 2379 }, { "epoch": 0.7683925905000202, "grad_norm": 0.35695707819244465, "learning_rate": 1.3324055918204969e-05, "loss": 0.2919, "step": 2380 }, { "epoch": 0.768715444529642, "grad_norm": 0.4168575918767098, "learning_rate": 1.3322373518525275e-05, "loss": 0.421, "step": 2381 }, { "epoch": 0.7690382985592639, "grad_norm": 0.4637251105606329, "learning_rate": 1.3320690381158877e-05, "loss": 0.3632, "step": 2382 }, { "epoch": 0.7693611525888857, "grad_norm": 0.4282857625135696, "learning_rate": 1.331900650631902e-05, "loss": 0.3847, "step": 2383 }, { "epoch": 0.7696840066185076, "grad_norm": 0.4374373155636324, "learning_rate": 1.3317321894219055e-05, "loss": 0.3961, "step": 2384 }, { "epoch": 0.7700068606481295, "grad_norm": 0.3890600290005391, "learning_rate": 1.3315636545072415e-05, "loss": 0.3948, "step": 2385 }, { "epoch": 0.7703297146777514, "grad_norm": 0.33992280103282324, "learning_rate": 1.3313950459092633e-05, "loss": 0.2895, "step": 2386 }, { "epoch": 0.7706525687073732, "grad_norm": 0.5402417924115626, "learning_rate": 1.3312263636493335e-05, "loss": 0.3871, "step": 2387 }, { "epoch": 0.770975422736995, "grad_norm": 0.38592231012525857, "learning_rate": 1.331057607748824e-05, "loss": 0.3087, "step": 2388 }, { "epoch": 0.7712982767666169, "grad_norm": 0.4447924873225348, "learning_rate": 1.3308887782291155e-05, "loss": 0.3994, "step": 2389 }, { "epoch": 0.7716211307962387, "grad_norm": 0.4251826602460497, "learning_rate": 1.3307198751115988e-05, "loss": 0.3399, "step": 2390 }, { "epoch": 0.7719439848258606, "grad_norm": 0.4347908116307863, "learning_rate": 1.3305508984176736e-05, "loss": 0.3556, "step": 2391 }, { "epoch": 0.7722668388554824, "grad_norm": 0.4767515630555655, "learning_rate": 1.3303818481687493e-05, "loss": 0.3381, "step": 2392 }, { "epoch": 0.7725896928851044, "grad_norm": 0.5530877978586212, "learning_rate": 1.3302127243862437e-05, "loss": 0.3171, "step": 2393 }, { "epoch": 0.7729125469147262, "grad_norm": 0.4572595247365817, "learning_rate": 1.3300435270915851e-05, "loss": 0.3751, "step": 2394 }, { "epoch": 0.7732354009443481, "grad_norm": 0.4590699428439604, "learning_rate": 1.3298742563062104e-05, "loss": 0.3371, "step": 2395 }, { "epoch": 0.7735582549739699, "grad_norm": 0.44608359126694724, "learning_rate": 1.329704912051566e-05, "loss": 0.3634, "step": 2396 }, { "epoch": 0.7738811090035917, "grad_norm": 0.4662591525684212, "learning_rate": 1.3295354943491074e-05, "loss": 0.4208, "step": 2397 }, { "epoch": 0.7742039630332136, "grad_norm": 0.41993131156860714, "learning_rate": 1.3293660032202998e-05, "loss": 0.3025, "step": 2398 }, { "epoch": 0.7745268170628354, "grad_norm": 0.5621945577835331, "learning_rate": 1.3291964386866175e-05, "loss": 0.2941, "step": 2399 }, { "epoch": 0.7748496710924573, "grad_norm": 0.4560398484353969, "learning_rate": 1.3290268007695438e-05, "loss": 0.3694, "step": 2400 }, { "epoch": 0.7751725251220791, "grad_norm": 0.4821368118572013, "learning_rate": 1.328857089490572e-05, "loss": 0.3922, "step": 2401 }, { "epoch": 0.7754953791517011, "grad_norm": 0.5399052626837019, "learning_rate": 1.328687304871204e-05, "loss": 0.4281, "step": 2402 }, { "epoch": 0.7758182331813229, "grad_norm": 0.4483863544113169, "learning_rate": 1.3285174469329516e-05, "loss": 0.3419, "step": 2403 }, { "epoch": 0.7761410872109448, "grad_norm": 0.34961370045705265, "learning_rate": 1.3283475156973351e-05, "loss": 0.3123, "step": 2404 }, { "epoch": 0.7764639412405666, "grad_norm": 0.4334701940611788, "learning_rate": 1.3281775111858848e-05, "loss": 0.336, "step": 2405 }, { "epoch": 0.7767867952701885, "grad_norm": 0.4626358760130181, "learning_rate": 1.32800743342014e-05, "loss": 0.3507, "step": 2406 }, { "epoch": 0.7771096492998103, "grad_norm": 0.4231256897304793, "learning_rate": 1.3278372824216495e-05, "loss": 0.3499, "step": 2407 }, { "epoch": 0.7774325033294321, "grad_norm": 0.4468914666141869, "learning_rate": 1.3276670582119709e-05, "loss": 0.3404, "step": 2408 }, { "epoch": 0.777755357359054, "grad_norm": 0.4621104366319235, "learning_rate": 1.3274967608126717e-05, "loss": 0.3585, "step": 2409 }, { "epoch": 0.7780782113886759, "grad_norm": 0.38864435038924505, "learning_rate": 1.327326390245328e-05, "loss": 0.402, "step": 2410 }, { "epoch": 0.7784010654182978, "grad_norm": 0.42810448974544824, "learning_rate": 1.3271559465315259e-05, "loss": 0.3696, "step": 2411 }, { "epoch": 0.7787239194479196, "grad_norm": 0.42457669222615024, "learning_rate": 1.32698542969286e-05, "loss": 0.4115, "step": 2412 }, { "epoch": 0.7790467734775415, "grad_norm": 0.4576447454858454, "learning_rate": 1.3268148397509347e-05, "loss": 0.3302, "step": 2413 }, { "epoch": 0.7793696275071633, "grad_norm": 0.5229818456177993, "learning_rate": 1.3266441767273636e-05, "loss": 0.3848, "step": 2414 }, { "epoch": 0.7796924815367852, "grad_norm": 1.1024413663126018, "learning_rate": 1.3264734406437694e-05, "loss": 0.3173, "step": 2415 }, { "epoch": 0.780015335566407, "grad_norm": 1.179982739111261, "learning_rate": 1.326302631521784e-05, "loss": 0.3365, "step": 2416 }, { "epoch": 0.7803381895960289, "grad_norm": 0.44533698209250977, "learning_rate": 1.3261317493830489e-05, "loss": 0.3645, "step": 2417 }, { "epoch": 0.7806610436256508, "grad_norm": 0.36074195867035436, "learning_rate": 1.3259607942492144e-05, "loss": 0.286, "step": 2418 }, { "epoch": 0.7809838976552727, "grad_norm": 0.5311320584129683, "learning_rate": 1.3257897661419406e-05, "loss": 0.3655, "step": 2419 }, { "epoch": 0.7813067516848945, "grad_norm": 0.5967529239119113, "learning_rate": 1.3256186650828961e-05, "loss": 0.3465, "step": 2420 }, { "epoch": 0.7816296057145163, "grad_norm": 0.4572917989190778, "learning_rate": 1.3254474910937597e-05, "loss": 0.3697, "step": 2421 }, { "epoch": 0.7819524597441382, "grad_norm": 0.4062417546593602, "learning_rate": 1.3252762441962185e-05, "loss": 0.401, "step": 2422 }, { "epoch": 0.78227531377376, "grad_norm": 0.5378644001832541, "learning_rate": 1.3251049244119693e-05, "loss": 0.4011, "step": 2423 }, { "epoch": 0.7825981678033819, "grad_norm": 0.4964440248765672, "learning_rate": 1.3249335317627182e-05, "loss": 0.3663, "step": 2424 }, { "epoch": 0.7829210218330037, "grad_norm": 0.33730589723810345, "learning_rate": 1.3247620662701803e-05, "loss": 0.3427, "step": 2425 }, { "epoch": 0.7832438758626256, "grad_norm": 0.43273745099274985, "learning_rate": 1.3245905279560802e-05, "loss": 0.3426, "step": 2426 }, { "epoch": 0.7835667298922475, "grad_norm": 0.3488441663549438, "learning_rate": 1.3244189168421515e-05, "loss": 0.2742, "step": 2427 }, { "epoch": 0.7838895839218694, "grad_norm": 0.3792241220867103, "learning_rate": 1.3242472329501369e-05, "loss": 0.3216, "step": 2428 }, { "epoch": 0.7842124379514912, "grad_norm": 0.5041729489330368, "learning_rate": 1.3240754763017887e-05, "loss": 0.3008, "step": 2429 }, { "epoch": 0.784535291981113, "grad_norm": 0.9733111864483831, "learning_rate": 1.3239036469188683e-05, "loss": 0.3993, "step": 2430 }, { "epoch": 0.7848581460107349, "grad_norm": 0.48269624379729653, "learning_rate": 1.323731744823146e-05, "loss": 0.3925, "step": 2431 }, { "epoch": 0.7851810000403567, "grad_norm": 0.448721850724732, "learning_rate": 1.3235597700364018e-05, "loss": 0.3389, "step": 2432 }, { "epoch": 0.7855038540699786, "grad_norm": 0.4361902599707475, "learning_rate": 1.3233877225804245e-05, "loss": 0.3915, "step": 2433 }, { "epoch": 0.7858267080996004, "grad_norm": 0.43710197139956924, "learning_rate": 1.3232156024770126e-05, "loss": 0.4022, "step": 2434 }, { "epoch": 0.7861495621292224, "grad_norm": 0.38163558248585183, "learning_rate": 1.3230434097479729e-05, "loss": 0.3361, "step": 2435 }, { "epoch": 0.7864724161588442, "grad_norm": 0.4891323757824703, "learning_rate": 1.3228711444151227e-05, "loss": 0.3185, "step": 2436 }, { "epoch": 0.7867952701884661, "grad_norm": 0.5058590649837164, "learning_rate": 1.322698806500287e-05, "loss": 0.37, "step": 2437 }, { "epoch": 0.7871181242180879, "grad_norm": 0.4318204841570018, "learning_rate": 1.3225263960253015e-05, "loss": 0.3545, "step": 2438 }, { "epoch": 0.7874409782477098, "grad_norm": 0.3775870257362822, "learning_rate": 1.3223539130120097e-05, "loss": 0.3218, "step": 2439 }, { "epoch": 0.7877638322773316, "grad_norm": 0.4200457404243007, "learning_rate": 1.3221813574822655e-05, "loss": 0.3857, "step": 2440 }, { "epoch": 0.7880866863069534, "grad_norm": 0.8000144273929725, "learning_rate": 1.3220087294579311e-05, "loss": 0.3989, "step": 2441 }, { "epoch": 0.7884095403365753, "grad_norm": 0.5741964731542775, "learning_rate": 1.3218360289608787e-05, "loss": 0.4009, "step": 2442 }, { "epoch": 0.7887323943661971, "grad_norm": 0.444975774102693, "learning_rate": 1.3216632560129887e-05, "loss": 0.3309, "step": 2443 }, { "epoch": 0.7890552483958191, "grad_norm": 0.3761077099432341, "learning_rate": 1.3214904106361514e-05, "loss": 0.3705, "step": 2444 }, { "epoch": 0.7893781024254409, "grad_norm": 0.4255476349107978, "learning_rate": 1.321317492852266e-05, "loss": 0.4199, "step": 2445 }, { "epoch": 0.7897009564550628, "grad_norm": 0.3907360499611609, "learning_rate": 1.321144502683241e-05, "loss": 0.2999, "step": 2446 }, { "epoch": 0.7900238104846846, "grad_norm": 0.3782280874832732, "learning_rate": 1.3209714401509941e-05, "loss": 0.344, "step": 2447 }, { "epoch": 0.7903466645143065, "grad_norm": 0.4511118934583056, "learning_rate": 1.3207983052774518e-05, "loss": 0.3039, "step": 2448 }, { "epoch": 0.7906695185439283, "grad_norm": 0.48559930716895294, "learning_rate": 1.3206250980845504e-05, "loss": 0.4102, "step": 2449 }, { "epoch": 0.7909923725735502, "grad_norm": 0.3438444529811745, "learning_rate": 1.3204518185942351e-05, "loss": 0.3287, "step": 2450 }, { "epoch": 0.791315226603172, "grad_norm": 0.3605419544051749, "learning_rate": 1.3202784668284597e-05, "loss": 0.2942, "step": 2451 }, { "epoch": 0.791638080632794, "grad_norm": 0.5355553946951312, "learning_rate": 1.320105042809188e-05, "loss": 0.3538, "step": 2452 }, { "epoch": 0.7919609346624158, "grad_norm": 0.40893951878980533, "learning_rate": 1.3199315465583925e-05, "loss": 0.332, "step": 2453 }, { "epoch": 0.7922837886920376, "grad_norm": 0.5150087904896505, "learning_rate": 1.3197579780980547e-05, "loss": 0.3566, "step": 2454 }, { "epoch": 0.7926066427216595, "grad_norm": 0.470662028364368, "learning_rate": 1.3195843374501654e-05, "loss": 0.3534, "step": 2455 }, { "epoch": 0.7929294967512813, "grad_norm": 0.4876087676296451, "learning_rate": 1.3194106246367255e-05, "loss": 0.3425, "step": 2456 }, { "epoch": 0.7932523507809032, "grad_norm": 0.44711495675916674, "learning_rate": 1.3192368396797434e-05, "loss": 0.4052, "step": 2457 }, { "epoch": 0.793575204810525, "grad_norm": 0.4453751153571868, "learning_rate": 1.3190629826012376e-05, "loss": 0.2884, "step": 2458 }, { "epoch": 0.7938980588401469, "grad_norm": 0.7927211710464777, "learning_rate": 1.3188890534232353e-05, "loss": 0.4125, "step": 2459 }, { "epoch": 0.7942209128697687, "grad_norm": 0.41249932604219747, "learning_rate": 1.3187150521677736e-05, "loss": 0.3726, "step": 2460 }, { "epoch": 0.7945437668993907, "grad_norm": 0.4329923516208189, "learning_rate": 1.3185409788568979e-05, "loss": 0.3603, "step": 2461 }, { "epoch": 0.7948666209290125, "grad_norm": 0.46831487572860137, "learning_rate": 1.3183668335126632e-05, "loss": 0.3515, "step": 2462 }, { "epoch": 0.7951894749586343, "grad_norm": 0.4177323594322801, "learning_rate": 1.3181926161571332e-05, "loss": 0.3942, "step": 2463 }, { "epoch": 0.7955123289882562, "grad_norm": 0.4165076247051086, "learning_rate": 1.3180183268123812e-05, "loss": 0.313, "step": 2464 }, { "epoch": 0.795835183017878, "grad_norm": 0.4341886955271393, "learning_rate": 1.3178439655004895e-05, "loss": 0.361, "step": 2465 }, { "epoch": 0.7961580370474999, "grad_norm": 0.5465955497787363, "learning_rate": 1.3176695322435494e-05, "loss": 0.3798, "step": 2466 }, { "epoch": 0.7964808910771217, "grad_norm": 0.5147885796060432, "learning_rate": 1.3174950270636612e-05, "loss": 0.3669, "step": 2467 }, { "epoch": 0.7968037451067436, "grad_norm": 0.4712339135080035, "learning_rate": 1.3173204499829345e-05, "loss": 0.3777, "step": 2468 }, { "epoch": 0.7971265991363655, "grad_norm": 0.477359935911043, "learning_rate": 1.317145801023488e-05, "loss": 0.4221, "step": 2469 }, { "epoch": 0.7974494531659874, "grad_norm": 0.4622485103657463, "learning_rate": 1.3169710802074496e-05, "loss": 0.3086, "step": 2470 }, { "epoch": 0.7977723071956092, "grad_norm": 0.5891656272392374, "learning_rate": 1.3167962875569563e-05, "loss": 0.3621, "step": 2471 }, { "epoch": 0.798095161225231, "grad_norm": 0.4858313436908084, "learning_rate": 1.3166214230941535e-05, "loss": 0.3645, "step": 2472 }, { "epoch": 0.7984180152548529, "grad_norm": 0.5227160265924301, "learning_rate": 1.3164464868411971e-05, "loss": 0.3655, "step": 2473 }, { "epoch": 0.7987408692844747, "grad_norm": 0.4265993547598817, "learning_rate": 1.3162714788202504e-05, "loss": 0.3521, "step": 2474 }, { "epoch": 0.7990637233140966, "grad_norm": 0.46960360271622625, "learning_rate": 1.3160963990534874e-05, "loss": 0.3509, "step": 2475 }, { "epoch": 0.7993865773437184, "grad_norm": 0.38072439056450497, "learning_rate": 1.3159212475630903e-05, "loss": 0.3544, "step": 2476 }, { "epoch": 0.7997094313733403, "grad_norm": 0.42880800212024295, "learning_rate": 1.3157460243712503e-05, "loss": 0.4301, "step": 2477 }, { "epoch": 0.8000322854029622, "grad_norm": 0.45443219357695713, "learning_rate": 1.3155707295001681e-05, "loss": 0.3825, "step": 2478 }, { "epoch": 0.8003551394325841, "grad_norm": 0.4645200107117853, "learning_rate": 1.3153953629720532e-05, "loss": 0.3699, "step": 2479 }, { "epoch": 0.8006779934622059, "grad_norm": 0.40766566624358713, "learning_rate": 1.3152199248091247e-05, "loss": 0.4143, "step": 2480 }, { "epoch": 0.8010008474918278, "grad_norm": 0.39057649335596895, "learning_rate": 1.3150444150336099e-05, "loss": 0.3293, "step": 2481 }, { "epoch": 0.8013237015214496, "grad_norm": 0.4943257290073196, "learning_rate": 1.3148688336677459e-05, "loss": 0.5035, "step": 2482 }, { "epoch": 0.8016465555510714, "grad_norm": 0.3840375242380603, "learning_rate": 1.3146931807337784e-05, "loss": 0.3501, "step": 2483 }, { "epoch": 0.8019694095806933, "grad_norm": 0.43992787782394105, "learning_rate": 1.3145174562539625e-05, "loss": 0.3657, "step": 2484 }, { "epoch": 0.8022922636103151, "grad_norm": 0.43016712697118437, "learning_rate": 1.3143416602505625e-05, "loss": 0.3604, "step": 2485 }, { "epoch": 0.8026151176399371, "grad_norm": 0.3536338673794118, "learning_rate": 1.3141657927458513e-05, "loss": 0.3855, "step": 2486 }, { "epoch": 0.8029379716695589, "grad_norm": 0.3797764750499699, "learning_rate": 1.313989853762111e-05, "loss": 0.3278, "step": 2487 }, { "epoch": 0.8032608256991808, "grad_norm": 0.4584652987782699, "learning_rate": 1.3138138433216325e-05, "loss": 0.3317, "step": 2488 }, { "epoch": 0.8035836797288026, "grad_norm": 0.3811454043153227, "learning_rate": 1.313637761446717e-05, "loss": 0.3415, "step": 2489 }, { "epoch": 0.8039065337584245, "grad_norm": 0.4091778500781509, "learning_rate": 1.313461608159673e-05, "loss": 0.3194, "step": 2490 }, { "epoch": 0.8042293877880463, "grad_norm": 0.3788658869070286, "learning_rate": 1.3132853834828194e-05, "loss": 0.3019, "step": 2491 }, { "epoch": 0.8045522418176682, "grad_norm": 0.3195463045451802, "learning_rate": 1.3131090874384833e-05, "loss": 0.349, "step": 2492 }, { "epoch": 0.80487509584729, "grad_norm": 0.43530632053096, "learning_rate": 1.3129327200490012e-05, "loss": 0.3967, "step": 2493 }, { "epoch": 0.805197949876912, "grad_norm": 0.3967957532212307, "learning_rate": 1.3127562813367188e-05, "loss": 0.3444, "step": 2494 }, { "epoch": 0.8055208039065338, "grad_norm": 0.4327007554655154, "learning_rate": 1.3125797713239905e-05, "loss": 0.2868, "step": 2495 }, { "epoch": 0.8058436579361556, "grad_norm": 0.4055580206173895, "learning_rate": 1.3124031900331797e-05, "loss": 0.4101, "step": 2496 }, { "epoch": 0.8061665119657775, "grad_norm": 0.3627353558985153, "learning_rate": 1.3122265374866593e-05, "loss": 0.3689, "step": 2497 }, { "epoch": 0.8064893659953993, "grad_norm": 0.5028674840760969, "learning_rate": 1.3120498137068109e-05, "loss": 0.3121, "step": 2498 }, { "epoch": 0.8068122200250212, "grad_norm": 0.34758580670074535, "learning_rate": 1.311873018716025e-05, "loss": 0.3952, "step": 2499 }, { "epoch": 0.807135074054643, "grad_norm": 0.3890252664820529, "learning_rate": 1.3116961525367015e-05, "loss": 0.3247, "step": 2500 }, { "epoch": 0.8074579280842649, "grad_norm": 0.4043827979282162, "learning_rate": 1.311519215191249e-05, "loss": 0.3706, "step": 2501 }, { "epoch": 0.8077807821138867, "grad_norm": 0.33599853799725304, "learning_rate": 1.3113422067020849e-05, "loss": 0.3301, "step": 2502 }, { "epoch": 0.8081036361435087, "grad_norm": 0.3894021468274031, "learning_rate": 1.3111651270916366e-05, "loss": 0.3061, "step": 2503 }, { "epoch": 0.8084264901731305, "grad_norm": 0.43294451949780843, "learning_rate": 1.3109879763823394e-05, "loss": 0.337, "step": 2504 }, { "epoch": 0.8087493442027524, "grad_norm": 0.4489244734261314, "learning_rate": 1.310810754596638e-05, "loss": 0.3276, "step": 2505 }, { "epoch": 0.8090721982323742, "grad_norm": 0.5855891082654686, "learning_rate": 1.3106334617569862e-05, "loss": 0.4119, "step": 2506 }, { "epoch": 0.809395052261996, "grad_norm": 0.3835315941993755, "learning_rate": 1.3104560978858469e-05, "loss": 0.4507, "step": 2507 }, { "epoch": 0.8097179062916179, "grad_norm": 0.5114313928121036, "learning_rate": 1.3102786630056919e-05, "loss": 0.3343, "step": 2508 }, { "epoch": 0.8100407603212397, "grad_norm": 0.4532034019882009, "learning_rate": 1.3101011571390017e-05, "loss": 0.349, "step": 2509 }, { "epoch": 0.8103636143508616, "grad_norm": 0.4736702390081286, "learning_rate": 1.3099235803082663e-05, "loss": 0.412, "step": 2510 }, { "epoch": 0.8106864683804835, "grad_norm": 0.4078853599413359, "learning_rate": 1.3097459325359842e-05, "loss": 0.3829, "step": 2511 }, { "epoch": 0.8110093224101054, "grad_norm": 0.46923868336974406, "learning_rate": 1.3095682138446635e-05, "loss": 0.3729, "step": 2512 }, { "epoch": 0.8113321764397272, "grad_norm": 0.49005966276812746, "learning_rate": 1.3093904242568204e-05, "loss": 0.3727, "step": 2513 }, { "epoch": 0.8116550304693491, "grad_norm": 0.39354049522977125, "learning_rate": 1.3092125637949811e-05, "loss": 0.3248, "step": 2514 }, { "epoch": 0.8119778844989709, "grad_norm": 0.4382045097443392, "learning_rate": 1.30903463248168e-05, "loss": 0.3694, "step": 2515 }, { "epoch": 0.8123007385285927, "grad_norm": 0.3915517587272867, "learning_rate": 1.3088566303394607e-05, "loss": 0.3149, "step": 2516 }, { "epoch": 0.8126235925582146, "grad_norm": 0.4193518549245794, "learning_rate": 1.308678557390876e-05, "loss": 0.3541, "step": 2517 }, { "epoch": 0.8129464465878364, "grad_norm": 0.475283379803755, "learning_rate": 1.3085004136584874e-05, "loss": 0.3304, "step": 2518 }, { "epoch": 0.8132693006174583, "grad_norm": 0.4380633214589404, "learning_rate": 1.3083221991648655e-05, "loss": 0.3253, "step": 2519 }, { "epoch": 0.8135921546470802, "grad_norm": 0.4520959267786747, "learning_rate": 1.3081439139325898e-05, "loss": 0.3451, "step": 2520 }, { "epoch": 0.8139150086767021, "grad_norm": 0.49104619848759035, "learning_rate": 1.3079655579842487e-05, "loss": 0.3477, "step": 2521 }, { "epoch": 0.8142378627063239, "grad_norm": 0.41185678916621127, "learning_rate": 1.30778713134244e-05, "loss": 0.3473, "step": 2522 }, { "epoch": 0.8145607167359458, "grad_norm": 0.4254036815429396, "learning_rate": 1.3076086340297699e-05, "loss": 0.3494, "step": 2523 }, { "epoch": 0.8148835707655676, "grad_norm": 0.44959440165413456, "learning_rate": 1.3074300660688536e-05, "loss": 0.3335, "step": 2524 }, { "epoch": 0.8152064247951895, "grad_norm": 0.4549992102287127, "learning_rate": 1.3072514274823156e-05, "loss": 0.3006, "step": 2525 }, { "epoch": 0.8155292788248113, "grad_norm": 0.4790023074057204, "learning_rate": 1.3070727182927889e-05, "loss": 0.3131, "step": 2526 }, { "epoch": 0.8158521328544331, "grad_norm": 0.3890907681762797, "learning_rate": 1.3068939385229164e-05, "loss": 0.3775, "step": 2527 }, { "epoch": 0.8161749868840551, "grad_norm": 0.3910179682975357, "learning_rate": 1.3067150881953484e-05, "loss": 0.2834, "step": 2528 }, { "epoch": 0.8164978409136769, "grad_norm": 0.44807482890942224, "learning_rate": 1.3065361673327454e-05, "loss": 0.3966, "step": 2529 }, { "epoch": 0.8168206949432988, "grad_norm": 0.3973115590802345, "learning_rate": 1.3063571759577765e-05, "loss": 0.4111, "step": 2530 }, { "epoch": 0.8171435489729206, "grad_norm": 0.39076599410949275, "learning_rate": 1.3061781140931196e-05, "loss": 0.3949, "step": 2531 }, { "epoch": 0.8174664030025425, "grad_norm": 0.4030583256540305, "learning_rate": 1.3059989817614614e-05, "loss": 0.3235, "step": 2532 }, { "epoch": 0.8177892570321643, "grad_norm": 0.4901670139428273, "learning_rate": 1.3058197789854979e-05, "loss": 0.3723, "step": 2533 }, { "epoch": 0.8181121110617862, "grad_norm": 0.47303986369772855, "learning_rate": 1.305640505787934e-05, "loss": 0.331, "step": 2534 }, { "epoch": 0.818434965091408, "grad_norm": 0.4244119089718119, "learning_rate": 1.305461162191483e-05, "loss": 0.3304, "step": 2535 }, { "epoch": 0.8187578191210299, "grad_norm": 0.5033979612165722, "learning_rate": 1.3052817482188677e-05, "loss": 0.3396, "step": 2536 }, { "epoch": 0.8190806731506518, "grad_norm": 0.39132149889526285, "learning_rate": 1.3051022638928193e-05, "loss": 0.3228, "step": 2537 }, { "epoch": 0.8194035271802736, "grad_norm": 0.3928728969844888, "learning_rate": 1.3049227092360789e-05, "loss": 0.3669, "step": 2538 }, { "epoch": 0.8197263812098955, "grad_norm": 0.4917846088493621, "learning_rate": 1.3047430842713954e-05, "loss": 0.3022, "step": 2539 }, { "epoch": 0.8200492352395173, "grad_norm": 0.42519176991891416, "learning_rate": 1.304563389021527e-05, "loss": 0.3746, "step": 2540 }, { "epoch": 0.8203720892691392, "grad_norm": 0.46792071770051136, "learning_rate": 1.3043836235092406e-05, "loss": 0.3476, "step": 2541 }, { "epoch": 0.820694943298761, "grad_norm": 0.445173452140247, "learning_rate": 1.3042037877573128e-05, "loss": 0.3801, "step": 2542 }, { "epoch": 0.8210177973283829, "grad_norm": 0.4044733773527878, "learning_rate": 1.3040238817885281e-05, "loss": 0.3364, "step": 2543 }, { "epoch": 0.8213406513580047, "grad_norm": 0.4213958051306942, "learning_rate": 1.3038439056256808e-05, "loss": 0.3369, "step": 2544 }, { "epoch": 0.8216635053876267, "grad_norm": 0.37304971067145654, "learning_rate": 1.303663859291573e-05, "loss": 0.3348, "step": 2545 }, { "epoch": 0.8219863594172485, "grad_norm": 0.4865003240922556, "learning_rate": 1.3034837428090169e-05, "loss": 0.4031, "step": 2546 }, { "epoch": 0.8223092134468704, "grad_norm": 0.37540480009312627, "learning_rate": 1.3033035562008328e-05, "loss": 0.3344, "step": 2547 }, { "epoch": 0.8226320674764922, "grad_norm": 0.44504292692160996, "learning_rate": 1.3031232994898499e-05, "loss": 0.347, "step": 2548 }, { "epoch": 0.822954921506114, "grad_norm": 0.38671353017704696, "learning_rate": 1.3029429726989067e-05, "loss": 0.3971, "step": 2549 }, { "epoch": 0.8232777755357359, "grad_norm": 0.3780855690416977, "learning_rate": 1.3027625758508504e-05, "loss": 0.3677, "step": 2550 }, { "epoch": 0.8236006295653577, "grad_norm": 0.4261298133079884, "learning_rate": 1.302582108968537e-05, "loss": 0.3813, "step": 2551 }, { "epoch": 0.8239234835949796, "grad_norm": 0.5551977134472368, "learning_rate": 1.302401572074831e-05, "loss": 0.417, "step": 2552 }, { "epoch": 0.8242463376246014, "grad_norm": 0.38807480282593265, "learning_rate": 1.3022209651926069e-05, "loss": 0.3784, "step": 2553 }, { "epoch": 0.8245691916542234, "grad_norm": 0.4589593405934984, "learning_rate": 1.3020402883447468e-05, "loss": 0.3763, "step": 2554 }, { "epoch": 0.8248920456838452, "grad_norm": 0.4266832468208747, "learning_rate": 1.3018595415541425e-05, "loss": 0.3069, "step": 2555 }, { "epoch": 0.8252148997134671, "grad_norm": 0.3869147314838819, "learning_rate": 1.301678724843694e-05, "loss": 0.3892, "step": 2556 }, { "epoch": 0.8255377537430889, "grad_norm": 0.3245573720347796, "learning_rate": 1.3014978382363108e-05, "loss": 0.3416, "step": 2557 }, { "epoch": 0.8258606077727108, "grad_norm": 0.3831798862725376, "learning_rate": 1.3013168817549113e-05, "loss": 0.3354, "step": 2558 }, { "epoch": 0.8261834618023326, "grad_norm": 0.5297273427615955, "learning_rate": 1.3011358554224217e-05, "loss": 0.3598, "step": 2559 }, { "epoch": 0.8265063158319544, "grad_norm": 0.4041920931688831, "learning_rate": 1.3009547592617784e-05, "loss": 0.3433, "step": 2560 }, { "epoch": 0.8268291698615763, "grad_norm": 0.4515144069354868, "learning_rate": 1.3007735932959256e-05, "loss": 0.4106, "step": 2561 }, { "epoch": 0.8271520238911982, "grad_norm": 0.4174930312095323, "learning_rate": 1.3005923575478172e-05, "loss": 0.3445, "step": 2562 }, { "epoch": 0.8274748779208201, "grad_norm": 0.4660234662687542, "learning_rate": 1.3004110520404151e-05, "loss": 0.3954, "step": 2563 }, { "epoch": 0.8277977319504419, "grad_norm": 0.3989658155867821, "learning_rate": 1.3002296767966909e-05, "loss": 0.3841, "step": 2564 }, { "epoch": 0.8281205859800638, "grad_norm": 0.5578587924486397, "learning_rate": 1.3000482318396244e-05, "loss": 0.3284, "step": 2565 }, { "epoch": 0.8284434400096856, "grad_norm": 0.477487169525381, "learning_rate": 1.2998667171922041e-05, "loss": 0.3585, "step": 2566 }, { "epoch": 0.8287662940393075, "grad_norm": 0.3992216303704858, "learning_rate": 1.299685132877428e-05, "loss": 0.3482, "step": 2567 }, { "epoch": 0.8290891480689293, "grad_norm": 0.4302985167317952, "learning_rate": 1.299503478918303e-05, "loss": 0.2821, "step": 2568 }, { "epoch": 0.8294120020985511, "grad_norm": 0.39796353934752376, "learning_rate": 1.2993217553378433e-05, "loss": 0.3805, "step": 2569 }, { "epoch": 0.8297348561281731, "grad_norm": 0.35470338498321397, "learning_rate": 1.299139962159074e-05, "loss": 0.2568, "step": 2570 }, { "epoch": 0.830057710157795, "grad_norm": 0.4225096017205905, "learning_rate": 1.2989580994050276e-05, "loss": 0.4158, "step": 2571 }, { "epoch": 0.8303805641874168, "grad_norm": 0.4535484996511939, "learning_rate": 1.298776167098746e-05, "loss": 0.4418, "step": 2572 }, { "epoch": 0.8307034182170386, "grad_norm": 0.39018968094553447, "learning_rate": 1.2985941652632797e-05, "loss": 0.4316, "step": 2573 }, { "epoch": 0.8310262722466605, "grad_norm": 0.43018054588019083, "learning_rate": 1.298412093921688e-05, "loss": 0.4118, "step": 2574 }, { "epoch": 0.8313491262762823, "grad_norm": 0.3940470892519734, "learning_rate": 1.2982299530970394e-05, "loss": 0.3861, "step": 2575 }, { "epoch": 0.8316719803059042, "grad_norm": 0.60255491544838, "learning_rate": 1.2980477428124107e-05, "loss": 0.3071, "step": 2576 }, { "epoch": 0.831994834335526, "grad_norm": 0.3342320818254577, "learning_rate": 1.2978654630908875e-05, "loss": 0.3742, "step": 2577 }, { "epoch": 0.8323176883651479, "grad_norm": 0.3402954558013053, "learning_rate": 1.297683113955565e-05, "loss": 0.3109, "step": 2578 }, { "epoch": 0.8326405423947698, "grad_norm": 0.4379300809782196, "learning_rate": 1.2975006954295457e-05, "loss": 0.3359, "step": 2579 }, { "epoch": 0.8329633964243917, "grad_norm": 0.3653333584767635, "learning_rate": 1.2973182075359424e-05, "loss": 0.4157, "step": 2580 }, { "epoch": 0.8332862504540135, "grad_norm": 0.40133157376389694, "learning_rate": 1.297135650297876e-05, "loss": 0.282, "step": 2581 }, { "epoch": 0.8336091044836353, "grad_norm": 0.3809888851910853, "learning_rate": 1.296953023738476e-05, "loss": 0.4081, "step": 2582 }, { "epoch": 0.8339319585132572, "grad_norm": 0.4057240715323932, "learning_rate": 1.2967703278808814e-05, "loss": 0.3801, "step": 2583 }, { "epoch": 0.834254812542879, "grad_norm": 0.4484309684420719, "learning_rate": 1.2965875627482387e-05, "loss": 0.3219, "step": 2584 }, { "epoch": 0.8345776665725009, "grad_norm": 0.4907936168269705, "learning_rate": 1.296404728363705e-05, "loss": 0.359, "step": 2585 }, { "epoch": 0.8349005206021227, "grad_norm": 0.38290929504193805, "learning_rate": 1.2962218247504444e-05, "loss": 0.3843, "step": 2586 }, { "epoch": 0.8352233746317447, "grad_norm": 0.8411646973904213, "learning_rate": 1.2960388519316308e-05, "loss": 0.3977, "step": 2587 }, { "epoch": 0.8355462286613665, "grad_norm": 0.4417917022487773, "learning_rate": 1.2958558099304466e-05, "loss": 0.3543, "step": 2588 }, { "epoch": 0.8358690826909884, "grad_norm": 0.42898314105387647, "learning_rate": 1.295672698770083e-05, "loss": 0.3957, "step": 2589 }, { "epoch": 0.8361919367206102, "grad_norm": 0.47001767721362536, "learning_rate": 1.29548951847374e-05, "loss": 0.4353, "step": 2590 }, { "epoch": 0.836514790750232, "grad_norm": 0.4123965125877992, "learning_rate": 1.2953062690646261e-05, "loss": 0.3533, "step": 2591 }, { "epoch": 0.8368376447798539, "grad_norm": 0.4290602168502843, "learning_rate": 1.295122950565959e-05, "loss": 0.3317, "step": 2592 }, { "epoch": 0.8371604988094757, "grad_norm": 0.38313265250936757, "learning_rate": 1.2949395630009647e-05, "loss": 0.4107, "step": 2593 }, { "epoch": 0.8374833528390976, "grad_norm": 0.43618453679709596, "learning_rate": 1.2947561063928778e-05, "loss": 0.3798, "step": 2594 }, { "epoch": 0.8378062068687194, "grad_norm": 0.4509655297453455, "learning_rate": 1.2945725807649427e-05, "loss": 0.3222, "step": 2595 }, { "epoch": 0.8381290608983414, "grad_norm": 0.46964484429214576, "learning_rate": 1.2943889861404115e-05, "loss": 0.2929, "step": 2596 }, { "epoch": 0.8384519149279632, "grad_norm": 0.4548133383569808, "learning_rate": 1.2942053225425455e-05, "loss": 0.3595, "step": 2597 }, { "epoch": 0.8387747689575851, "grad_norm": 0.47517152998279516, "learning_rate": 1.2940215899946145e-05, "loss": 0.3641, "step": 2598 }, { "epoch": 0.8390976229872069, "grad_norm": 0.5922699625812613, "learning_rate": 1.2938377885198973e-05, "loss": 0.356, "step": 2599 }, { "epoch": 0.8394204770168288, "grad_norm": 0.4339369008650523, "learning_rate": 1.293653918141681e-05, "loss": 0.3759, "step": 2600 }, { "epoch": 0.8397433310464506, "grad_norm": 0.49624541582184156, "learning_rate": 1.2934699788832621e-05, "loss": 0.4508, "step": 2601 }, { "epoch": 0.8400661850760724, "grad_norm": 0.4762445009333319, "learning_rate": 1.2932859707679452e-05, "loss": 0.3651, "step": 2602 }, { "epoch": 0.8403890391056943, "grad_norm": 0.4526548590565446, "learning_rate": 1.293101893819044e-05, "loss": 0.4546, "step": 2603 }, { "epoch": 0.8407118931353162, "grad_norm": 0.38131967913294607, "learning_rate": 1.2929177480598807e-05, "loss": 0.3903, "step": 2604 }, { "epoch": 0.8410347471649381, "grad_norm": 0.346259305775013, "learning_rate": 1.2927335335137864e-05, "loss": 0.3552, "step": 2605 }, { "epoch": 0.8413576011945599, "grad_norm": 0.6052366393874633, "learning_rate": 1.2925492502041007e-05, "loss": 0.3973, "step": 2606 }, { "epoch": 0.8416804552241818, "grad_norm": 0.4021268785561407, "learning_rate": 1.2923648981541722e-05, "loss": 0.3801, "step": 2607 }, { "epoch": 0.8420033092538036, "grad_norm": 0.4775538918695294, "learning_rate": 1.2921804773873581e-05, "loss": 0.3292, "step": 2608 }, { "epoch": 0.8423261632834255, "grad_norm": 0.4403838363578718, "learning_rate": 1.291995987927024e-05, "loss": 0.4444, "step": 2609 }, { "epoch": 0.8426490173130473, "grad_norm": 0.5758170601976673, "learning_rate": 1.2918114297965446e-05, "loss": 0.3564, "step": 2610 }, { "epoch": 0.8429718713426692, "grad_norm": 0.4090671936709387, "learning_rate": 1.2916268030193033e-05, "loss": 0.3327, "step": 2611 }, { "epoch": 0.843294725372291, "grad_norm": 0.4157936801630739, "learning_rate": 1.2914421076186916e-05, "loss": 0.3634, "step": 2612 }, { "epoch": 0.843617579401913, "grad_norm": 0.5252767554400986, "learning_rate": 1.2912573436181107e-05, "loss": 0.3705, "step": 2613 }, { "epoch": 0.8439404334315348, "grad_norm": 0.41992455420846486, "learning_rate": 1.2910725110409696e-05, "loss": 0.422, "step": 2614 }, { "epoch": 0.8442632874611566, "grad_norm": 0.4821043752142886, "learning_rate": 1.2908876099106867e-05, "loss": 0.2975, "step": 2615 }, { "epoch": 0.8445861414907785, "grad_norm": 0.43801981557126657, "learning_rate": 1.2907026402506883e-05, "loss": 0.326, "step": 2616 }, { "epoch": 0.8449089955204003, "grad_norm": 0.4496522549160452, "learning_rate": 1.2905176020844101e-05, "loss": 0.3991, "step": 2617 }, { "epoch": 0.8452318495500222, "grad_norm": 0.41699700146189744, "learning_rate": 1.2903324954352961e-05, "loss": 0.3717, "step": 2618 }, { "epoch": 0.845554703579644, "grad_norm": 0.3192696011124786, "learning_rate": 1.2901473203267988e-05, "loss": 0.3414, "step": 2619 }, { "epoch": 0.8458775576092659, "grad_norm": 0.47874004509778545, "learning_rate": 1.28996207678238e-05, "loss": 0.3593, "step": 2620 }, { "epoch": 0.8462004116388878, "grad_norm": 0.5170953056302647, "learning_rate": 1.2897767648255094e-05, "loss": 0.3955, "step": 2621 }, { "epoch": 0.8465232656685097, "grad_norm": 0.3620141171668808, "learning_rate": 1.289591384479666e-05, "loss": 0.3041, "step": 2622 }, { "epoch": 0.8468461196981315, "grad_norm": 0.45485567339491323, "learning_rate": 1.2894059357683376e-05, "loss": 0.3263, "step": 2623 }, { "epoch": 0.8471689737277533, "grad_norm": 0.4593532268739952, "learning_rate": 1.2892204187150198e-05, "loss": 0.3707, "step": 2624 }, { "epoch": 0.8474918277573752, "grad_norm": 0.48661890366145844, "learning_rate": 1.2890348333432174e-05, "loss": 0.3858, "step": 2625 }, { "epoch": 0.847814681786997, "grad_norm": 0.42847102330720377, "learning_rate": 1.2888491796764441e-05, "loss": 0.3236, "step": 2626 }, { "epoch": 0.8481375358166189, "grad_norm": 0.33857319524675256, "learning_rate": 1.2886634577382215e-05, "loss": 0.2904, "step": 2627 }, { "epoch": 0.8484603898462407, "grad_norm": 0.5599128313609383, "learning_rate": 1.2884776675520807e-05, "loss": 0.4016, "step": 2628 }, { "epoch": 0.8487832438758627, "grad_norm": 0.4624770924295889, "learning_rate": 1.2882918091415612e-05, "loss": 0.3333, "step": 2629 }, { "epoch": 0.8491060979054845, "grad_norm": 0.5063730089282789, "learning_rate": 1.2881058825302103e-05, "loss": 0.3788, "step": 2630 }, { "epoch": 0.8494289519351064, "grad_norm": 0.4594913274129024, "learning_rate": 1.2879198877415853e-05, "loss": 0.3302, "step": 2631 }, { "epoch": 0.8497518059647282, "grad_norm": 0.4637874264494957, "learning_rate": 1.2877338247992514e-05, "loss": 0.3597, "step": 2632 }, { "epoch": 0.8500746599943501, "grad_norm": 0.41454498849148674, "learning_rate": 1.2875476937267825e-05, "loss": 0.3319, "step": 2633 }, { "epoch": 0.8503975140239719, "grad_norm": 0.3975320594898808, "learning_rate": 1.2873614945477608e-05, "loss": 0.3382, "step": 2634 }, { "epoch": 0.8507203680535937, "grad_norm": 0.4336278078310516, "learning_rate": 1.2871752272857779e-05, "loss": 0.4068, "step": 2635 }, { "epoch": 0.8510432220832156, "grad_norm": 0.46323126296726974, "learning_rate": 1.286988891964433e-05, "loss": 0.3254, "step": 2636 }, { "epoch": 0.8513660761128374, "grad_norm": 0.39046010492386446, "learning_rate": 1.2868024886073355e-05, "loss": 0.2829, "step": 2637 }, { "epoch": 0.8516889301424594, "grad_norm": 0.54980596676775, "learning_rate": 1.2866160172381018e-05, "loss": 0.4152, "step": 2638 }, { "epoch": 0.8520117841720812, "grad_norm": 0.42025439413553917, "learning_rate": 1.2864294778803576e-05, "loss": 0.415, "step": 2639 }, { "epoch": 0.8523346382017031, "grad_norm": 0.3718549006929693, "learning_rate": 1.2862428705577372e-05, "loss": 0.3877, "step": 2640 }, { "epoch": 0.8526574922313249, "grad_norm": 0.4737453520988824, "learning_rate": 1.2860561952938837e-05, "loss": 0.3578, "step": 2641 }, { "epoch": 0.8529803462609468, "grad_norm": 0.4554123364045124, "learning_rate": 1.2858694521124484e-05, "loss": 0.3501, "step": 2642 }, { "epoch": 0.8533032002905686, "grad_norm": 0.4748718353087352, "learning_rate": 1.2856826410370912e-05, "loss": 0.3265, "step": 2643 }, { "epoch": 0.8536260543201905, "grad_norm": 0.4294317135200271, "learning_rate": 1.2854957620914814e-05, "loss": 0.3406, "step": 2644 }, { "epoch": 0.8539489083498123, "grad_norm": 0.4634558666597168, "learning_rate": 1.2853088152992959e-05, "loss": 0.3668, "step": 2645 }, { "epoch": 0.8542717623794343, "grad_norm": 0.4075924366714571, "learning_rate": 1.2851218006842206e-05, "loss": 0.3243, "step": 2646 }, { "epoch": 0.8545946164090561, "grad_norm": 0.43845788928057927, "learning_rate": 1.28493471826995e-05, "loss": 0.367, "step": 2647 }, { "epoch": 0.8549174704386779, "grad_norm": 0.4241554441756401, "learning_rate": 1.2847475680801874e-05, "loss": 0.4181, "step": 2648 }, { "epoch": 0.8552403244682998, "grad_norm": 0.43931795114228456, "learning_rate": 1.2845603501386443e-05, "loss": 0.3328, "step": 2649 }, { "epoch": 0.8555631784979216, "grad_norm": 0.43415725881066125, "learning_rate": 1.2843730644690407e-05, "loss": 0.3382, "step": 2650 }, { "epoch": 0.8558860325275435, "grad_norm": 0.4338682543858008, "learning_rate": 1.2841857110951059e-05, "loss": 0.3604, "step": 2651 }, { "epoch": 0.8562088865571653, "grad_norm": 0.53915057892568, "learning_rate": 1.283998290040577e-05, "loss": 0.4263, "step": 2652 }, { "epoch": 0.8565317405867872, "grad_norm": 0.42691258043059904, "learning_rate": 1.2838108013292005e-05, "loss": 0.3111, "step": 2653 }, { "epoch": 0.856854594616409, "grad_norm": 0.4707301323886959, "learning_rate": 1.2836232449847303e-05, "loss": 0.3754, "step": 2654 }, { "epoch": 0.857177448646031, "grad_norm": 0.46722370918565714, "learning_rate": 1.2834356210309299e-05, "loss": 0.3512, "step": 2655 }, { "epoch": 0.8575003026756528, "grad_norm": 0.42983029120379884, "learning_rate": 1.2832479294915707e-05, "loss": 0.3438, "step": 2656 }, { "epoch": 0.8578231567052746, "grad_norm": 0.4052356884412845, "learning_rate": 1.2830601703904335e-05, "loss": 0.3481, "step": 2657 }, { "epoch": 0.8581460107348965, "grad_norm": 0.4470810610156402, "learning_rate": 1.2828723437513064e-05, "loss": 0.3006, "step": 2658 }, { "epoch": 0.8584688647645183, "grad_norm": 0.6007945226474508, "learning_rate": 1.2826844495979872e-05, "loss": 0.2985, "step": 2659 }, { "epoch": 0.8587917187941402, "grad_norm": 0.49374811454989437, "learning_rate": 1.2824964879542819e-05, "loss": 0.3874, "step": 2660 }, { "epoch": 0.859114572823762, "grad_norm": 0.3841114365602885, "learning_rate": 1.282308458844005e-05, "loss": 0.3668, "step": 2661 }, { "epoch": 0.8594374268533839, "grad_norm": 0.4204406099415518, "learning_rate": 1.2821203622909791e-05, "loss": 0.3783, "step": 2662 }, { "epoch": 0.8597602808830058, "grad_norm": 0.4373073677105492, "learning_rate": 1.2819321983190365e-05, "loss": 0.3587, "step": 2663 }, { "epoch": 0.8600831349126277, "grad_norm": 0.41820517937165425, "learning_rate": 1.2817439669520166e-05, "loss": 0.3444, "step": 2664 }, { "epoch": 0.8604059889422495, "grad_norm": 0.38510194806842796, "learning_rate": 1.2815556682137682e-05, "loss": 0.3581, "step": 2665 }, { "epoch": 0.8607288429718714, "grad_norm": 0.5119701768279341, "learning_rate": 1.2813673021281486e-05, "loss": 0.3247, "step": 2666 }, { "epoch": 0.8610516970014932, "grad_norm": 0.4980868112790105, "learning_rate": 1.2811788687190238e-05, "loss": 0.3739, "step": 2667 }, { "epoch": 0.861374551031115, "grad_norm": 0.3622309217548434, "learning_rate": 1.2809903680102678e-05, "loss": 0.3072, "step": 2668 }, { "epoch": 0.8616974050607369, "grad_norm": 0.4227441745054033, "learning_rate": 1.2808018000257634e-05, "loss": 0.3116, "step": 2669 }, { "epoch": 0.8620202590903587, "grad_norm": 0.4062028698608029, "learning_rate": 1.2806131647894018e-05, "loss": 0.3885, "step": 2670 }, { "epoch": 0.8623431131199806, "grad_norm": 0.5259061104579338, "learning_rate": 1.2804244623250831e-05, "loss": 0.4107, "step": 2671 }, { "epoch": 0.8626659671496025, "grad_norm": 0.5000251093139165, "learning_rate": 1.2802356926567152e-05, "loss": 0.4152, "step": 2672 }, { "epoch": 0.8629888211792244, "grad_norm": 0.4213625005337561, "learning_rate": 1.2800468558082153e-05, "loss": 0.358, "step": 2673 }, { "epoch": 0.8633116752088462, "grad_norm": 1.3759575924706953, "learning_rate": 1.279857951803509e-05, "loss": 0.341, "step": 2674 }, { "epoch": 0.8636345292384681, "grad_norm": 0.40086895203901984, "learning_rate": 1.2796689806665296e-05, "loss": 0.3475, "step": 2675 }, { "epoch": 0.8639573832680899, "grad_norm": 0.46903343875664594, "learning_rate": 1.2794799424212196e-05, "loss": 0.3868, "step": 2676 }, { "epoch": 0.8642802372977118, "grad_norm": 0.33292058786575845, "learning_rate": 1.2792908370915302e-05, "loss": 0.4022, "step": 2677 }, { "epoch": 0.8646030913273336, "grad_norm": 0.4270632731666142, "learning_rate": 1.2791016647014205e-05, "loss": 0.3285, "step": 2678 }, { "epoch": 0.8649259453569554, "grad_norm": 0.6209843711261309, "learning_rate": 1.2789124252748587e-05, "loss": 0.3069, "step": 2679 }, { "epoch": 0.8652487993865774, "grad_norm": 0.42605411137766214, "learning_rate": 1.2787231188358208e-05, "loss": 0.381, "step": 2680 }, { "epoch": 0.8655716534161992, "grad_norm": 0.4858752271845869, "learning_rate": 1.278533745408292e-05, "loss": 0.3818, "step": 2681 }, { "epoch": 0.8658945074458211, "grad_norm": 0.39049250116437145, "learning_rate": 1.2783443050162654e-05, "loss": 0.3157, "step": 2682 }, { "epoch": 0.8662173614754429, "grad_norm": 0.3881160082292323, "learning_rate": 1.2781547976837427e-05, "loss": 0.4387, "step": 2683 }, { "epoch": 0.8665402155050648, "grad_norm": 0.3940452052927177, "learning_rate": 1.277965223434735e-05, "loss": 0.3854, "step": 2684 }, { "epoch": 0.8668630695346866, "grad_norm": 0.4102895399669038, "learning_rate": 1.27777558229326e-05, "loss": 0.3328, "step": 2685 }, { "epoch": 0.8671859235643085, "grad_norm": 0.4764315900004233, "learning_rate": 1.2775858742833459e-05, "loss": 0.4254, "step": 2686 }, { "epoch": 0.8675087775939303, "grad_norm": 0.4691951749187074, "learning_rate": 1.2773960994290276e-05, "loss": 0.4009, "step": 2687 }, { "epoch": 0.8678316316235521, "grad_norm": 0.3491104356575305, "learning_rate": 1.2772062577543503e-05, "loss": 0.3392, "step": 2688 }, { "epoch": 0.8681544856531741, "grad_norm": 0.42930374812439065, "learning_rate": 1.2770163492833658e-05, "loss": 0.3639, "step": 2689 }, { "epoch": 0.868477339682796, "grad_norm": 0.4930547183530461, "learning_rate": 1.2768263740401357e-05, "loss": 0.4097, "step": 2690 }, { "epoch": 0.8688001937124178, "grad_norm": 0.46181046914396645, "learning_rate": 1.2766363320487295e-05, "loss": 0.3419, "step": 2691 }, { "epoch": 0.8691230477420396, "grad_norm": 0.3917044121469092, "learning_rate": 1.2764462233332251e-05, "loss": 0.2941, "step": 2692 }, { "epoch": 0.8694459017716615, "grad_norm": 0.4312735704155393, "learning_rate": 1.2762560479177093e-05, "loss": 0.3783, "step": 2693 }, { "epoch": 0.8697687558012833, "grad_norm": 0.3519686072708877, "learning_rate": 1.276065805826277e-05, "loss": 0.3513, "step": 2694 }, { "epoch": 0.8700916098309052, "grad_norm": 0.40332610175270356, "learning_rate": 1.2758754970830316e-05, "loss": 0.4039, "step": 2695 }, { "epoch": 0.870414463860527, "grad_norm": 0.34732416319080217, "learning_rate": 1.2756851217120847e-05, "loss": 0.3082, "step": 2696 }, { "epoch": 0.870737317890149, "grad_norm": 0.39539468797269534, "learning_rate": 1.275494679737557e-05, "loss": 0.4081, "step": 2697 }, { "epoch": 0.8710601719197708, "grad_norm": 0.39150825083603225, "learning_rate": 1.2753041711835767e-05, "loss": 0.3324, "step": 2698 }, { "epoch": 0.8713830259493927, "grad_norm": 0.5176130454718652, "learning_rate": 1.2751135960742815e-05, "loss": 0.4034, "step": 2699 }, { "epoch": 0.8717058799790145, "grad_norm": 0.5659468975929106, "learning_rate": 1.2749229544338167e-05, "loss": 0.3058, "step": 2700 }, { "epoch": 0.8720287340086363, "grad_norm": 0.597689186001696, "learning_rate": 1.2747322462863366e-05, "loss": 0.3494, "step": 2701 }, { "epoch": 0.8723515880382582, "grad_norm": 0.3799036038594716, "learning_rate": 1.2745414716560034e-05, "loss": 0.3243, "step": 2702 }, { "epoch": 0.87267444206788, "grad_norm": 0.5075764243399058, "learning_rate": 1.274350630566988e-05, "loss": 0.3369, "step": 2703 }, { "epoch": 0.8729972960975019, "grad_norm": 0.5304723116222487, "learning_rate": 1.27415972304347e-05, "loss": 0.337, "step": 2704 }, { "epoch": 0.8733201501271238, "grad_norm": 0.5258587894683602, "learning_rate": 1.2739687491096368e-05, "loss": 0.3359, "step": 2705 }, { "epoch": 0.8736430041567457, "grad_norm": 0.42086153772580087, "learning_rate": 1.2737777087896845e-05, "loss": 0.3919, "step": 2706 }, { "epoch": 0.8739658581863675, "grad_norm": 0.463676283409098, "learning_rate": 1.273586602107818e-05, "loss": 0.3198, "step": 2707 }, { "epoch": 0.8742887122159894, "grad_norm": 0.35068250108882343, "learning_rate": 1.2733954290882498e-05, "loss": 0.3054, "step": 2708 }, { "epoch": 0.8746115662456112, "grad_norm": 0.3554637249254715, "learning_rate": 1.2732041897552015e-05, "loss": 0.3275, "step": 2709 }, { "epoch": 0.874934420275233, "grad_norm": 0.40413907714769187, "learning_rate": 1.273012884132903e-05, "loss": 0.3814, "step": 2710 }, { "epoch": 0.8752572743048549, "grad_norm": 0.41048510986363795, "learning_rate": 1.2728215122455923e-05, "loss": 0.3613, "step": 2711 }, { "epoch": 0.8755801283344767, "grad_norm": 0.47352275283950146, "learning_rate": 1.272630074117516e-05, "loss": 0.3427, "step": 2712 }, { "epoch": 0.8759029823640986, "grad_norm": 0.3922152162220124, "learning_rate": 1.272438569772929e-05, "loss": 0.3809, "step": 2713 }, { "epoch": 0.8762258363937205, "grad_norm": 0.43832753745664876, "learning_rate": 1.2722469992360944e-05, "loss": 0.3775, "step": 2714 }, { "epoch": 0.8765486904233424, "grad_norm": 0.4037229528806279, "learning_rate": 1.2720553625312845e-05, "loss": 0.3411, "step": 2715 }, { "epoch": 0.8768715444529642, "grad_norm": 0.4257648789876882, "learning_rate": 1.2718636596827789e-05, "loss": 0.2988, "step": 2716 }, { "epoch": 0.8771943984825861, "grad_norm": 0.48683763429570187, "learning_rate": 1.2716718907148663e-05, "loss": 0.3272, "step": 2717 }, { "epoch": 0.8775172525122079, "grad_norm": 0.4363772490757629, "learning_rate": 1.2714800556518436e-05, "loss": 0.3944, "step": 2718 }, { "epoch": 0.8778401065418298, "grad_norm": 0.5147307953027541, "learning_rate": 1.2712881545180159e-05, "loss": 0.3569, "step": 2719 }, { "epoch": 0.8781629605714516, "grad_norm": 0.4967052568019653, "learning_rate": 1.271096187337697e-05, "loss": 0.3179, "step": 2720 }, { "epoch": 0.8784858146010734, "grad_norm": 0.7130252927180023, "learning_rate": 1.2709041541352086e-05, "loss": 0.369, "step": 2721 }, { "epoch": 0.8788086686306954, "grad_norm": 0.42423318925076836, "learning_rate": 1.2707120549348812e-05, "loss": 0.4123, "step": 2722 }, { "epoch": 0.8791315226603172, "grad_norm": 0.3604859215451339, "learning_rate": 1.2705198897610536e-05, "loss": 0.3726, "step": 2723 }, { "epoch": 0.8794543766899391, "grad_norm": 0.37873381166497616, "learning_rate": 1.2703276586380728e-05, "loss": 0.3507, "step": 2724 }, { "epoch": 0.8797772307195609, "grad_norm": 0.4572960386446172, "learning_rate": 1.2701353615902941e-05, "loss": 0.3646, "step": 2725 }, { "epoch": 0.8801000847491828, "grad_norm": 0.7459229012921705, "learning_rate": 1.2699429986420813e-05, "loss": 0.3923, "step": 2726 }, { "epoch": 0.8804229387788046, "grad_norm": 0.37954643812510164, "learning_rate": 1.2697505698178068e-05, "loss": 0.3343, "step": 2727 }, { "epoch": 0.8807457928084265, "grad_norm": 0.4185139365151412, "learning_rate": 1.2695580751418506e-05, "loss": 0.3551, "step": 2728 }, { "epoch": 0.8810686468380483, "grad_norm": 0.3599314129397154, "learning_rate": 1.2693655146386018e-05, "loss": 0.3059, "step": 2729 }, { "epoch": 0.8813915008676702, "grad_norm": 0.4013416787771874, "learning_rate": 1.2691728883324574e-05, "loss": 0.3047, "step": 2730 }, { "epoch": 0.8817143548972921, "grad_norm": 0.3716843355811413, "learning_rate": 1.2689801962478233e-05, "loss": 0.3667, "step": 2731 }, { "epoch": 0.882037208926914, "grad_norm": 0.6051571634154027, "learning_rate": 1.2687874384091128e-05, "loss": 0.363, "step": 2732 }, { "epoch": 0.8823600629565358, "grad_norm": 0.46160460096369227, "learning_rate": 1.2685946148407483e-05, "loss": 0.3383, "step": 2733 }, { "epoch": 0.8826829169861576, "grad_norm": 0.3459395248680566, "learning_rate": 1.2684017255671604e-05, "loss": 0.404, "step": 2734 }, { "epoch": 0.8830057710157795, "grad_norm": 0.38438163348619947, "learning_rate": 1.2682087706127876e-05, "loss": 0.3656, "step": 2735 }, { "epoch": 0.8833286250454013, "grad_norm": 0.5212951873683765, "learning_rate": 1.268015750002077e-05, "loss": 0.3497, "step": 2736 }, { "epoch": 0.8836514790750232, "grad_norm": 0.34350212790409085, "learning_rate": 1.2678226637594845e-05, "loss": 0.3245, "step": 2737 }, { "epoch": 0.883974333104645, "grad_norm": 0.4018584751940867, "learning_rate": 1.2676295119094735e-05, "loss": 0.3696, "step": 2738 }, { "epoch": 0.884297187134267, "grad_norm": 0.421954378797634, "learning_rate": 1.2674362944765162e-05, "loss": 0.3705, "step": 2739 }, { "epoch": 0.8846200411638888, "grad_norm": 0.3549587602103855, "learning_rate": 1.267243011485093e-05, "loss": 0.3415, "step": 2740 }, { "epoch": 0.8849428951935107, "grad_norm": 0.368208001425988, "learning_rate": 1.2670496629596924e-05, "loss": 0.3921, "step": 2741 }, { "epoch": 0.8852657492231325, "grad_norm": 0.5208397489652361, "learning_rate": 1.2668562489248116e-05, "loss": 0.3169, "step": 2742 }, { "epoch": 0.8855886032527543, "grad_norm": 0.4051278272369945, "learning_rate": 1.2666627694049558e-05, "loss": 0.398, "step": 2743 }, { "epoch": 0.8859114572823762, "grad_norm": 0.4808385592398251, "learning_rate": 1.2664692244246386e-05, "loss": 0.4117, "step": 2744 }, { "epoch": 0.886234311311998, "grad_norm": 0.4273375397020591, "learning_rate": 1.2662756140083818e-05, "loss": 0.3928, "step": 2745 }, { "epoch": 0.8865571653416199, "grad_norm": 0.39815590110342636, "learning_rate": 1.2660819381807156e-05, "loss": 0.3218, "step": 2746 }, { "epoch": 0.8868800193712417, "grad_norm": 0.5035800569719554, "learning_rate": 1.2658881969661788e-05, "loss": 0.4051, "step": 2747 }, { "epoch": 0.8872028734008637, "grad_norm": 0.44036589481661814, "learning_rate": 1.2656943903893178e-05, "loss": 0.3779, "step": 2748 }, { "epoch": 0.8875257274304855, "grad_norm": 0.3854086138153916, "learning_rate": 1.2655005184746873e-05, "loss": 0.3313, "step": 2749 }, { "epoch": 0.8878485814601074, "grad_norm": 0.38141184064154127, "learning_rate": 1.2653065812468514e-05, "loss": 0.3507, "step": 2750 }, { "epoch": 0.8881714354897292, "grad_norm": 0.3926842267260824, "learning_rate": 1.2651125787303812e-05, "loss": 0.301, "step": 2751 }, { "epoch": 0.888494289519351, "grad_norm": 0.3768221289804095, "learning_rate": 1.2649185109498566e-05, "loss": 0.3012, "step": 2752 }, { "epoch": 0.8888171435489729, "grad_norm": 0.5068259135962208, "learning_rate": 1.2647243779298657e-05, "loss": 0.3552, "step": 2753 }, { "epoch": 0.8891399975785947, "grad_norm": 0.48063459441066647, "learning_rate": 1.264530179695005e-05, "loss": 0.3594, "step": 2754 }, { "epoch": 0.8894628516082166, "grad_norm": 0.616307938451342, "learning_rate": 1.2643359162698793e-05, "loss": 0.3275, "step": 2755 }, { "epoch": 0.8897857056378385, "grad_norm": 0.36367879161502153, "learning_rate": 1.2641415876791009e-05, "loss": 0.386, "step": 2756 }, { "epoch": 0.8901085596674604, "grad_norm": 0.4113879310268718, "learning_rate": 1.2639471939472914e-05, "loss": 0.4218, "step": 2757 }, { "epoch": 0.8904314136970822, "grad_norm": 0.4929199832188705, "learning_rate": 1.2637527350990805e-05, "loss": 0.2924, "step": 2758 }, { "epoch": 0.8907542677267041, "grad_norm": 0.7054819727434327, "learning_rate": 1.2635582111591053e-05, "loss": 0.3211, "step": 2759 }, { "epoch": 0.8910771217563259, "grad_norm": 0.3686037875671813, "learning_rate": 1.2633636221520118e-05, "loss": 0.349, "step": 2760 }, { "epoch": 0.8913999757859478, "grad_norm": 0.39628669752439244, "learning_rate": 1.2631689681024547e-05, "loss": 0.3764, "step": 2761 }, { "epoch": 0.8917228298155696, "grad_norm": 0.5688482491852523, "learning_rate": 1.2629742490350958e-05, "loss": 0.3396, "step": 2762 }, { "epoch": 0.8920456838451915, "grad_norm": 0.47797382978651537, "learning_rate": 1.262779464974606e-05, "loss": 0.3399, "step": 2763 }, { "epoch": 0.8923685378748133, "grad_norm": 0.4263437988070165, "learning_rate": 1.262584615945664e-05, "loss": 0.3357, "step": 2764 }, { "epoch": 0.8926913919044352, "grad_norm": 0.4568602995395829, "learning_rate": 1.262389701972957e-05, "loss": 0.3147, "step": 2765 }, { "epoch": 0.8930142459340571, "grad_norm": 0.4407555112703878, "learning_rate": 1.2621947230811806e-05, "loss": 0.3157, "step": 2766 }, { "epoch": 0.8933370999636789, "grad_norm": 0.4114394739735186, "learning_rate": 1.261999679295038e-05, "loss": 0.3527, "step": 2767 }, { "epoch": 0.8936599539933008, "grad_norm": 0.4995792379770341, "learning_rate": 1.2618045706392411e-05, "loss": 0.3883, "step": 2768 }, { "epoch": 0.8939828080229226, "grad_norm": 0.46546193064147307, "learning_rate": 1.26160939713851e-05, "loss": 0.3358, "step": 2769 }, { "epoch": 0.8943056620525445, "grad_norm": 0.4147427184539853, "learning_rate": 1.2614141588175727e-05, "loss": 0.3577, "step": 2770 }, { "epoch": 0.8946285160821663, "grad_norm": 0.4635084893376462, "learning_rate": 1.2612188557011658e-05, "loss": 0.3189, "step": 2771 }, { "epoch": 0.8949513701117882, "grad_norm": 0.4222298632443165, "learning_rate": 1.2610234878140337e-05, "loss": 0.332, "step": 2772 }, { "epoch": 0.8952742241414101, "grad_norm": 0.5176553270307813, "learning_rate": 1.2608280551809298e-05, "loss": 0.3362, "step": 2773 }, { "epoch": 0.895597078171032, "grad_norm": 0.43180620706012446, "learning_rate": 1.2606325578266144e-05, "loss": 0.3472, "step": 2774 }, { "epoch": 0.8959199322006538, "grad_norm": 0.958851879854104, "learning_rate": 1.2604369957758572e-05, "loss": 0.3222, "step": 2775 }, { "epoch": 0.8962427862302756, "grad_norm": 0.429868434745617, "learning_rate": 1.2602413690534356e-05, "loss": 0.3574, "step": 2776 }, { "epoch": 0.8965656402598975, "grad_norm": 0.6959848884021942, "learning_rate": 1.2600456776841353e-05, "loss": 0.3979, "step": 2777 }, { "epoch": 0.8968884942895193, "grad_norm": 0.3924649373871996, "learning_rate": 1.2598499216927498e-05, "loss": 0.3409, "step": 2778 }, { "epoch": 0.8972113483191412, "grad_norm": 0.5113875164891383, "learning_rate": 1.2596541011040811e-05, "loss": 0.3089, "step": 2779 }, { "epoch": 0.897534202348763, "grad_norm": 0.45291147626847095, "learning_rate": 1.2594582159429402e-05, "loss": 0.3864, "step": 2780 }, { "epoch": 0.897857056378385, "grad_norm": 0.4225241222827599, "learning_rate": 1.2592622662341447e-05, "loss": 0.3918, "step": 2781 }, { "epoch": 0.8981799104080068, "grad_norm": 0.6007077683718187, "learning_rate": 1.2590662520025211e-05, "loss": 0.3478, "step": 2782 }, { "epoch": 0.8985027644376287, "grad_norm": 0.40647289317157337, "learning_rate": 1.2588701732729046e-05, "loss": 0.331, "step": 2783 }, { "epoch": 0.8988256184672505, "grad_norm": 0.3752975358953948, "learning_rate": 1.2586740300701377e-05, "loss": 0.3292, "step": 2784 }, { "epoch": 0.8991484724968724, "grad_norm": 0.3973314519967439, "learning_rate": 1.2584778224190719e-05, "loss": 0.3404, "step": 2785 }, { "epoch": 0.8994713265264942, "grad_norm": 0.43754669725177014, "learning_rate": 1.2582815503445662e-05, "loss": 0.3741, "step": 2786 }, { "epoch": 0.899794180556116, "grad_norm": 0.3479888869052701, "learning_rate": 1.2580852138714881e-05, "loss": 0.3347, "step": 2787 }, { "epoch": 0.9001170345857379, "grad_norm": 0.5586359544128289, "learning_rate": 1.257888813024713e-05, "loss": 0.3575, "step": 2788 }, { "epoch": 0.9004398886153597, "grad_norm": 0.487137295050233, "learning_rate": 1.2576923478291246e-05, "loss": 0.3102, "step": 2789 }, { "epoch": 0.9007627426449817, "grad_norm": 0.4035926511358478, "learning_rate": 1.257495818309615e-05, "loss": 0.3696, "step": 2790 }, { "epoch": 0.9010855966746035, "grad_norm": 0.45005540246385095, "learning_rate": 1.257299224491084e-05, "loss": 0.3226, "step": 2791 }, { "epoch": 0.9014084507042254, "grad_norm": 0.4273932203336058, "learning_rate": 1.25710256639844e-05, "loss": 0.4087, "step": 2792 }, { "epoch": 0.9017313047338472, "grad_norm": 0.4222904416342333, "learning_rate": 1.2569058440565988e-05, "loss": 0.2932, "step": 2793 }, { "epoch": 0.9020541587634691, "grad_norm": 0.3374159612286641, "learning_rate": 1.2567090574904854e-05, "loss": 0.2964, "step": 2794 }, { "epoch": 0.9023770127930909, "grad_norm": 0.4385299721552262, "learning_rate": 1.2565122067250323e-05, "loss": 0.3642, "step": 2795 }, { "epoch": 0.9026998668227127, "grad_norm": 0.37897288879198565, "learning_rate": 1.25631529178518e-05, "loss": 0.3415, "step": 2796 }, { "epoch": 0.9030227208523346, "grad_norm": 0.5037259147694146, "learning_rate": 1.2561183126958776e-05, "loss": 0.3729, "step": 2797 }, { "epoch": 0.9033455748819565, "grad_norm": 0.4128423611271215, "learning_rate": 1.2559212694820818e-05, "loss": 0.3369, "step": 2798 }, { "epoch": 0.9036684289115784, "grad_norm": 0.3818863695594365, "learning_rate": 1.2557241621687581e-05, "loss": 0.3417, "step": 2799 }, { "epoch": 0.9039912829412002, "grad_norm": 0.4807061123371679, "learning_rate": 1.2555269907808791e-05, "loss": 0.405, "step": 2800 }, { "epoch": 0.9043141369708221, "grad_norm": 0.44387347156410795, "learning_rate": 1.2553297553434268e-05, "loss": 0.3972, "step": 2801 }, { "epoch": 0.9046369910004439, "grad_norm": 0.5148683959361328, "learning_rate": 1.2551324558813902e-05, "loss": 0.43, "step": 2802 }, { "epoch": 0.9049598450300658, "grad_norm": 0.40504138043059773, "learning_rate": 1.2549350924197671e-05, "loss": 0.3432, "step": 2803 }, { "epoch": 0.9052826990596876, "grad_norm": 0.36411342264002466, "learning_rate": 1.2547376649835631e-05, "loss": 0.2971, "step": 2804 }, { "epoch": 0.9056055530893095, "grad_norm": 0.3564749346065, "learning_rate": 1.2545401735977922e-05, "loss": 0.3164, "step": 2805 }, { "epoch": 0.9059284071189313, "grad_norm": 0.400512285157782, "learning_rate": 1.254342618287476e-05, "loss": 0.3052, "step": 2806 }, { "epoch": 0.9062512611485533, "grad_norm": 0.4400678110502638, "learning_rate": 1.2541449990776446e-05, "loss": 0.3685, "step": 2807 }, { "epoch": 0.9065741151781751, "grad_norm": 0.4816162612522582, "learning_rate": 1.253947315993336e-05, "loss": 0.317, "step": 2808 }, { "epoch": 0.9068969692077969, "grad_norm": 0.41856737451179826, "learning_rate": 1.2537495690595965e-05, "loss": 0.3867, "step": 2809 }, { "epoch": 0.9072198232374188, "grad_norm": 0.3508648601059889, "learning_rate": 1.2535517583014802e-05, "loss": 0.3414, "step": 2810 }, { "epoch": 0.9075426772670406, "grad_norm": 0.388259566920064, "learning_rate": 1.2533538837440496e-05, "loss": 0.3977, "step": 2811 }, { "epoch": 0.9078655312966625, "grad_norm": 0.4611973775311184, "learning_rate": 1.253155945412375e-05, "loss": 0.3748, "step": 2812 }, { "epoch": 0.9081883853262843, "grad_norm": 0.3722640399121417, "learning_rate": 1.2529579433315351e-05, "loss": 0.3293, "step": 2813 }, { "epoch": 0.9085112393559062, "grad_norm": 0.3424761448327958, "learning_rate": 1.2527598775266166e-05, "loss": 0.3485, "step": 2814 }, { "epoch": 0.9088340933855281, "grad_norm": 0.3978640050142474, "learning_rate": 1.2525617480227136e-05, "loss": 0.2848, "step": 2815 }, { "epoch": 0.90915694741515, "grad_norm": 0.4433539730599383, "learning_rate": 1.2523635548449296e-05, "loss": 0.3607, "step": 2816 }, { "epoch": 0.9094798014447718, "grad_norm": 0.5094109848113421, "learning_rate": 1.2521652980183748e-05, "loss": 0.3102, "step": 2817 }, { "epoch": 0.9098026554743937, "grad_norm": 0.42399939130235065, "learning_rate": 1.251966977568168e-05, "loss": 0.328, "step": 2818 }, { "epoch": 0.9101255095040155, "grad_norm": 0.38862194858179383, "learning_rate": 1.2517685935194367e-05, "loss": 0.3321, "step": 2819 }, { "epoch": 0.9104483635336373, "grad_norm": 0.3990012848845049, "learning_rate": 1.2515701458973155e-05, "loss": 0.3486, "step": 2820 }, { "epoch": 0.9107712175632592, "grad_norm": 0.5557237310462728, "learning_rate": 1.2513716347269474e-05, "loss": 0.362, "step": 2821 }, { "epoch": 0.911094071592881, "grad_norm": 0.47415472626349037, "learning_rate": 1.2511730600334838e-05, "loss": 0.3468, "step": 2822 }, { "epoch": 0.9114169256225029, "grad_norm": 0.4495462392616143, "learning_rate": 1.2509744218420834e-05, "loss": 0.3848, "step": 2823 }, { "epoch": 0.9117397796521248, "grad_norm": 0.39035940573374145, "learning_rate": 1.250775720177914e-05, "loss": 0.3358, "step": 2824 }, { "epoch": 0.9120626336817467, "grad_norm": 0.4247674679412692, "learning_rate": 1.2505769550661501e-05, "loss": 0.3408, "step": 2825 }, { "epoch": 0.9123854877113685, "grad_norm": 0.43895873840972377, "learning_rate": 1.2503781265319757e-05, "loss": 0.3561, "step": 2826 }, { "epoch": 0.9127083417409904, "grad_norm": 0.39515510412098254, "learning_rate": 1.2501792346005814e-05, "loss": 0.3473, "step": 2827 }, { "epoch": 0.9130311957706122, "grad_norm": 0.6001453860380528, "learning_rate": 1.249980279297167e-05, "loss": 0.3453, "step": 2828 }, { "epoch": 0.913354049800234, "grad_norm": 0.3881252531724565, "learning_rate": 1.2497812606469397e-05, "loss": 0.3975, "step": 2829 }, { "epoch": 0.9136769038298559, "grad_norm": 0.3821198697632926, "learning_rate": 1.249582178675115e-05, "loss": 0.3675, "step": 2830 }, { "epoch": 0.9139997578594777, "grad_norm": 0.5014436172298925, "learning_rate": 1.2493830334069162e-05, "loss": 0.3737, "step": 2831 }, { "epoch": 0.9143226118890997, "grad_norm": 0.6539577864834588, "learning_rate": 1.2491838248675748e-05, "loss": 0.4252, "step": 2832 }, { "epoch": 0.9146454659187215, "grad_norm": 0.45142324774575227, "learning_rate": 1.24898455308233e-05, "loss": 0.3224, "step": 2833 }, { "epoch": 0.9149683199483434, "grad_norm": 0.4095604943870871, "learning_rate": 1.2487852180764295e-05, "loss": 0.3608, "step": 2834 }, { "epoch": 0.9152911739779652, "grad_norm": 0.37793658309872347, "learning_rate": 1.2485858198751289e-05, "loss": 0.3971, "step": 2835 }, { "epoch": 0.9156140280075871, "grad_norm": 0.4266832647662789, "learning_rate": 1.2483863585036915e-05, "loss": 0.3635, "step": 2836 }, { "epoch": 0.9159368820372089, "grad_norm": 0.40057605967463533, "learning_rate": 1.2481868339873887e-05, "loss": 0.3467, "step": 2837 }, { "epoch": 0.9162597360668308, "grad_norm": 0.38853506736434906, "learning_rate": 1.2479872463515001e-05, "loss": 0.3582, "step": 2838 }, { "epoch": 0.9165825900964526, "grad_norm": 0.3652476431829238, "learning_rate": 1.247787595621313e-05, "loss": 0.3881, "step": 2839 }, { "epoch": 0.9169054441260746, "grad_norm": 0.4285798729132094, "learning_rate": 1.2475878818221233e-05, "loss": 0.3251, "step": 2840 }, { "epoch": 0.9172282981556964, "grad_norm": 0.4549328253181473, "learning_rate": 1.2473881049792341e-05, "loss": 0.3238, "step": 2841 }, { "epoch": 0.9175511521853182, "grad_norm": 0.4579363326664821, "learning_rate": 1.247188265117957e-05, "loss": 0.4064, "step": 2842 }, { "epoch": 0.9178740062149401, "grad_norm": 0.38390984544954393, "learning_rate": 1.2469883622636114e-05, "loss": 0.3351, "step": 2843 }, { "epoch": 0.9181968602445619, "grad_norm": 0.386297912273229, "learning_rate": 1.2467883964415246e-05, "loss": 0.3512, "step": 2844 }, { "epoch": 0.9185197142741838, "grad_norm": 0.38158101600209965, "learning_rate": 1.2465883676770321e-05, "loss": 0.3369, "step": 2845 }, { "epoch": 0.9188425683038056, "grad_norm": 0.33315438588190904, "learning_rate": 1.246388275995477e-05, "loss": 0.3223, "step": 2846 }, { "epoch": 0.9191654223334275, "grad_norm": 0.4023528599172178, "learning_rate": 1.2461881214222115e-05, "loss": 0.3879, "step": 2847 }, { "epoch": 0.9194882763630493, "grad_norm": 0.44843133373288296, "learning_rate": 1.2459879039825939e-05, "loss": 0.3648, "step": 2848 }, { "epoch": 0.9198111303926713, "grad_norm": 0.3965228649925411, "learning_rate": 1.245787623701992e-05, "loss": 0.3626, "step": 2849 }, { "epoch": 0.9201339844222931, "grad_norm": 0.5035427690924409, "learning_rate": 1.2455872806057809e-05, "loss": 0.4069, "step": 2850 }, { "epoch": 0.920456838451915, "grad_norm": 0.5075579311017893, "learning_rate": 1.2453868747193438e-05, "loss": 0.3294, "step": 2851 }, { "epoch": 0.9207796924815368, "grad_norm": 0.4284566562646836, "learning_rate": 1.2451864060680717e-05, "loss": 0.409, "step": 2852 }, { "epoch": 0.9211025465111586, "grad_norm": 0.3852586185013833, "learning_rate": 1.2449858746773641e-05, "loss": 0.3796, "step": 2853 }, { "epoch": 0.9214254005407805, "grad_norm": 0.37218536769928146, "learning_rate": 1.2447852805726276e-05, "loss": 0.3074, "step": 2854 }, { "epoch": 0.9217482545704023, "grad_norm": 0.4305922309962267, "learning_rate": 1.2445846237792776e-05, "loss": 0.3319, "step": 2855 }, { "epoch": 0.9220711086000242, "grad_norm": 0.36092490368143737, "learning_rate": 1.2443839043227369e-05, "loss": 0.3734, "step": 2856 }, { "epoch": 0.9223939626296461, "grad_norm": 0.396595430692966, "learning_rate": 1.244183122228436e-05, "loss": 0.3201, "step": 2857 }, { "epoch": 0.922716816659268, "grad_norm": 0.438917373879066, "learning_rate": 1.2439822775218142e-05, "loss": 0.3973, "step": 2858 }, { "epoch": 0.9230396706888898, "grad_norm": 0.4047728753675887, "learning_rate": 1.2437813702283182e-05, "loss": 0.3391, "step": 2859 }, { "epoch": 0.9233625247185117, "grad_norm": 0.466477145404804, "learning_rate": 1.2435804003734026e-05, "loss": 0.3666, "step": 2860 }, { "epoch": 0.9236853787481335, "grad_norm": 0.3836107746087361, "learning_rate": 1.2433793679825298e-05, "loss": 0.2821, "step": 2861 }, { "epoch": 0.9240082327777553, "grad_norm": 0.428450773860967, "learning_rate": 1.2431782730811707e-05, "loss": 0.3192, "step": 2862 }, { "epoch": 0.9243310868073772, "grad_norm": 0.4284176057673311, "learning_rate": 1.2429771156948034e-05, "loss": 0.355, "step": 2863 }, { "epoch": 0.924653940836999, "grad_norm": 0.43110812662425374, "learning_rate": 1.2427758958489144e-05, "loss": 0.3676, "step": 2864 }, { "epoch": 0.9249767948666209, "grad_norm": 0.3683531297722991, "learning_rate": 1.242574613568998e-05, "loss": 0.34, "step": 2865 }, { "epoch": 0.9252996488962428, "grad_norm": 0.5049985849720342, "learning_rate": 1.2423732688805565e-05, "loss": 0.3033, "step": 2866 }, { "epoch": 0.9256225029258647, "grad_norm": 0.43035410788068335, "learning_rate": 1.2421718618090998e-05, "loss": 0.3031, "step": 2867 }, { "epoch": 0.9259453569554865, "grad_norm": 0.4736644347016428, "learning_rate": 1.241970392380146e-05, "loss": 0.3336, "step": 2868 }, { "epoch": 0.9262682109851084, "grad_norm": 0.3574621434656716, "learning_rate": 1.2417688606192209e-05, "loss": 0.2961, "step": 2869 }, { "epoch": 0.9265910650147302, "grad_norm": 0.41412088663040936, "learning_rate": 1.2415672665518585e-05, "loss": 0.3494, "step": 2870 }, { "epoch": 0.926913919044352, "grad_norm": 0.5158871530664304, "learning_rate": 1.2413656102036004e-05, "loss": 0.3391, "step": 2871 }, { "epoch": 0.9272367730739739, "grad_norm": 0.3571722741214048, "learning_rate": 1.241163891599996e-05, "loss": 0.3826, "step": 2872 }, { "epoch": 0.9275596271035957, "grad_norm": 0.4049662126942133, "learning_rate": 1.2409621107666027e-05, "loss": 0.3347, "step": 2873 }, { "epoch": 0.9278824811332177, "grad_norm": 0.38333723648033946, "learning_rate": 1.2407602677289864e-05, "loss": 0.3137, "step": 2874 }, { "epoch": 0.9282053351628395, "grad_norm": 0.4242088728802654, "learning_rate": 1.2405583625127195e-05, "loss": 0.3114, "step": 2875 }, { "epoch": 0.9285281891924614, "grad_norm": 0.4831106275673817, "learning_rate": 1.240356395143384e-05, "loss": 0.3614, "step": 2876 }, { "epoch": 0.9288510432220832, "grad_norm": 0.44054065120127717, "learning_rate": 1.2401543656465687e-05, "loss": 0.3666, "step": 2877 }, { "epoch": 0.9291738972517051, "grad_norm": 0.3958310060463739, "learning_rate": 1.2399522740478697e-05, "loss": 0.3404, "step": 2878 }, { "epoch": 0.9294967512813269, "grad_norm": 0.4044166791197965, "learning_rate": 1.2397501203728927e-05, "loss": 0.3014, "step": 2879 }, { "epoch": 0.9298196053109488, "grad_norm": 0.3709397769516438, "learning_rate": 1.2395479046472498e-05, "loss": 0.3322, "step": 2880 }, { "epoch": 0.9301424593405706, "grad_norm": 0.40914961526902355, "learning_rate": 1.2393456268965614e-05, "loss": 0.3883, "step": 2881 }, { "epoch": 0.9304653133701924, "grad_norm": 0.4500783854212478, "learning_rate": 1.2391432871464561e-05, "loss": 0.4149, "step": 2882 }, { "epoch": 0.9307881673998144, "grad_norm": 0.3938299964117058, "learning_rate": 1.23894088542257e-05, "loss": 0.3265, "step": 2883 }, { "epoch": 0.9311110214294362, "grad_norm": 0.5143601851593094, "learning_rate": 1.2387384217505468e-05, "loss": 0.393, "step": 2884 }, { "epoch": 0.9314338754590581, "grad_norm": 0.517266212173769, "learning_rate": 1.2385358961560386e-05, "loss": 0.3961, "step": 2885 }, { "epoch": 0.9317567294886799, "grad_norm": 0.47601614850214913, "learning_rate": 1.2383333086647053e-05, "loss": 0.3947, "step": 2886 }, { "epoch": 0.9320795835183018, "grad_norm": 0.46154056969463564, "learning_rate": 1.2381306593022141e-05, "loss": 0.3247, "step": 2887 }, { "epoch": 0.9324024375479236, "grad_norm": 0.386457692115784, "learning_rate": 1.2379279480942407e-05, "loss": 0.3457, "step": 2888 }, { "epoch": 0.9327252915775455, "grad_norm": 0.42838816391674833, "learning_rate": 1.2377251750664681e-05, "loss": 0.294, "step": 2889 }, { "epoch": 0.9330481456071673, "grad_norm": 0.36195308760256284, "learning_rate": 1.2375223402445874e-05, "loss": 0.3348, "step": 2890 }, { "epoch": 0.9333709996367893, "grad_norm": 0.4218421594159836, "learning_rate": 1.2373194436542975e-05, "loss": 0.2888, "step": 2891 }, { "epoch": 0.9336938536664111, "grad_norm": 0.446148953463246, "learning_rate": 1.2371164853213055e-05, "loss": 0.3866, "step": 2892 }, { "epoch": 0.934016707696033, "grad_norm": 0.37858722811341844, "learning_rate": 1.2369134652713253e-05, "loss": 0.3228, "step": 2893 }, { "epoch": 0.9343395617256548, "grad_norm": 0.3219591257831602, "learning_rate": 1.2367103835300794e-05, "loss": 0.3001, "step": 2894 }, { "epoch": 0.9346624157552766, "grad_norm": 0.431951472323822, "learning_rate": 1.2365072401232986e-05, "loss": 0.3842, "step": 2895 }, { "epoch": 0.9349852697848985, "grad_norm": 0.3949896323876552, "learning_rate": 1.2363040350767202e-05, "loss": 0.388, "step": 2896 }, { "epoch": 0.9353081238145203, "grad_norm": 0.5747250818579736, "learning_rate": 1.2361007684160902e-05, "loss": 0.3298, "step": 2897 }, { "epoch": 0.9356309778441422, "grad_norm": 0.36928064478100187, "learning_rate": 1.2358974401671625e-05, "loss": 0.3218, "step": 2898 }, { "epoch": 0.935953831873764, "grad_norm": 0.37737014459512386, "learning_rate": 1.2356940503556979e-05, "loss": 0.3414, "step": 2899 }, { "epoch": 0.936276685903386, "grad_norm": 0.37487635370076516, "learning_rate": 1.2354905990074662e-05, "loss": 0.3548, "step": 2900 }, { "epoch": 0.9365995399330078, "grad_norm": 0.5178632046514929, "learning_rate": 1.2352870861482441e-05, "loss": 0.3368, "step": 2901 }, { "epoch": 0.9369223939626297, "grad_norm": 0.5095935149751819, "learning_rate": 1.2350835118038166e-05, "loss": 0.2685, "step": 2902 }, { "epoch": 0.9372452479922515, "grad_norm": 0.41689586250046484, "learning_rate": 1.2348798759999759e-05, "loss": 0.3416, "step": 2903 }, { "epoch": 0.9375681020218734, "grad_norm": 0.3651111087554626, "learning_rate": 1.2346761787625228e-05, "loss": 0.3159, "step": 2904 }, { "epoch": 0.9378909560514952, "grad_norm": 0.4822767203507459, "learning_rate": 1.2344724201172652e-05, "loss": 0.2602, "step": 2905 }, { "epoch": 0.938213810081117, "grad_norm": 0.33155711611763017, "learning_rate": 1.2342686000900192e-05, "loss": 0.3317, "step": 2906 }, { "epoch": 0.9385366641107389, "grad_norm": 0.503823628559553, "learning_rate": 1.2340647187066085e-05, "loss": 0.3991, "step": 2907 }, { "epoch": 0.9388595181403608, "grad_norm": 0.4005317595138773, "learning_rate": 1.2338607759928644e-05, "loss": 0.3187, "step": 2908 }, { "epoch": 0.9391823721699827, "grad_norm": 0.39437952019733585, "learning_rate": 1.2336567719746263e-05, "loss": 0.3382, "step": 2909 }, { "epoch": 0.9395052261996045, "grad_norm": 0.38785067895431047, "learning_rate": 1.2334527066777414e-05, "loss": 0.3674, "step": 2910 }, { "epoch": 0.9398280802292264, "grad_norm": 0.32157318066567864, "learning_rate": 1.233248580128064e-05, "loss": 0.2813, "step": 2911 }, { "epoch": 0.9401509342588482, "grad_norm": 0.3782817277128121, "learning_rate": 1.2330443923514573e-05, "loss": 0.3414, "step": 2912 }, { "epoch": 0.9404737882884701, "grad_norm": 0.36557216551817634, "learning_rate": 1.232840143373791e-05, "loss": 0.2563, "step": 2913 }, { "epoch": 0.9407966423180919, "grad_norm": 0.36732603443776346, "learning_rate": 1.2326358332209437e-05, "loss": 0.3857, "step": 2914 }, { "epoch": 0.9411194963477137, "grad_norm": 0.3505069160400495, "learning_rate": 1.2324314619188008e-05, "loss": 0.3669, "step": 2915 }, { "epoch": 0.9414423503773357, "grad_norm": 0.4579166336768523, "learning_rate": 1.2322270294932561e-05, "loss": 0.3695, "step": 2916 }, { "epoch": 0.9417652044069575, "grad_norm": 0.4324827786268144, "learning_rate": 1.2320225359702107e-05, "loss": 0.3895, "step": 2917 }, { "epoch": 0.9420880584365794, "grad_norm": 0.38399612497846497, "learning_rate": 1.231817981375574e-05, "loss": 0.328, "step": 2918 }, { "epoch": 0.9424109124662012, "grad_norm": 0.4425227650443068, "learning_rate": 1.2316133657352625e-05, "loss": 0.2776, "step": 2919 }, { "epoch": 0.9427337664958231, "grad_norm": 0.48501055066758003, "learning_rate": 1.2314086890752007e-05, "loss": 0.3668, "step": 2920 }, { "epoch": 0.9430566205254449, "grad_norm": 0.44072957403349505, "learning_rate": 1.2312039514213208e-05, "loss": 0.3455, "step": 2921 }, { "epoch": 0.9433794745550668, "grad_norm": 0.4447208765933493, "learning_rate": 1.2309991527995632e-05, "loss": 0.3808, "step": 2922 }, { "epoch": 0.9437023285846886, "grad_norm": 0.5723477831058272, "learning_rate": 1.2307942932358751e-05, "loss": 0.3444, "step": 2923 }, { "epoch": 0.9440251826143105, "grad_norm": 0.43987853319523607, "learning_rate": 1.2305893727562122e-05, "loss": 0.3348, "step": 2924 }, { "epoch": 0.9443480366439324, "grad_norm": 0.3554928233573324, "learning_rate": 1.2303843913865376e-05, "loss": 0.3719, "step": 2925 }, { "epoch": 0.9446708906735543, "grad_norm": 0.4473433139583673, "learning_rate": 1.2301793491528222e-05, "loss": 0.368, "step": 2926 }, { "epoch": 0.9449937447031761, "grad_norm": 0.41659778146974985, "learning_rate": 1.2299742460810445e-05, "loss": 0.3077, "step": 2927 }, { "epoch": 0.9453165987327979, "grad_norm": 0.4032962095057728, "learning_rate": 1.2297690821971907e-05, "loss": 0.3242, "step": 2928 }, { "epoch": 0.9456394527624198, "grad_norm": 0.4702649701592017, "learning_rate": 1.2295638575272548e-05, "loss": 0.3951, "step": 2929 }, { "epoch": 0.9459623067920416, "grad_norm": 0.36452034490280044, "learning_rate": 1.2293585720972389e-05, "loss": 0.3148, "step": 2930 }, { "epoch": 0.9462851608216635, "grad_norm": 0.45608346354173235, "learning_rate": 1.2291532259331517e-05, "loss": 0.3136, "step": 2931 }, { "epoch": 0.9466080148512853, "grad_norm": 0.4284225102298138, "learning_rate": 1.2289478190610107e-05, "loss": 0.3752, "step": 2932 }, { "epoch": 0.9469308688809073, "grad_norm": 0.40326710620119016, "learning_rate": 1.2287423515068404e-05, "loss": 0.4196, "step": 2933 }, { "epoch": 0.9472537229105291, "grad_norm": 0.3868295253614959, "learning_rate": 1.2285368232966737e-05, "loss": 0.3819, "step": 2934 }, { "epoch": 0.947576576940151, "grad_norm": 0.4623721867396476, "learning_rate": 1.2283312344565504e-05, "loss": 0.3332, "step": 2935 }, { "epoch": 0.9478994309697728, "grad_norm": 0.48238826504222637, "learning_rate": 1.2281255850125182e-05, "loss": 0.3726, "step": 2936 }, { "epoch": 0.9482222849993946, "grad_norm": 0.42629636351753475, "learning_rate": 1.2279198749906331e-05, "loss": 0.3395, "step": 2937 }, { "epoch": 0.9485451390290165, "grad_norm": 0.4311362798348454, "learning_rate": 1.2277141044169579e-05, "loss": 0.3472, "step": 2938 }, { "epoch": 0.9488679930586383, "grad_norm": 0.45449522111032964, "learning_rate": 1.2275082733175633e-05, "loss": 0.3464, "step": 2939 }, { "epoch": 0.9491908470882602, "grad_norm": 0.3239061125660452, "learning_rate": 1.2273023817185284e-05, "loss": 0.3175, "step": 2940 }, { "epoch": 0.949513701117882, "grad_norm": 0.49015030814858174, "learning_rate": 1.2270964296459387e-05, "loss": 0.3971, "step": 2941 }, { "epoch": 0.949836555147504, "grad_norm": 0.3828839048789399, "learning_rate": 1.2268904171258886e-05, "loss": 0.3866, "step": 2942 }, { "epoch": 0.9501594091771258, "grad_norm": 0.3872347950625705, "learning_rate": 1.2266843441844791e-05, "loss": 0.3597, "step": 2943 }, { "epoch": 0.9504822632067477, "grad_norm": 0.39490049955272577, "learning_rate": 1.2264782108478199e-05, "loss": 0.4202, "step": 2944 }, { "epoch": 0.9508051172363695, "grad_norm": 0.3992494416626229, "learning_rate": 1.2262720171420275e-05, "loss": 0.3803, "step": 2945 }, { "epoch": 0.9511279712659914, "grad_norm": 0.4502170596451674, "learning_rate": 1.2260657630932262e-05, "loss": 0.3592, "step": 2946 }, { "epoch": 0.9514508252956132, "grad_norm": 0.500147281518187, "learning_rate": 1.2258594487275485e-05, "loss": 0.3017, "step": 2947 }, { "epoch": 0.951773679325235, "grad_norm": 0.45268688838945287, "learning_rate": 1.225653074071134e-05, "loss": 0.3788, "step": 2948 }, { "epoch": 0.9520965333548569, "grad_norm": 0.49568376930583025, "learning_rate": 1.2254466391501296e-05, "loss": 0.3817, "step": 2949 }, { "epoch": 0.9524193873844788, "grad_norm": 0.3812921043973889, "learning_rate": 1.225240143990691e-05, "loss": 0.3841, "step": 2950 }, { "epoch": 0.9527422414141007, "grad_norm": 0.46306260096696833, "learning_rate": 1.225033588618981e-05, "loss": 0.3843, "step": 2951 }, { "epoch": 0.9530650954437225, "grad_norm": 0.3905395731609217, "learning_rate": 1.2248269730611688e-05, "loss": 0.3233, "step": 2952 }, { "epoch": 0.9533879494733444, "grad_norm": 0.46103741526146924, "learning_rate": 1.2246202973434334e-05, "loss": 0.3465, "step": 2953 }, { "epoch": 0.9537108035029662, "grad_norm": 0.4873280884693344, "learning_rate": 1.2244135614919598e-05, "loss": 0.3541, "step": 2954 }, { "epoch": 0.9540336575325881, "grad_norm": 0.5858082540661149, "learning_rate": 1.2242067655329414e-05, "loss": 0.3255, "step": 2955 }, { "epoch": 0.9543565115622099, "grad_norm": 0.33721349474276424, "learning_rate": 1.2239999094925785e-05, "loss": 0.3016, "step": 2956 }, { "epoch": 0.9546793655918318, "grad_norm": 0.3735543665097669, "learning_rate": 1.22379299339708e-05, "loss": 0.3325, "step": 2957 }, { "epoch": 0.9550022196214536, "grad_norm": 0.49869132859538007, "learning_rate": 1.2235860172726618e-05, "loss": 0.339, "step": 2958 }, { "epoch": 0.9553250736510756, "grad_norm": 0.4956631266480101, "learning_rate": 1.2233789811455473e-05, "loss": 0.3293, "step": 2959 }, { "epoch": 0.9556479276806974, "grad_norm": 0.45236035905311783, "learning_rate": 1.2231718850419679e-05, "loss": 0.4012, "step": 2960 }, { "epoch": 0.9559707817103192, "grad_norm": 0.36667986499144295, "learning_rate": 1.222964728988162e-05, "loss": 0.3014, "step": 2961 }, { "epoch": 0.9562936357399411, "grad_norm": 0.45824217861482547, "learning_rate": 1.2227575130103764e-05, "loss": 0.343, "step": 2962 }, { "epoch": 0.9566164897695629, "grad_norm": 0.41187524896579836, "learning_rate": 1.2225502371348648e-05, "loss": 0.3802, "step": 2963 }, { "epoch": 0.9569393437991848, "grad_norm": 0.4066633505881953, "learning_rate": 1.222342901387889e-05, "loss": 0.3595, "step": 2964 }, { "epoch": 0.9572621978288066, "grad_norm": 0.44281794667612556, "learning_rate": 1.2221355057957182e-05, "loss": 0.384, "step": 2965 }, { "epoch": 0.9575850518584285, "grad_norm": 0.40417940640863453, "learning_rate": 1.2219280503846287e-05, "loss": 0.3976, "step": 2966 }, { "epoch": 0.9579079058880504, "grad_norm": 0.4536730758193306, "learning_rate": 1.221720535180905e-05, "loss": 0.3381, "step": 2967 }, { "epoch": 0.9582307599176723, "grad_norm": 0.36110441839315627, "learning_rate": 1.2215129602108394e-05, "loss": 0.3838, "step": 2968 }, { "epoch": 0.9585536139472941, "grad_norm": 0.4446522347889125, "learning_rate": 1.221305325500731e-05, "loss": 0.4018, "step": 2969 }, { "epoch": 0.958876467976916, "grad_norm": 0.40103427909131883, "learning_rate": 1.2210976310768868e-05, "loss": 0.4136, "step": 2970 }, { "epoch": 0.9591993220065378, "grad_norm": 0.5614398812998662, "learning_rate": 1.2208898769656217e-05, "loss": 0.3286, "step": 2971 }, { "epoch": 0.9595221760361596, "grad_norm": 1.0015561460198266, "learning_rate": 1.2206820631932572e-05, "loss": 0.3855, "step": 2972 }, { "epoch": 0.9598450300657815, "grad_norm": 0.3643663472674865, "learning_rate": 1.220474189786124e-05, "loss": 0.3809, "step": 2973 }, { "epoch": 0.9601678840954033, "grad_norm": 0.455012750916411, "learning_rate": 1.2202662567705586e-05, "loss": 0.352, "step": 2974 }, { "epoch": 0.9604907381250253, "grad_norm": 0.4498800193826464, "learning_rate": 1.220058264172906e-05, "loss": 0.3516, "step": 2975 }, { "epoch": 0.9608135921546471, "grad_norm": 0.5544925540139278, "learning_rate": 1.2198502120195185e-05, "loss": 0.367, "step": 2976 }, { "epoch": 0.961136446184269, "grad_norm": 0.40460107146194085, "learning_rate": 1.2196421003367564e-05, "loss": 0.397, "step": 2977 }, { "epoch": 0.9614593002138908, "grad_norm": 0.452682872963805, "learning_rate": 1.2194339291509871e-05, "loss": 0.3218, "step": 2978 }, { "epoch": 0.9617821542435127, "grad_norm": 0.36467069964787796, "learning_rate": 1.2192256984885852e-05, "loss": 0.4036, "step": 2979 }, { "epoch": 0.9621050082731345, "grad_norm": 0.35183365545951856, "learning_rate": 1.2190174083759334e-05, "loss": 0.3334, "step": 2980 }, { "epoch": 0.9624278623027563, "grad_norm": 0.48861230603795724, "learning_rate": 1.2188090588394217e-05, "loss": 0.3433, "step": 2981 }, { "epoch": 0.9627507163323782, "grad_norm": 0.4771739068173978, "learning_rate": 1.2186006499054482e-05, "loss": 0.2987, "step": 2982 }, { "epoch": 0.963073570362, "grad_norm": 0.45113822282378274, "learning_rate": 1.2183921816004173e-05, "loss": 0.3674, "step": 2983 }, { "epoch": 0.963396424391622, "grad_norm": 0.4407229276273073, "learning_rate": 1.2181836539507423e-05, "loss": 0.3704, "step": 2984 }, { "epoch": 0.9637192784212438, "grad_norm": 0.5119628724368583, "learning_rate": 1.2179750669828428e-05, "loss": 0.3354, "step": 2985 }, { "epoch": 0.9640421324508657, "grad_norm": 0.4302802606475726, "learning_rate": 1.2177664207231471e-05, "loss": 0.435, "step": 2986 }, { "epoch": 0.9643649864804875, "grad_norm": 0.3717513499798219, "learning_rate": 1.2175577151980897e-05, "loss": 0.3297, "step": 2987 }, { "epoch": 0.9646878405101094, "grad_norm": 0.40119922206254, "learning_rate": 1.2173489504341136e-05, "loss": 0.4044, "step": 2988 }, { "epoch": 0.9650106945397312, "grad_norm": 0.42000555427855896, "learning_rate": 1.2171401264576695e-05, "loss": 0.3629, "step": 2989 }, { "epoch": 0.965333548569353, "grad_norm": 0.396369618694451, "learning_rate": 1.2169312432952145e-05, "loss": 0.3249, "step": 2990 }, { "epoch": 0.9656564025989749, "grad_norm": 0.39182187892874926, "learning_rate": 1.2167223009732136e-05, "loss": 0.2796, "step": 2991 }, { "epoch": 0.9659792566285968, "grad_norm": 0.4437959172565454, "learning_rate": 1.2165132995181404e-05, "loss": 0.2943, "step": 2992 }, { "epoch": 0.9663021106582187, "grad_norm": 0.4715577800240589, "learning_rate": 1.2163042389564744e-05, "loss": 0.3609, "step": 2993 }, { "epoch": 0.9666249646878405, "grad_norm": 0.3612760048943095, "learning_rate": 1.2160951193147034e-05, "loss": 0.307, "step": 2994 }, { "epoch": 0.9669478187174624, "grad_norm": 0.4710403975499467, "learning_rate": 1.2158859406193226e-05, "loss": 0.3211, "step": 2995 }, { "epoch": 0.9672706727470842, "grad_norm": 0.45268402176579997, "learning_rate": 1.2156767028968347e-05, "loss": 0.3081, "step": 2996 }, { "epoch": 0.9675935267767061, "grad_norm": 0.4147586683618703, "learning_rate": 1.21546740617375e-05, "loss": 0.3729, "step": 2997 }, { "epoch": 0.9679163808063279, "grad_norm": 0.6790759094694807, "learning_rate": 1.2152580504765857e-05, "loss": 0.3488, "step": 2998 }, { "epoch": 0.9682392348359498, "grad_norm": 0.37485724275105264, "learning_rate": 1.2150486358318674e-05, "loss": 0.3528, "step": 2999 }, { "epoch": 0.9685620888655716, "grad_norm": 0.5279062835998641, "learning_rate": 1.214839162266127e-05, "loss": 0.3709, "step": 3000 }, { "epoch": 0.9688849428951936, "grad_norm": 0.6694450775221298, "learning_rate": 1.214629629805905e-05, "loss": 0.3375, "step": 3001 }, { "epoch": 0.9692077969248154, "grad_norm": 0.40882208426538413, "learning_rate": 1.2144200384777486e-05, "loss": 0.3444, "step": 3002 }, { "epoch": 0.9695306509544372, "grad_norm": 0.4660900313580999, "learning_rate": 1.2142103883082133e-05, "loss": 0.3199, "step": 3003 }, { "epoch": 0.9698535049840591, "grad_norm": 0.39528176422855843, "learning_rate": 1.2140006793238603e-05, "loss": 0.3019, "step": 3004 }, { "epoch": 0.9701763590136809, "grad_norm": 0.45210534267360714, "learning_rate": 1.213790911551261e-05, "loss": 0.3527, "step": 3005 }, { "epoch": 0.9704992130433028, "grad_norm": 0.41856441080358114, "learning_rate": 1.2135810850169914e-05, "loss": 0.3376, "step": 3006 }, { "epoch": 0.9708220670729246, "grad_norm": 0.3612254301773118, "learning_rate": 1.213371199747637e-05, "loss": 0.3092, "step": 3007 }, { "epoch": 0.9711449211025465, "grad_norm": 0.40861815203843826, "learning_rate": 1.2131612557697896e-05, "loss": 0.3951, "step": 3008 }, { "epoch": 0.9714677751321684, "grad_norm": 0.4614817805433779, "learning_rate": 1.2129512531100487e-05, "loss": 0.3549, "step": 3009 }, { "epoch": 0.9717906291617903, "grad_norm": 0.47940335248436156, "learning_rate": 1.2127411917950219e-05, "loss": 0.3368, "step": 3010 }, { "epoch": 0.9721134831914121, "grad_norm": 0.45957838124118106, "learning_rate": 1.2125310718513229e-05, "loss": 0.3791, "step": 3011 }, { "epoch": 0.972436337221034, "grad_norm": 0.33627045375161463, "learning_rate": 1.2123208933055744e-05, "loss": 0.3377, "step": 3012 }, { "epoch": 0.9727591912506558, "grad_norm": 0.36386753321281745, "learning_rate": 1.2121106561844053e-05, "loss": 0.333, "step": 3013 }, { "epoch": 0.9730820452802776, "grad_norm": 0.46399644713954097, "learning_rate": 1.2119003605144524e-05, "loss": 0.3085, "step": 3014 }, { "epoch": 0.9734048993098995, "grad_norm": 0.44271944060446894, "learning_rate": 1.2116900063223599e-05, "loss": 0.3225, "step": 3015 }, { "epoch": 0.9737277533395213, "grad_norm": 0.39575777802800644, "learning_rate": 1.2114795936347792e-05, "loss": 0.3588, "step": 3016 }, { "epoch": 0.9740506073691432, "grad_norm": 0.4901094045034274, "learning_rate": 1.2112691224783698e-05, "loss": 0.3112, "step": 3017 }, { "epoch": 0.9743734613987651, "grad_norm": 0.45874405402286456, "learning_rate": 1.2110585928797976e-05, "loss": 0.3613, "step": 3018 }, { "epoch": 0.974696315428387, "grad_norm": 0.5050303575498586, "learning_rate": 1.2108480048657367e-05, "loss": 0.3023, "step": 3019 }, { "epoch": 0.9750191694580088, "grad_norm": 0.4216577318972352, "learning_rate": 1.2106373584628682e-05, "loss": 0.4088, "step": 3020 }, { "epoch": 0.9753420234876307, "grad_norm": 0.386315234690423, "learning_rate": 1.2104266536978806e-05, "loss": 0.2879, "step": 3021 }, { "epoch": 0.9756648775172525, "grad_norm": 0.371567772729797, "learning_rate": 1.2102158905974701e-05, "loss": 0.3966, "step": 3022 }, { "epoch": 0.9759877315468743, "grad_norm": 0.3555603657885529, "learning_rate": 1.2100050691883397e-05, "loss": 0.3341, "step": 3023 }, { "epoch": 0.9763105855764962, "grad_norm": 0.5237932239535917, "learning_rate": 1.2097941894972009e-05, "loss": 0.3858, "step": 3024 }, { "epoch": 0.976633439606118, "grad_norm": 0.3765692355731033, "learning_rate": 1.2095832515507716e-05, "loss": 0.3475, "step": 3025 }, { "epoch": 0.97695629363574, "grad_norm": 0.41171924966913737, "learning_rate": 1.2093722553757768e-05, "loss": 0.3331, "step": 3026 }, { "epoch": 0.9772791476653618, "grad_norm": 0.38449830506252913, "learning_rate": 1.20916120099895e-05, "loss": 0.3443, "step": 3027 }, { "epoch": 0.9776020016949837, "grad_norm": 0.41273087598385705, "learning_rate": 1.2089500884470314e-05, "loss": 0.3625, "step": 3028 }, { "epoch": 0.9779248557246055, "grad_norm": 0.4478755654195249, "learning_rate": 1.2087389177467687e-05, "loss": 0.3699, "step": 3029 }, { "epoch": 0.9782477097542274, "grad_norm": 0.38799208594876455, "learning_rate": 1.2085276889249166e-05, "loss": 0.4242, "step": 3030 }, { "epoch": 0.9785705637838492, "grad_norm": 0.40394033579073235, "learning_rate": 1.2083164020082378e-05, "loss": 0.3344, "step": 3031 }, { "epoch": 0.9788934178134711, "grad_norm": 0.43631617241399917, "learning_rate": 1.2081050570235021e-05, "loss": 0.3928, "step": 3032 }, { "epoch": 0.9792162718430929, "grad_norm": 0.4368096810758206, "learning_rate": 1.2078936539974865e-05, "loss": 0.3351, "step": 3033 }, { "epoch": 0.9795391258727147, "grad_norm": 0.552774684170223, "learning_rate": 1.2076821929569755e-05, "loss": 0.3386, "step": 3034 }, { "epoch": 0.9798619799023367, "grad_norm": 0.33304579961340797, "learning_rate": 1.2074706739287608e-05, "loss": 0.3166, "step": 3035 }, { "epoch": 0.9801848339319585, "grad_norm": 0.3299544155205845, "learning_rate": 1.2072590969396419e-05, "loss": 0.3112, "step": 3036 }, { "epoch": 0.9805076879615804, "grad_norm": 0.43925025434789194, "learning_rate": 1.2070474620164247e-05, "loss": 0.3524, "step": 3037 }, { "epoch": 0.9808305419912022, "grad_norm": 0.5024993841577757, "learning_rate": 1.2068357691859239e-05, "loss": 0.3964, "step": 3038 }, { "epoch": 0.9811533960208241, "grad_norm": 0.4033655614325291, "learning_rate": 1.20662401847496e-05, "loss": 0.3647, "step": 3039 }, { "epoch": 0.9814762500504459, "grad_norm": 0.3876238329753923, "learning_rate": 1.2064122099103618e-05, "loss": 0.3963, "step": 3040 }, { "epoch": 0.9817991040800678, "grad_norm": 0.4244157451127379, "learning_rate": 1.2062003435189652e-05, "loss": 0.4227, "step": 3041 }, { "epoch": 0.9821219581096896, "grad_norm": 0.402444050959513, "learning_rate": 1.2059884193276133e-05, "loss": 0.3477, "step": 3042 }, { "epoch": 0.9824448121393116, "grad_norm": 0.4347153929790065, "learning_rate": 1.2057764373631564e-05, "loss": 0.4207, "step": 3043 }, { "epoch": 0.9827676661689334, "grad_norm": 0.39562094333338543, "learning_rate": 1.2055643976524528e-05, "loss": 0.3901, "step": 3044 }, { "epoch": 0.9830905201985553, "grad_norm": 0.4758723200524293, "learning_rate": 1.2053523002223672e-05, "loss": 0.3897, "step": 3045 }, { "epoch": 0.9834133742281771, "grad_norm": 0.47784184325241885, "learning_rate": 1.2051401450997723e-05, "loss": 0.4335, "step": 3046 }, { "epoch": 0.9837362282577989, "grad_norm": 0.3950899352957644, "learning_rate": 1.2049279323115477e-05, "loss": 0.3554, "step": 3047 }, { "epoch": 0.9840590822874208, "grad_norm": 0.3959111471163805, "learning_rate": 1.2047156618845809e-05, "loss": 0.3415, "step": 3048 }, { "epoch": 0.9843819363170426, "grad_norm": 0.49919563744534834, "learning_rate": 1.2045033338457653e-05, "loss": 0.3648, "step": 3049 }, { "epoch": 0.9847047903466645, "grad_norm": 0.41115324268531794, "learning_rate": 1.2042909482220037e-05, "loss": 0.3548, "step": 3050 }, { "epoch": 0.9850276443762864, "grad_norm": 0.4405873711036808, "learning_rate": 1.2040785050402044e-05, "loss": 0.376, "step": 3051 }, { "epoch": 0.9853504984059083, "grad_norm": 0.37037126647730834, "learning_rate": 1.2038660043272837e-05, "loss": 0.3627, "step": 3052 }, { "epoch": 0.9856733524355301, "grad_norm": 0.47900038624461666, "learning_rate": 1.2036534461101654e-05, "loss": 0.4007, "step": 3053 }, { "epoch": 0.985996206465152, "grad_norm": 0.39673043418271337, "learning_rate": 1.2034408304157801e-05, "loss": 0.3308, "step": 3054 }, { "epoch": 0.9863190604947738, "grad_norm": 0.4755976185393055, "learning_rate": 1.2032281572710661e-05, "loss": 0.3543, "step": 3055 }, { "epoch": 0.9866419145243956, "grad_norm": 0.5201768303144921, "learning_rate": 1.2030154267029687e-05, "loss": 0.347, "step": 3056 }, { "epoch": 0.9869647685540175, "grad_norm": 0.3895545104392051, "learning_rate": 1.2028026387384405e-05, "loss": 0.3361, "step": 3057 }, { "epoch": 0.9872876225836393, "grad_norm": 0.3594902886980607, "learning_rate": 1.2025897934044417e-05, "loss": 0.3126, "step": 3058 }, { "epoch": 0.9876104766132612, "grad_norm": 0.4590922553478531, "learning_rate": 1.2023768907279393e-05, "loss": 0.3673, "step": 3059 }, { "epoch": 0.9879333306428831, "grad_norm": 0.3520197609008487, "learning_rate": 1.2021639307359073e-05, "loss": 0.3479, "step": 3060 }, { "epoch": 0.988256184672505, "grad_norm": 0.3955752359115871, "learning_rate": 1.2019509134553286e-05, "loss": 0.3401, "step": 3061 }, { "epoch": 0.9885790387021268, "grad_norm": 0.36318336719306504, "learning_rate": 1.201737838913191e-05, "loss": 0.2633, "step": 3062 }, { "epoch": 0.9889018927317487, "grad_norm": 0.5257888607937192, "learning_rate": 1.2015247071364916e-05, "loss": 0.3332, "step": 3063 }, { "epoch": 0.9892247467613705, "grad_norm": 0.31088222051344905, "learning_rate": 1.2013115181522337e-05, "loss": 0.3032, "step": 3064 }, { "epoch": 0.9895476007909924, "grad_norm": 0.4405780759652841, "learning_rate": 1.2010982719874277e-05, "loss": 0.3669, "step": 3065 }, { "epoch": 0.9898704548206142, "grad_norm": 0.49208766110452434, "learning_rate": 1.200884968669092e-05, "loss": 0.2923, "step": 3066 }, { "epoch": 0.990193308850236, "grad_norm": 0.6165675412169117, "learning_rate": 1.2006716082242515e-05, "loss": 0.3431, "step": 3067 }, { "epoch": 0.990516162879858, "grad_norm": 0.35716753035069815, "learning_rate": 1.2004581906799392e-05, "loss": 0.297, "step": 3068 }, { "epoch": 0.9908390169094798, "grad_norm": 0.3467553876002782, "learning_rate": 1.2002447160631942e-05, "loss": 0.3625, "step": 3069 }, { "epoch": 0.9911618709391017, "grad_norm": 0.38420903541419793, "learning_rate": 1.2000311844010636e-05, "loss": 0.3985, "step": 3070 }, { "epoch": 0.9914847249687235, "grad_norm": 0.5240233126898277, "learning_rate": 1.199817595720602e-05, "loss": 0.3646, "step": 3071 }, { "epoch": 0.9918075789983454, "grad_norm": 0.42301021313436543, "learning_rate": 1.1996039500488702e-05, "loss": 0.336, "step": 3072 }, { "epoch": 0.9921304330279672, "grad_norm": 0.37408160960886583, "learning_rate": 1.1993902474129372e-05, "loss": 0.3326, "step": 3073 }, { "epoch": 0.9924532870575891, "grad_norm": 0.5432451722477853, "learning_rate": 1.1991764878398785e-05, "loss": 0.349, "step": 3074 }, { "epoch": 0.9927761410872109, "grad_norm": 0.42204827819424523, "learning_rate": 1.1989626713567774e-05, "loss": 0.3674, "step": 3075 }, { "epoch": 0.9930989951168328, "grad_norm": 0.4520031086701255, "learning_rate": 1.198748797990724e-05, "loss": 0.3332, "step": 3076 }, { "epoch": 0.9934218491464547, "grad_norm": 0.6612615103012696, "learning_rate": 1.1985348677688162e-05, "loss": 0.3355, "step": 3077 }, { "epoch": 0.9937447031760765, "grad_norm": 0.43967545944910585, "learning_rate": 1.1983208807181581e-05, "loss": 0.3923, "step": 3078 }, { "epoch": 0.9940675572056984, "grad_norm": 0.4666337064980386, "learning_rate": 1.1981068368658617e-05, "loss": 0.3474, "step": 3079 }, { "epoch": 0.9943904112353202, "grad_norm": 0.39683421906124594, "learning_rate": 1.1978927362390463e-05, "loss": 0.3539, "step": 3080 }, { "epoch": 0.9947132652649421, "grad_norm": 0.4339852112048939, "learning_rate": 1.197678578864838e-05, "loss": 0.4124, "step": 3081 }, { "epoch": 0.9950361192945639, "grad_norm": 0.355338208390581, "learning_rate": 1.1974643647703702e-05, "loss": 0.3643, "step": 3082 }, { "epoch": 0.9953589733241858, "grad_norm": 0.37481331626454983, "learning_rate": 1.1972500939827837e-05, "loss": 0.3285, "step": 3083 }, { "epoch": 0.9956818273538076, "grad_norm": 0.5085497641854676, "learning_rate": 1.1970357665292261e-05, "loss": 0.3428, "step": 3084 }, { "epoch": 0.9960046813834296, "grad_norm": 0.40796977651846494, "learning_rate": 1.1968213824368526e-05, "loss": 0.3516, "step": 3085 }, { "epoch": 0.9963275354130514, "grad_norm": 0.4238107780742868, "learning_rate": 1.196606941732825e-05, "loss": 0.3903, "step": 3086 }, { "epoch": 0.9966503894426733, "grad_norm": 0.4162582132371185, "learning_rate": 1.196392444444313e-05, "loss": 0.371, "step": 3087 }, { "epoch": 0.9969732434722951, "grad_norm": 0.41341147404068135, "learning_rate": 1.1961778905984932e-05, "loss": 0.3412, "step": 3088 }, { "epoch": 0.997296097501917, "grad_norm": 0.38200316889737645, "learning_rate": 1.1959632802225488e-05, "loss": 0.3188, "step": 3089 }, { "epoch": 0.9976189515315388, "grad_norm": 0.46995152972105614, "learning_rate": 1.195748613343671e-05, "loss": 0.3803, "step": 3090 }, { "epoch": 0.9979418055611606, "grad_norm": 0.4680880006903165, "learning_rate": 1.195533889989058e-05, "loss": 0.4084, "step": 3091 }, { "epoch": 0.9982646595907825, "grad_norm": 0.3639332834437348, "learning_rate": 1.1953191101859144e-05, "loss": 0.2937, "step": 3092 }, { "epoch": 0.9985875136204043, "grad_norm": 0.4150901532774408, "learning_rate": 1.1951042739614527e-05, "loss": 0.3967, "step": 3093 }, { "epoch": 0.9989103676500263, "grad_norm": 0.37962845860520794, "learning_rate": 1.1948893813428927e-05, "loss": 0.341, "step": 3094 }, { "epoch": 0.9992332216796481, "grad_norm": 0.3678740528280721, "learning_rate": 1.1946744323574606e-05, "loss": 0.3136, "step": 3095 }, { "epoch": 0.99955607570927, "grad_norm": 0.37437217824973984, "learning_rate": 1.1944594270323902e-05, "loss": 0.3237, "step": 3096 }, { "epoch": 0.9998789297388918, "grad_norm": 0.4836732200695029, "learning_rate": 1.1942443653949226e-05, "loss": 0.3201, "step": 3097 }, { "epoch": 1.0, "grad_norm": 1.0388417018351515, "learning_rate": 1.1940292474723054e-05, "loss": 0.4114, "step": 3098 }, { "epoch": 1.0003228540296218, "grad_norm": 0.4973544988679974, "learning_rate": 1.1938140732917944e-05, "loss": 0.2628, "step": 3099 }, { "epoch": 1.0006457080592437, "grad_norm": 0.39719602895637524, "learning_rate": 1.193598842880651e-05, "loss": 0.3027, "step": 3100 }, { "epoch": 1.0009685620888655, "grad_norm": 0.42117522398766777, "learning_rate": 1.1933835562661456e-05, "loss": 0.323, "step": 3101 }, { "epoch": 1.0012914161184874, "grad_norm": 0.35755679510256994, "learning_rate": 1.193168213475554e-05, "loss": 0.3038, "step": 3102 }, { "epoch": 1.0016142701481092, "grad_norm": 0.4168402233027418, "learning_rate": 1.19295281453616e-05, "loss": 0.3007, "step": 3103 }, { "epoch": 1.001937124177731, "grad_norm": 0.3186395948866922, "learning_rate": 1.1927373594752542e-05, "loss": 0.2571, "step": 3104 }, { "epoch": 1.002259978207353, "grad_norm": 0.4379640442054533, "learning_rate": 1.1925218483201349e-05, "loss": 0.2801, "step": 3105 }, { "epoch": 1.0025828322369748, "grad_norm": 0.5283015153935992, "learning_rate": 1.1923062810981066e-05, "loss": 0.2794, "step": 3106 }, { "epoch": 1.0029056862665968, "grad_norm": 0.5266347677877131, "learning_rate": 1.192090657836482e-05, "loss": 0.308, "step": 3107 }, { "epoch": 1.0032285402962187, "grad_norm": 0.3562136832063903, "learning_rate": 1.1918749785625793e-05, "loss": 0.2711, "step": 3108 }, { "epoch": 1.0035513943258405, "grad_norm": 0.40639115572521745, "learning_rate": 1.1916592433037256e-05, "loss": 0.3091, "step": 3109 }, { "epoch": 1.0038742483554624, "grad_norm": 0.3750427896313809, "learning_rate": 1.191443452087254e-05, "loss": 0.2561, "step": 3110 }, { "epoch": 1.0041971023850842, "grad_norm": 0.4778880300020411, "learning_rate": 1.1912276049405049e-05, "loss": 0.416, "step": 3111 }, { "epoch": 1.004519956414706, "grad_norm": 0.4067849889704422, "learning_rate": 1.1910117018908257e-05, "loss": 0.3218, "step": 3112 }, { "epoch": 1.0048428104443279, "grad_norm": 0.6883989315272351, "learning_rate": 1.1907957429655713e-05, "loss": 0.3271, "step": 3113 }, { "epoch": 1.0051656644739497, "grad_norm": 0.49187665488546206, "learning_rate": 1.1905797281921032e-05, "loss": 0.3173, "step": 3114 }, { "epoch": 1.0054885185035716, "grad_norm": 0.40044419467465475, "learning_rate": 1.1903636575977902e-05, "loss": 0.2907, "step": 3115 }, { "epoch": 1.0058113725331934, "grad_norm": 0.331486870667273, "learning_rate": 1.1901475312100083e-05, "loss": 0.2679, "step": 3116 }, { "epoch": 1.0061342265628153, "grad_norm": 0.4124900296514769, "learning_rate": 1.1899313490561405e-05, "loss": 0.3419, "step": 3117 }, { "epoch": 1.006457080592437, "grad_norm": 0.409711895063535, "learning_rate": 1.1897151111635764e-05, "loss": 0.2513, "step": 3118 }, { "epoch": 1.006779934622059, "grad_norm": 0.43300899324183606, "learning_rate": 1.1894988175597132e-05, "loss": 0.3207, "step": 3119 }, { "epoch": 1.0071027886516808, "grad_norm": 0.3807560197218319, "learning_rate": 1.1892824682719551e-05, "loss": 0.3058, "step": 3120 }, { "epoch": 1.0074256426813026, "grad_norm": 0.39580824908942763, "learning_rate": 1.1890660633277132e-05, "loss": 0.3117, "step": 3121 }, { "epoch": 1.0077484967109245, "grad_norm": 0.4915386460574527, "learning_rate": 1.1888496027544058e-05, "loss": 0.3289, "step": 3122 }, { "epoch": 1.0080713507405463, "grad_norm": 0.48316513756127816, "learning_rate": 1.188633086579458e-05, "loss": 0.2811, "step": 3123 }, { "epoch": 1.0083942047701684, "grad_norm": 0.38558458772306226, "learning_rate": 1.1884165148303023e-05, "loss": 0.3292, "step": 3124 }, { "epoch": 1.0087170587997902, "grad_norm": 0.3978483557561998, "learning_rate": 1.1881998875343775e-05, "loss": 0.3476, "step": 3125 }, { "epoch": 1.009039912829412, "grad_norm": 0.4024783089179332, "learning_rate": 1.1879832047191308e-05, "loss": 0.3532, "step": 3126 }, { "epoch": 1.009362766859034, "grad_norm": 0.4125274833680995, "learning_rate": 1.187766466412015e-05, "loss": 0.2911, "step": 3127 }, { "epoch": 1.0096856208886558, "grad_norm": 0.37096473272214087, "learning_rate": 1.1875496726404907e-05, "loss": 0.3345, "step": 3128 }, { "epoch": 1.0100084749182776, "grad_norm": 0.4268581350445673, "learning_rate": 1.1873328234320256e-05, "loss": 0.3009, "step": 3129 }, { "epoch": 1.0103313289478995, "grad_norm": 0.40371582507693454, "learning_rate": 1.1871159188140941e-05, "loss": 0.2944, "step": 3130 }, { "epoch": 1.0106541829775213, "grad_norm": 0.5460512258384362, "learning_rate": 1.1868989588141773e-05, "loss": 0.3008, "step": 3131 }, { "epoch": 1.0109770370071431, "grad_norm": 0.37657032744344676, "learning_rate": 1.1866819434597641e-05, "loss": 0.2465, "step": 3132 }, { "epoch": 1.011299891036765, "grad_norm": 0.33911754223177015, "learning_rate": 1.1864648727783503e-05, "loss": 0.2728, "step": 3133 }, { "epoch": 1.0116227450663868, "grad_norm": 0.4701056781961189, "learning_rate": 1.1862477467974378e-05, "loss": 0.3477, "step": 3134 }, { "epoch": 1.0119455990960087, "grad_norm": 0.624154526379695, "learning_rate": 1.1860305655445368e-05, "loss": 0.2587, "step": 3135 }, { "epoch": 1.0122684531256305, "grad_norm": 0.36852014025872404, "learning_rate": 1.1858133290471634e-05, "loss": 0.2852, "step": 3136 }, { "epoch": 1.0125913071552524, "grad_norm": 0.39509546897339676, "learning_rate": 1.1855960373328416e-05, "loss": 0.2643, "step": 3137 }, { "epoch": 1.0129141611848742, "grad_norm": 0.4174651930377931, "learning_rate": 1.1853786904291016e-05, "loss": 0.28, "step": 3138 }, { "epoch": 1.013237015214496, "grad_norm": 0.3942921478985842, "learning_rate": 1.1851612883634811e-05, "loss": 0.3168, "step": 3139 }, { "epoch": 1.013559869244118, "grad_norm": 0.4409819756187125, "learning_rate": 1.1849438311635245e-05, "loss": 0.3024, "step": 3140 }, { "epoch": 1.01388272327374, "grad_norm": 0.35935713663103963, "learning_rate": 1.1847263188567836e-05, "loss": 0.2862, "step": 3141 }, { "epoch": 1.0142055773033618, "grad_norm": 0.5330981738750403, "learning_rate": 1.1845087514708167e-05, "loss": 0.3947, "step": 3142 }, { "epoch": 1.0145284313329836, "grad_norm": 0.3242469377241775, "learning_rate": 1.1842911290331894e-05, "loss": 0.3347, "step": 3143 }, { "epoch": 1.0148512853626055, "grad_norm": 0.3656195010114672, "learning_rate": 1.1840734515714739e-05, "loss": 0.2881, "step": 3144 }, { "epoch": 1.0151741393922273, "grad_norm": 0.48746408047510076, "learning_rate": 1.18385571911325e-05, "loss": 0.2668, "step": 3145 }, { "epoch": 1.0154969934218492, "grad_norm": 0.4090708733364688, "learning_rate": 1.1836379316861041e-05, "loss": 0.2903, "step": 3146 }, { "epoch": 1.015819847451471, "grad_norm": 0.4210928208381953, "learning_rate": 1.183420089317629e-05, "loss": 0.2945, "step": 3147 }, { "epoch": 1.0161427014810929, "grad_norm": 0.4073564918391538, "learning_rate": 1.1832021920354257e-05, "loss": 0.3389, "step": 3148 }, { "epoch": 1.0164655555107147, "grad_norm": 0.4071235682363763, "learning_rate": 1.1829842398671013e-05, "loss": 0.3042, "step": 3149 }, { "epoch": 1.0167884095403366, "grad_norm": 0.3771802791596265, "learning_rate": 1.18276623284027e-05, "loss": 0.3161, "step": 3150 }, { "epoch": 1.0171112635699584, "grad_norm": 0.3861505284067382, "learning_rate": 1.1825481709825526e-05, "loss": 0.3105, "step": 3151 }, { "epoch": 1.0174341175995802, "grad_norm": 0.40197463935480227, "learning_rate": 1.1823300543215777e-05, "loss": 0.2792, "step": 3152 }, { "epoch": 1.017756971629202, "grad_norm": 0.41108356118293005, "learning_rate": 1.1821118828849805e-05, "loss": 0.3682, "step": 3153 }, { "epoch": 1.018079825658824, "grad_norm": 0.34517466988281037, "learning_rate": 1.1818936567004027e-05, "loss": 0.3127, "step": 3154 }, { "epoch": 1.0184026796884458, "grad_norm": 0.4085453308907246, "learning_rate": 1.1816753757954934e-05, "loss": 0.3076, "step": 3155 }, { "epoch": 1.0187255337180676, "grad_norm": 0.4066178258211535, "learning_rate": 1.1814570401979082e-05, "loss": 0.3232, "step": 3156 }, { "epoch": 1.0190483877476897, "grad_norm": 0.5234818699784155, "learning_rate": 1.1812386499353104e-05, "loss": 0.2642, "step": 3157 }, { "epoch": 1.0193712417773115, "grad_norm": 0.4065113041579344, "learning_rate": 1.1810202050353693e-05, "loss": 0.2774, "step": 3158 }, { "epoch": 1.0196940958069334, "grad_norm": 0.42402508945502576, "learning_rate": 1.1808017055257618e-05, "loss": 0.2645, "step": 3159 }, { "epoch": 1.0200169498365552, "grad_norm": 0.3685871889316505, "learning_rate": 1.1805831514341717e-05, "loss": 0.3259, "step": 3160 }, { "epoch": 1.020339803866177, "grad_norm": 0.3841152246518712, "learning_rate": 1.1803645427882893e-05, "loss": 0.306, "step": 3161 }, { "epoch": 1.020662657895799, "grad_norm": 0.43761239249035716, "learning_rate": 1.1801458796158116e-05, "loss": 0.3585, "step": 3162 }, { "epoch": 1.0209855119254208, "grad_norm": 0.3951649905953798, "learning_rate": 1.1799271619444435e-05, "loss": 0.3218, "step": 3163 }, { "epoch": 1.0213083659550426, "grad_norm": 0.441389647340389, "learning_rate": 1.1797083898018958e-05, "loss": 0.2923, "step": 3164 }, { "epoch": 1.0216312199846644, "grad_norm": 0.4244324643186847, "learning_rate": 1.1794895632158873e-05, "loss": 0.3068, "step": 3165 }, { "epoch": 1.0219540740142863, "grad_norm": 0.6872380562483787, "learning_rate": 1.1792706822141423e-05, "loss": 0.306, "step": 3166 }, { "epoch": 1.0222769280439081, "grad_norm": 0.3990311134145014, "learning_rate": 1.179051746824393e-05, "loss": 0.3478, "step": 3167 }, { "epoch": 1.02259978207353, "grad_norm": 0.3677700195571162, "learning_rate": 1.1788327570743784e-05, "loss": 0.3349, "step": 3168 }, { "epoch": 1.0229226361031518, "grad_norm": 0.40112634039611755, "learning_rate": 1.1786137129918439e-05, "loss": 0.3462, "step": 3169 }, { "epoch": 1.0232454901327737, "grad_norm": 0.39723176067066085, "learning_rate": 1.1783946146045422e-05, "loss": 0.27, "step": 3170 }, { "epoch": 1.0235683441623955, "grad_norm": 0.4475197650389077, "learning_rate": 1.1781754619402328e-05, "loss": 0.2738, "step": 3171 }, { "epoch": 1.0238911981920173, "grad_norm": 0.5025917506979066, "learning_rate": 1.1779562550266818e-05, "loss": 0.3126, "step": 3172 }, { "epoch": 1.0242140522216392, "grad_norm": 0.46665008947097886, "learning_rate": 1.1777369938916627e-05, "loss": 0.3426, "step": 3173 }, { "epoch": 1.024536906251261, "grad_norm": 0.46388026671241195, "learning_rate": 1.1775176785629557e-05, "loss": 0.2533, "step": 3174 }, { "epoch": 1.024859760280883, "grad_norm": 0.42035187500186033, "learning_rate": 1.1772983090683472e-05, "loss": 0.3047, "step": 3175 }, { "epoch": 1.025182614310505, "grad_norm": 0.36422732324962276, "learning_rate": 1.1770788854356318e-05, "loss": 0.3576, "step": 3176 }, { "epoch": 1.0255054683401268, "grad_norm": 0.3481737747315477, "learning_rate": 1.1768594076926095e-05, "loss": 0.3019, "step": 3177 }, { "epoch": 1.0258283223697486, "grad_norm": 0.3710711797972702, "learning_rate": 1.1766398758670881e-05, "loss": 0.2902, "step": 3178 }, { "epoch": 1.0261511763993705, "grad_norm": 0.4229522816081565, "learning_rate": 1.176420289986882e-05, "loss": 0.3143, "step": 3179 }, { "epoch": 1.0264740304289923, "grad_norm": 0.4442496358120626, "learning_rate": 1.1762006500798123e-05, "loss": 0.3608, "step": 3180 }, { "epoch": 1.0267968844586142, "grad_norm": 0.44086165739953564, "learning_rate": 1.1759809561737072e-05, "loss": 0.336, "step": 3181 }, { "epoch": 1.027119738488236, "grad_norm": 0.38211097363425606, "learning_rate": 1.1757612082964014e-05, "loss": 0.3158, "step": 3182 }, { "epoch": 1.0274425925178579, "grad_norm": 0.4004202774063881, "learning_rate": 1.1755414064757369e-05, "loss": 0.3448, "step": 3183 }, { "epoch": 1.0277654465474797, "grad_norm": 0.4499330554145203, "learning_rate": 1.1753215507395625e-05, "loss": 0.3716, "step": 3184 }, { "epoch": 1.0280883005771015, "grad_norm": 0.5297914048886039, "learning_rate": 1.1751016411157331e-05, "loss": 0.3307, "step": 3185 }, { "epoch": 1.0284111546067234, "grad_norm": 0.40211100941590716, "learning_rate": 1.174881677632111e-05, "loss": 0.2813, "step": 3186 }, { "epoch": 1.0287340086363452, "grad_norm": 0.429222611106531, "learning_rate": 1.1746616603165657e-05, "loss": 0.2807, "step": 3187 }, { "epoch": 1.029056862665967, "grad_norm": 0.4857294941350797, "learning_rate": 1.1744415891969727e-05, "loss": 0.2855, "step": 3188 }, { "epoch": 1.029379716695589, "grad_norm": 0.38708139131830693, "learning_rate": 1.1742214643012147e-05, "loss": 0.3233, "step": 3189 }, { "epoch": 1.0297025707252108, "grad_norm": 0.4230111361204129, "learning_rate": 1.1740012856571816e-05, "loss": 0.3164, "step": 3190 }, { "epoch": 1.0300254247548328, "grad_norm": 0.4233122506301514, "learning_rate": 1.1737810532927694e-05, "loss": 0.3115, "step": 3191 }, { "epoch": 1.0303482787844547, "grad_norm": 0.49103289676957135, "learning_rate": 1.1735607672358813e-05, "loss": 0.4023, "step": 3192 }, { "epoch": 1.0306711328140765, "grad_norm": 0.4124885170344019, "learning_rate": 1.1733404275144275e-05, "loss": 0.3085, "step": 3193 }, { "epoch": 1.0309939868436984, "grad_norm": 0.4498208959741758, "learning_rate": 1.1731200341563243e-05, "loss": 0.3025, "step": 3194 }, { "epoch": 1.0313168408733202, "grad_norm": 0.3858864195239608, "learning_rate": 1.1728995871894954e-05, "loss": 0.2972, "step": 3195 }, { "epoch": 1.031639694902942, "grad_norm": 0.5112472473579484, "learning_rate": 1.1726790866418711e-05, "loss": 0.3142, "step": 3196 }, { "epoch": 1.031962548932564, "grad_norm": 0.4985397441422396, "learning_rate": 1.1724585325413889e-05, "loss": 0.2661, "step": 3197 }, { "epoch": 1.0322854029621857, "grad_norm": 0.45778763237887304, "learning_rate": 1.1722379249159921e-05, "loss": 0.2988, "step": 3198 }, { "epoch": 1.0326082569918076, "grad_norm": 0.4163398901309678, "learning_rate": 1.1720172637936318e-05, "loss": 0.2693, "step": 3199 }, { "epoch": 1.0329311110214294, "grad_norm": 0.37768895386213946, "learning_rate": 1.1717965492022653e-05, "loss": 0.3095, "step": 3200 }, { "epoch": 1.0332539650510513, "grad_norm": 0.3780200058538394, "learning_rate": 1.171575781169857e-05, "loss": 0.3108, "step": 3201 }, { "epoch": 1.0335768190806731, "grad_norm": 0.35354182575464577, "learning_rate": 1.1713549597243774e-05, "loss": 0.3356, "step": 3202 }, { "epoch": 1.033899673110295, "grad_norm": 0.41591923674601894, "learning_rate": 1.171134084893805e-05, "loss": 0.321, "step": 3203 }, { "epoch": 1.0342225271399168, "grad_norm": 0.4728869075525744, "learning_rate": 1.1709131567061239e-05, "loss": 0.3145, "step": 3204 }, { "epoch": 1.0345453811695386, "grad_norm": 0.5163749159178179, "learning_rate": 1.1706921751893255e-05, "loss": 0.3274, "step": 3205 }, { "epoch": 1.0348682351991605, "grad_norm": 0.4041611713955415, "learning_rate": 1.1704711403714078e-05, "loss": 0.2729, "step": 3206 }, { "epoch": 1.0351910892287823, "grad_norm": 0.44414652807002386, "learning_rate": 1.1702500522803755e-05, "loss": 0.3073, "step": 3207 }, { "epoch": 1.0355139432584044, "grad_norm": 0.535169822224103, "learning_rate": 1.1700289109442403e-05, "loss": 0.2956, "step": 3208 }, { "epoch": 1.0358367972880262, "grad_norm": 0.9278148021663927, "learning_rate": 1.1698077163910207e-05, "loss": 0.2581, "step": 3209 }, { "epoch": 1.036159651317648, "grad_norm": 0.5817302036110974, "learning_rate": 1.1695864686487413e-05, "loss": 0.2614, "step": 3210 }, { "epoch": 1.03648250534727, "grad_norm": 0.43795709635510227, "learning_rate": 1.1693651677454345e-05, "loss": 0.3132, "step": 3211 }, { "epoch": 1.0368053593768918, "grad_norm": 0.5094143024474453, "learning_rate": 1.1691438137091383e-05, "loss": 0.3336, "step": 3212 }, { "epoch": 1.0371282134065136, "grad_norm": 0.5112996643814245, "learning_rate": 1.168922406567898e-05, "loss": 0.3252, "step": 3213 }, { "epoch": 1.0374510674361355, "grad_norm": 0.4745146720250896, "learning_rate": 1.1687009463497657e-05, "loss": 0.3031, "step": 3214 }, { "epoch": 1.0377739214657573, "grad_norm": 0.42074239655601287, "learning_rate": 1.1684794330828e-05, "loss": 0.2825, "step": 3215 }, { "epoch": 1.0380967754953792, "grad_norm": 0.5286325097075977, "learning_rate": 1.1682578667950665e-05, "loss": 0.3215, "step": 3216 }, { "epoch": 1.038419629525001, "grad_norm": 0.40126726166772475, "learning_rate": 1.1680362475146373e-05, "loss": 0.2857, "step": 3217 }, { "epoch": 1.0387424835546228, "grad_norm": 0.4679948708532106, "learning_rate": 1.167814575269591e-05, "loss": 0.3421, "step": 3218 }, { "epoch": 1.0390653375842447, "grad_norm": 0.34303600596838996, "learning_rate": 1.1675928500880136e-05, "loss": 0.2824, "step": 3219 }, { "epoch": 1.0393881916138665, "grad_norm": 0.4210314144782469, "learning_rate": 1.1673710719979972e-05, "loss": 0.3392, "step": 3220 }, { "epoch": 1.0397110456434884, "grad_norm": 0.33659746003573526, "learning_rate": 1.1671492410276405e-05, "loss": 0.3171, "step": 3221 }, { "epoch": 1.0400338996731102, "grad_norm": 0.465679590376242, "learning_rate": 1.1669273572050496e-05, "loss": 0.2998, "step": 3222 }, { "epoch": 1.040356753702732, "grad_norm": 0.36458234999143924, "learning_rate": 1.1667054205583366e-05, "loss": 0.3091, "step": 3223 }, { "epoch": 1.040679607732354, "grad_norm": 0.5703007114779547, "learning_rate": 1.1664834311156207e-05, "loss": 0.3066, "step": 3224 }, { "epoch": 1.041002461761976, "grad_norm": 0.3616181675647446, "learning_rate": 1.1662613889050275e-05, "loss": 0.2876, "step": 3225 }, { "epoch": 1.0413253157915978, "grad_norm": 0.41457621576568915, "learning_rate": 1.1660392939546896e-05, "loss": 0.2823, "step": 3226 }, { "epoch": 1.0416481698212197, "grad_norm": 1.8833074447074143, "learning_rate": 1.1658171462927462e-05, "loss": 0.3209, "step": 3227 }, { "epoch": 1.0419710238508415, "grad_norm": 0.44761751852676546, "learning_rate": 1.165594945947343e-05, "loss": 0.3245, "step": 3228 }, { "epoch": 1.0422938778804633, "grad_norm": 0.3964689509874384, "learning_rate": 1.165372692946632e-05, "loss": 0.2885, "step": 3229 }, { "epoch": 1.0426167319100852, "grad_norm": 0.39122900599386173, "learning_rate": 1.1651503873187734e-05, "loss": 0.2564, "step": 3230 }, { "epoch": 1.042939585939707, "grad_norm": 0.4612643355155146, "learning_rate": 1.1649280290919321e-05, "loss": 0.2669, "step": 3231 }, { "epoch": 1.0432624399693289, "grad_norm": 0.46380913861478174, "learning_rate": 1.1647056182942808e-05, "loss": 0.3918, "step": 3232 }, { "epoch": 1.0435852939989507, "grad_norm": 0.4795909447847717, "learning_rate": 1.164483154953999e-05, "loss": 0.3419, "step": 3233 }, { "epoch": 1.0439081480285726, "grad_norm": 0.38885202436364597, "learning_rate": 1.1642606390992723e-05, "loss": 0.2908, "step": 3234 }, { "epoch": 1.0442310020581944, "grad_norm": 0.4333197744911468, "learning_rate": 1.164038070758293e-05, "loss": 0.2771, "step": 3235 }, { "epoch": 1.0445538560878163, "grad_norm": 0.43409642333262916, "learning_rate": 1.16381544995926e-05, "loss": 0.2698, "step": 3236 }, { "epoch": 1.044876710117438, "grad_norm": 0.40153943813049464, "learning_rate": 1.1635927767303796e-05, "loss": 0.2717, "step": 3237 }, { "epoch": 1.04519956414706, "grad_norm": 0.41816801259162695, "learning_rate": 1.1633700510998641e-05, "loss": 0.3596, "step": 3238 }, { "epoch": 1.0455224181766818, "grad_norm": 0.4399588515595115, "learning_rate": 1.1631472730959322e-05, "loss": 0.2865, "step": 3239 }, { "epoch": 1.0458452722063036, "grad_norm": 0.41058768130801726, "learning_rate": 1.1629244427468098e-05, "loss": 0.3512, "step": 3240 }, { "epoch": 1.0461681262359255, "grad_norm": 0.5289013163209212, "learning_rate": 1.1627015600807292e-05, "loss": 0.3343, "step": 3241 }, { "epoch": 1.0464909802655475, "grad_norm": 0.3752382994870465, "learning_rate": 1.162478625125929e-05, "loss": 0.3328, "step": 3242 }, { "epoch": 1.0468138342951694, "grad_norm": 0.393088512777299, "learning_rate": 1.1622556379106552e-05, "loss": 0.3076, "step": 3243 }, { "epoch": 1.0471366883247912, "grad_norm": 0.42824173149483047, "learning_rate": 1.1620325984631596e-05, "loss": 0.3203, "step": 3244 }, { "epoch": 1.047459542354413, "grad_norm": 0.3854286518636982, "learning_rate": 1.1618095068117013e-05, "loss": 0.3302, "step": 3245 }, { "epoch": 1.047782396384035, "grad_norm": 0.424921748664788, "learning_rate": 1.1615863629845457e-05, "loss": 0.2592, "step": 3246 }, { "epoch": 1.0481052504136568, "grad_norm": 0.47158637116736835, "learning_rate": 1.1613631670099645e-05, "loss": 0.2775, "step": 3247 }, { "epoch": 1.0484281044432786, "grad_norm": 0.39617375107293634, "learning_rate": 1.1611399189162367e-05, "loss": 0.3268, "step": 3248 }, { "epoch": 1.0487509584729005, "grad_norm": 0.4487819744593639, "learning_rate": 1.1609166187316472e-05, "loss": 0.2537, "step": 3249 }, { "epoch": 1.0490738125025223, "grad_norm": 0.41512923460305984, "learning_rate": 1.1606932664844881e-05, "loss": 0.3205, "step": 3250 }, { "epoch": 1.0493966665321441, "grad_norm": 0.44707670477955497, "learning_rate": 1.1604698622030577e-05, "loss": 0.3743, "step": 3251 }, { "epoch": 1.049719520561766, "grad_norm": 0.3338649545858964, "learning_rate": 1.160246405915661e-05, "loss": 0.2942, "step": 3252 }, { "epoch": 1.0500423745913878, "grad_norm": 0.39207185553283674, "learning_rate": 1.1600228976506095e-05, "loss": 0.3035, "step": 3253 }, { "epoch": 1.0503652286210097, "grad_norm": 0.4530071833565272, "learning_rate": 1.1597993374362217e-05, "loss": 0.2798, "step": 3254 }, { "epoch": 1.0506880826506315, "grad_norm": 0.4299187005878944, "learning_rate": 1.159575725300822e-05, "loss": 0.2534, "step": 3255 }, { "epoch": 1.0510109366802534, "grad_norm": 0.31721926388503463, "learning_rate": 1.1593520612727418e-05, "loss": 0.2967, "step": 3256 }, { "epoch": 1.0513337907098752, "grad_norm": 0.4041635742299768, "learning_rate": 1.1591283453803194e-05, "loss": 0.3294, "step": 3257 }, { "epoch": 1.051656644739497, "grad_norm": 0.38947823691663896, "learning_rate": 1.158904577651899e-05, "loss": 0.2989, "step": 3258 }, { "epoch": 1.0519794987691191, "grad_norm": 0.3698796828809157, "learning_rate": 1.1586807581158316e-05, "loss": 0.2348, "step": 3259 }, { "epoch": 1.052302352798741, "grad_norm": 0.7475615451699278, "learning_rate": 1.1584568868004751e-05, "loss": 0.3161, "step": 3260 }, { "epoch": 1.0526252068283628, "grad_norm": 0.4261405903499818, "learning_rate": 1.1582329637341934e-05, "loss": 0.2962, "step": 3261 }, { "epoch": 1.0529480608579846, "grad_norm": 0.5662928769030992, "learning_rate": 1.1580089889453575e-05, "loss": 0.2633, "step": 3262 }, { "epoch": 1.0532709148876065, "grad_norm": 0.4958468799377045, "learning_rate": 1.1577849624623444e-05, "loss": 0.274, "step": 3263 }, { "epoch": 1.0535937689172283, "grad_norm": 0.44939947640053024, "learning_rate": 1.157560884313538e-05, "loss": 0.3223, "step": 3264 }, { "epoch": 1.0539166229468502, "grad_norm": 0.35437909602780576, "learning_rate": 1.1573367545273287e-05, "loss": 0.2841, "step": 3265 }, { "epoch": 1.054239476976472, "grad_norm": 0.35643900822332164, "learning_rate": 1.157112573132114e-05, "loss": 0.2213, "step": 3266 }, { "epoch": 1.0545623310060939, "grad_norm": 0.37466258409427006, "learning_rate": 1.1568883401562966e-05, "loss": 0.2687, "step": 3267 }, { "epoch": 1.0548851850357157, "grad_norm": 0.505445157334223, "learning_rate": 1.1566640556282872e-05, "loss": 0.3177, "step": 3268 }, { "epoch": 1.0552080390653376, "grad_norm": 0.45162743999388855, "learning_rate": 1.1564397195765015e-05, "loss": 0.2936, "step": 3269 }, { "epoch": 1.0555308930949594, "grad_norm": 0.4094991848845644, "learning_rate": 1.1562153320293634e-05, "loss": 0.2551, "step": 3270 }, { "epoch": 1.0558537471245812, "grad_norm": 0.4430466478374135, "learning_rate": 1.155990893015302e-05, "loss": 0.2624, "step": 3271 }, { "epoch": 1.056176601154203, "grad_norm": 0.38811028036837236, "learning_rate": 1.1557664025627533e-05, "loss": 0.3742, "step": 3272 }, { "epoch": 1.056499455183825, "grad_norm": 0.45303714569609677, "learning_rate": 1.1555418607001603e-05, "loss": 0.3263, "step": 3273 }, { "epoch": 1.0568223092134468, "grad_norm": 0.4682045747129746, "learning_rate": 1.155317267455972e-05, "loss": 0.3044, "step": 3274 }, { "epoch": 1.0571451632430686, "grad_norm": 0.42464359880568786, "learning_rate": 1.1550926228586443e-05, "loss": 0.2875, "step": 3275 }, { "epoch": 1.0574680172726907, "grad_norm": 0.4716240553666264, "learning_rate": 1.1548679269366388e-05, "loss": 0.3401, "step": 3276 }, { "epoch": 1.0577908713023125, "grad_norm": 0.44699467477922994, "learning_rate": 1.1546431797184248e-05, "loss": 0.3376, "step": 3277 }, { "epoch": 1.0581137253319344, "grad_norm": 0.3975815382026158, "learning_rate": 1.1544183812324768e-05, "loss": 0.3099, "step": 3278 }, { "epoch": 1.0584365793615562, "grad_norm": 0.3709314027607587, "learning_rate": 1.1541935315072774e-05, "loss": 0.3437, "step": 3279 }, { "epoch": 1.058759433391178, "grad_norm": 0.4167717205690876, "learning_rate": 1.1539686305713136e-05, "loss": 0.2957, "step": 3280 }, { "epoch": 1.0590822874208, "grad_norm": 0.35693536889654937, "learning_rate": 1.1537436784530807e-05, "loss": 0.2512, "step": 3281 }, { "epoch": 1.0594051414504218, "grad_norm": 0.5430325851095799, "learning_rate": 1.1535186751810799e-05, "loss": 0.2405, "step": 3282 }, { "epoch": 1.0597279954800436, "grad_norm": 0.42146316340273415, "learning_rate": 1.1532936207838186e-05, "loss": 0.2766, "step": 3283 }, { "epoch": 1.0600508495096654, "grad_norm": 0.357654611228382, "learning_rate": 1.1530685152898109e-05, "loss": 0.314, "step": 3284 }, { "epoch": 1.0603737035392873, "grad_norm": 0.41763430101468557, "learning_rate": 1.1528433587275772e-05, "loss": 0.2671, "step": 3285 }, { "epoch": 1.0606965575689091, "grad_norm": 0.4573411028087847, "learning_rate": 1.152618151125645e-05, "loss": 0.355, "step": 3286 }, { "epoch": 1.061019411598531, "grad_norm": 0.36454268311702853, "learning_rate": 1.1523928925125473e-05, "loss": 0.3172, "step": 3287 }, { "epoch": 1.0613422656281528, "grad_norm": 0.3914664447674399, "learning_rate": 1.1521675829168242e-05, "loss": 0.2981, "step": 3288 }, { "epoch": 1.0616651196577747, "grad_norm": 0.4643107791856813, "learning_rate": 1.1519422223670224e-05, "loss": 0.3106, "step": 3289 }, { "epoch": 1.0619879736873965, "grad_norm": 0.38871167327858264, "learning_rate": 1.1517168108916942e-05, "loss": 0.2809, "step": 3290 }, { "epoch": 1.0623108277170183, "grad_norm": 0.39860332752104316, "learning_rate": 1.1514913485193995e-05, "loss": 0.2994, "step": 3291 }, { "epoch": 1.0626336817466404, "grad_norm": 0.4876383363104702, "learning_rate": 1.1512658352787038e-05, "loss": 0.2926, "step": 3292 }, { "epoch": 1.0629565357762623, "grad_norm": 0.560332374561345, "learning_rate": 1.1510402711981793e-05, "loss": 0.2818, "step": 3293 }, { "epoch": 1.063279389805884, "grad_norm": 0.4432069972825009, "learning_rate": 1.1508146563064047e-05, "loss": 0.2871, "step": 3294 }, { "epoch": 1.063602243835506, "grad_norm": 0.45138331125581227, "learning_rate": 1.150588990631965e-05, "loss": 0.3012, "step": 3295 }, { "epoch": 1.0639250978651278, "grad_norm": 0.33085858721779265, "learning_rate": 1.150363274203452e-05, "loss": 0.2178, "step": 3296 }, { "epoch": 1.0642479518947496, "grad_norm": 0.3680506124515326, "learning_rate": 1.1501375070494635e-05, "loss": 0.3209, "step": 3297 }, { "epoch": 1.0645708059243715, "grad_norm": 0.5986449194412506, "learning_rate": 1.1499116891986038e-05, "loss": 0.2453, "step": 3298 }, { "epoch": 1.0648936599539933, "grad_norm": 0.3918688706517467, "learning_rate": 1.1496858206794839e-05, "loss": 0.3442, "step": 3299 }, { "epoch": 1.0652165139836152, "grad_norm": 0.4728658637883631, "learning_rate": 1.1494599015207207e-05, "loss": 0.3885, "step": 3300 }, { "epoch": 1.065539368013237, "grad_norm": 0.36236363547133255, "learning_rate": 1.1492339317509382e-05, "loss": 0.337, "step": 3301 }, { "epoch": 1.0658622220428589, "grad_norm": 0.38324089360841285, "learning_rate": 1.1490079113987661e-05, "loss": 0.3806, "step": 3302 }, { "epoch": 1.0661850760724807, "grad_norm": 0.4028680813922029, "learning_rate": 1.1487818404928415e-05, "loss": 0.3073, "step": 3303 }, { "epoch": 1.0665079301021025, "grad_norm": 0.4129188792569228, "learning_rate": 1.1485557190618066e-05, "loss": 0.2894, "step": 3304 }, { "epoch": 1.0668307841317244, "grad_norm": 0.4376022726880038, "learning_rate": 1.148329547134311e-05, "loss": 0.3433, "step": 3305 }, { "epoch": 1.0671536381613462, "grad_norm": 0.7005573568315073, "learning_rate": 1.1481033247390103e-05, "loss": 0.2841, "step": 3306 }, { "epoch": 1.067476492190968, "grad_norm": 0.4022029521164865, "learning_rate": 1.1478770519045668e-05, "loss": 0.3438, "step": 3307 }, { "epoch": 1.06779934622059, "grad_norm": 0.4458197862663426, "learning_rate": 1.1476507286596485e-05, "loss": 0.3079, "step": 3308 }, { "epoch": 1.0681222002502118, "grad_norm": 0.44142730208670905, "learning_rate": 1.1474243550329307e-05, "loss": 0.2604, "step": 3309 }, { "epoch": 1.0684450542798338, "grad_norm": 0.45695802220336557, "learning_rate": 1.1471979310530939e-05, "loss": 0.3694, "step": 3310 }, { "epoch": 1.0687679083094557, "grad_norm": 0.38495605296406205, "learning_rate": 1.1469714567488266e-05, "loss": 0.2609, "step": 3311 }, { "epoch": 1.0690907623390775, "grad_norm": 0.5017987453401724, "learning_rate": 1.1467449321488223e-05, "loss": 0.3124, "step": 3312 }, { "epoch": 1.0694136163686994, "grad_norm": 0.5355313738006615, "learning_rate": 1.1465183572817815e-05, "loss": 0.2898, "step": 3313 }, { "epoch": 1.0697364703983212, "grad_norm": 0.4551416459156414, "learning_rate": 1.146291732176411e-05, "loss": 0.3044, "step": 3314 }, { "epoch": 1.070059324427943, "grad_norm": 0.4897181886211879, "learning_rate": 1.1460650568614234e-05, "loss": 0.2769, "step": 3315 }, { "epoch": 1.070382178457565, "grad_norm": 0.5877825227972763, "learning_rate": 1.1458383313655388e-05, "loss": 0.3463, "step": 3316 }, { "epoch": 1.0707050324871867, "grad_norm": 0.484762601333966, "learning_rate": 1.1456115557174825e-05, "loss": 0.3421, "step": 3317 }, { "epoch": 1.0710278865168086, "grad_norm": 0.3835795928893714, "learning_rate": 1.1453847299459868e-05, "loss": 0.3066, "step": 3318 }, { "epoch": 1.0713507405464304, "grad_norm": 0.36427900082676545, "learning_rate": 1.1451578540797906e-05, "loss": 0.288, "step": 3319 }, { "epoch": 1.0716735945760523, "grad_norm": 0.38892487331847475, "learning_rate": 1.1449309281476384e-05, "loss": 0.3614, "step": 3320 }, { "epoch": 1.0719964486056741, "grad_norm": 0.4037015049026727, "learning_rate": 1.1447039521782815e-05, "loss": 0.3225, "step": 3321 }, { "epoch": 1.072319302635296, "grad_norm": 0.35962844903924807, "learning_rate": 1.1444769262004771e-05, "loss": 0.2664, "step": 3322 }, { "epoch": 1.0726421566649178, "grad_norm": 0.38046740926971356, "learning_rate": 1.1442498502429897e-05, "loss": 0.3124, "step": 3323 }, { "epoch": 1.0729650106945396, "grad_norm": 0.3909849606212096, "learning_rate": 1.144022724334589e-05, "loss": 0.3496, "step": 3324 }, { "epoch": 1.0732878647241615, "grad_norm": 0.393416190298201, "learning_rate": 1.143795548504052e-05, "loss": 0.3453, "step": 3325 }, { "epoch": 1.0736107187537836, "grad_norm": 0.38510519351242767, "learning_rate": 1.143568322780161e-05, "loss": 0.2977, "step": 3326 }, { "epoch": 1.0739335727834054, "grad_norm": 0.41524300641048056, "learning_rate": 1.1433410471917055e-05, "loss": 0.3128, "step": 3327 }, { "epoch": 1.0742564268130272, "grad_norm": 0.36208985603628463, "learning_rate": 1.143113721767481e-05, "loss": 0.3098, "step": 3328 }, { "epoch": 1.074579280842649, "grad_norm": 0.36683741656061486, "learning_rate": 1.1428863465362895e-05, "loss": 0.3025, "step": 3329 }, { "epoch": 1.074902134872271, "grad_norm": 0.4313546061082764, "learning_rate": 1.1426589215269386e-05, "loss": 0.3005, "step": 3330 }, { "epoch": 1.0752249889018928, "grad_norm": 0.4173703385975774, "learning_rate": 1.1424314467682435e-05, "loss": 0.2655, "step": 3331 }, { "epoch": 1.0755478429315146, "grad_norm": 0.38080140414741437, "learning_rate": 1.1422039222890242e-05, "loss": 0.2938, "step": 3332 }, { "epoch": 1.0758706969611365, "grad_norm": 0.43280442088038246, "learning_rate": 1.141976348118108e-05, "loss": 0.3253, "step": 3333 }, { "epoch": 1.0761935509907583, "grad_norm": 0.3799834360311146, "learning_rate": 1.1417487242843284e-05, "loss": 0.2744, "step": 3334 }, { "epoch": 1.0765164050203802, "grad_norm": 0.3835804465773085, "learning_rate": 1.1415210508165248e-05, "loss": 0.3355, "step": 3335 }, { "epoch": 1.076839259050002, "grad_norm": 0.4760700361596323, "learning_rate": 1.1412933277435434e-05, "loss": 0.2978, "step": 3336 }, { "epoch": 1.0771621130796238, "grad_norm": 0.581844808857931, "learning_rate": 1.141065555094236e-05, "loss": 0.2957, "step": 3337 }, { "epoch": 1.0774849671092457, "grad_norm": 0.3868051589298461, "learning_rate": 1.1408377328974615e-05, "loss": 0.241, "step": 3338 }, { "epoch": 1.0778078211388675, "grad_norm": 0.4784888068364716, "learning_rate": 1.1406098611820843e-05, "loss": 0.3091, "step": 3339 }, { "epoch": 1.0781306751684894, "grad_norm": 0.39080049609135925, "learning_rate": 1.1403819399769755e-05, "loss": 0.3064, "step": 3340 }, { "epoch": 1.0784535291981112, "grad_norm": 0.3932302455210032, "learning_rate": 1.1401539693110128e-05, "loss": 0.3379, "step": 3341 }, { "epoch": 1.078776383227733, "grad_norm": 0.4124405161486858, "learning_rate": 1.1399259492130793e-05, "loss": 0.2996, "step": 3342 }, { "epoch": 1.079099237257355, "grad_norm": 0.5559862861079145, "learning_rate": 1.1396978797120651e-05, "loss": 0.2796, "step": 3343 }, { "epoch": 1.079422091286977, "grad_norm": 0.41750070901521114, "learning_rate": 1.139469760836866e-05, "loss": 0.3146, "step": 3344 }, { "epoch": 1.0797449453165988, "grad_norm": 0.39021912247770074, "learning_rate": 1.1392415926163847e-05, "loss": 0.3294, "step": 3345 }, { "epoch": 1.0800677993462207, "grad_norm": 0.458302287773993, "learning_rate": 1.1390133750795296e-05, "loss": 0.3195, "step": 3346 }, { "epoch": 1.0803906533758425, "grad_norm": 0.3776257638953978, "learning_rate": 1.1387851082552155e-05, "loss": 0.3049, "step": 3347 }, { "epoch": 1.0807135074054643, "grad_norm": 0.4478906401095702, "learning_rate": 1.1385567921723638e-05, "loss": 0.3062, "step": 3348 }, { "epoch": 1.0810363614350862, "grad_norm": 0.372107696721039, "learning_rate": 1.1383284268599012e-05, "loss": 0.3076, "step": 3349 }, { "epoch": 1.081359215464708, "grad_norm": 0.39055765427471095, "learning_rate": 1.1381000123467619e-05, "loss": 0.2894, "step": 3350 }, { "epoch": 1.0816820694943299, "grad_norm": 0.4079823830202834, "learning_rate": 1.1378715486618855e-05, "loss": 0.3429, "step": 3351 }, { "epoch": 1.0820049235239517, "grad_norm": 0.41886490142528493, "learning_rate": 1.1376430358342181e-05, "loss": 0.314, "step": 3352 }, { "epoch": 1.0823277775535736, "grad_norm": 0.3132551554898045, "learning_rate": 1.1374144738927117e-05, "loss": 0.2875, "step": 3353 }, { "epoch": 1.0826506315831954, "grad_norm": 0.3965954079971687, "learning_rate": 1.137185862866325e-05, "loss": 0.3277, "step": 3354 }, { "epoch": 1.0829734856128173, "grad_norm": 0.37338408354277103, "learning_rate": 1.1369572027840225e-05, "loss": 0.2779, "step": 3355 }, { "epoch": 1.083296339642439, "grad_norm": 0.46481679659377045, "learning_rate": 1.1367284936747756e-05, "loss": 0.2652, "step": 3356 }, { "epoch": 1.083619193672061, "grad_norm": 0.34599305423280935, "learning_rate": 1.1364997355675608e-05, "loss": 0.3284, "step": 3357 }, { "epoch": 1.0839420477016828, "grad_norm": 0.4060431722030522, "learning_rate": 1.136270928491362e-05, "loss": 0.3104, "step": 3358 }, { "epoch": 1.0842649017313046, "grad_norm": 0.46234764286693014, "learning_rate": 1.1360420724751685e-05, "loss": 0.3436, "step": 3359 }, { "epoch": 1.0845877557609267, "grad_norm": 0.38128022577482434, "learning_rate": 1.135813167547976e-05, "loss": 0.3448, "step": 3360 }, { "epoch": 1.0849106097905485, "grad_norm": 0.3779352920377369, "learning_rate": 1.1355842137387865e-05, "loss": 0.3059, "step": 3361 }, { "epoch": 1.0852334638201704, "grad_norm": 0.35775890039362834, "learning_rate": 1.135355211076608e-05, "loss": 0.2817, "step": 3362 }, { "epoch": 1.0855563178497922, "grad_norm": 0.4504061148765678, "learning_rate": 1.1351261595904549e-05, "loss": 0.319, "step": 3363 }, { "epoch": 1.085879171879414, "grad_norm": 0.5681343078793912, "learning_rate": 1.1348970593093476e-05, "loss": 0.2895, "step": 3364 }, { "epoch": 1.086202025909036, "grad_norm": 0.48629906118929267, "learning_rate": 1.1346679102623132e-05, "loss": 0.3387, "step": 3365 }, { "epoch": 1.0865248799386578, "grad_norm": 0.44007319210334006, "learning_rate": 1.1344387124783842e-05, "loss": 0.2651, "step": 3366 }, { "epoch": 1.0868477339682796, "grad_norm": 0.36964094053891383, "learning_rate": 1.1342094659865996e-05, "loss": 0.3282, "step": 3367 }, { "epoch": 1.0871705879979014, "grad_norm": 0.43479701916925906, "learning_rate": 1.133980170816005e-05, "loss": 0.3, "step": 3368 }, { "epoch": 1.0874934420275233, "grad_norm": 0.4618125968210604, "learning_rate": 1.1337508269956513e-05, "loss": 0.3405, "step": 3369 }, { "epoch": 1.0878162960571451, "grad_norm": 0.48492062051302925, "learning_rate": 1.1335214345545964e-05, "loss": 0.306, "step": 3370 }, { "epoch": 1.088139150086767, "grad_norm": 0.5235121373946282, "learning_rate": 1.133291993521904e-05, "loss": 0.2904, "step": 3371 }, { "epoch": 1.0884620041163888, "grad_norm": 0.47360155832558415, "learning_rate": 1.1330625039266436e-05, "loss": 0.3563, "step": 3372 }, { "epoch": 1.0887848581460107, "grad_norm": 0.49696161478001766, "learning_rate": 1.1328329657978914e-05, "loss": 0.3288, "step": 3373 }, { "epoch": 1.0891077121756325, "grad_norm": 0.35650771296599176, "learning_rate": 1.1326033791647297e-05, "loss": 0.287, "step": 3374 }, { "epoch": 1.0894305662052544, "grad_norm": 0.4650098507904134, "learning_rate": 1.1323737440562469e-05, "loss": 0.3238, "step": 3375 }, { "epoch": 1.0897534202348762, "grad_norm": 0.4476153697342474, "learning_rate": 1.1321440605015372e-05, "loss": 0.2961, "step": 3376 }, { "epoch": 1.090076274264498, "grad_norm": 0.3874507170001616, "learning_rate": 1.1319143285297013e-05, "loss": 0.2754, "step": 3377 }, { "epoch": 1.0903991282941201, "grad_norm": 0.5405619414815762, "learning_rate": 1.1316845481698458e-05, "loss": 0.3352, "step": 3378 }, { "epoch": 1.090721982323742, "grad_norm": 0.49185697847088333, "learning_rate": 1.1314547194510839e-05, "loss": 0.2991, "step": 3379 }, { "epoch": 1.0910448363533638, "grad_norm": 0.454144454142604, "learning_rate": 1.1312248424025342e-05, "loss": 0.306, "step": 3380 }, { "epoch": 1.0913676903829856, "grad_norm": 0.4061380237205261, "learning_rate": 1.1309949170533218e-05, "loss": 0.2902, "step": 3381 }, { "epoch": 1.0916905444126075, "grad_norm": 0.4260349407476014, "learning_rate": 1.130764943432578e-05, "loss": 0.2973, "step": 3382 }, { "epoch": 1.0920133984422293, "grad_norm": 0.3846376062143684, "learning_rate": 1.1305349215694406e-05, "loss": 0.2991, "step": 3383 }, { "epoch": 1.0923362524718512, "grad_norm": 0.39833893776802715, "learning_rate": 1.1303048514930524e-05, "loss": 0.3549, "step": 3384 }, { "epoch": 1.092659106501473, "grad_norm": 0.4368251384216377, "learning_rate": 1.1300747332325635e-05, "loss": 0.3096, "step": 3385 }, { "epoch": 1.0929819605310949, "grad_norm": 0.39426458628687605, "learning_rate": 1.129844566817129e-05, "loss": 0.2363, "step": 3386 }, { "epoch": 1.0933048145607167, "grad_norm": 0.33916188614238235, "learning_rate": 1.129614352275911e-05, "loss": 0.3048, "step": 3387 }, { "epoch": 1.0936276685903386, "grad_norm": 0.3830317558789297, "learning_rate": 1.1293840896380773e-05, "loss": 0.3385, "step": 3388 }, { "epoch": 1.0939505226199604, "grad_norm": 0.39898115326440553, "learning_rate": 1.129153778932802e-05, "loss": 0.2525, "step": 3389 }, { "epoch": 1.0942733766495822, "grad_norm": 0.37175748541141795, "learning_rate": 1.1289234201892651e-05, "loss": 0.3162, "step": 3390 }, { "epoch": 1.094596230679204, "grad_norm": 0.4568480958888799, "learning_rate": 1.1286930134366524e-05, "loss": 0.3525, "step": 3391 }, { "epoch": 1.094919084708826, "grad_norm": 0.418832891913155, "learning_rate": 1.1284625587041567e-05, "loss": 0.3224, "step": 3392 }, { "epoch": 1.095241938738448, "grad_norm": 0.3981906311974513, "learning_rate": 1.1282320560209756e-05, "loss": 0.3219, "step": 3393 }, { "epoch": 1.0955647927680698, "grad_norm": 0.4088813315794091, "learning_rate": 1.1280015054163144e-05, "loss": 0.3224, "step": 3394 }, { "epoch": 1.0958876467976917, "grad_norm": 0.4310526550149113, "learning_rate": 1.1277709069193827e-05, "loss": 0.2792, "step": 3395 }, { "epoch": 1.0962105008273135, "grad_norm": 0.3895997876496565, "learning_rate": 1.127540260559397e-05, "loss": 0.2937, "step": 3396 }, { "epoch": 1.0965333548569354, "grad_norm": 0.42413386909161654, "learning_rate": 1.1273095663655807e-05, "loss": 0.3603, "step": 3397 }, { "epoch": 1.0968562088865572, "grad_norm": 0.3968134937515018, "learning_rate": 1.1270788243671615e-05, "loss": 0.3506, "step": 3398 }, { "epoch": 1.097179062916179, "grad_norm": 0.4013620447188084, "learning_rate": 1.126848034593375e-05, "loss": 0.2759, "step": 3399 }, { "epoch": 1.097501916945801, "grad_norm": 0.41854431948073917, "learning_rate": 1.1266171970734613e-05, "loss": 0.3196, "step": 3400 }, { "epoch": 1.0978247709754227, "grad_norm": 0.468272796025812, "learning_rate": 1.1263863118366672e-05, "loss": 0.3046, "step": 3401 }, { "epoch": 1.0981476250050446, "grad_norm": 0.4124350367925084, "learning_rate": 1.1261553789122459e-05, "loss": 0.2949, "step": 3402 }, { "epoch": 1.0984704790346664, "grad_norm": 0.45705545496202654, "learning_rate": 1.1259243983294558e-05, "loss": 0.2915, "step": 3403 }, { "epoch": 1.0987933330642883, "grad_norm": 0.3932796034726098, "learning_rate": 1.1256933701175625e-05, "loss": 0.3206, "step": 3404 }, { "epoch": 1.0991161870939101, "grad_norm": 0.42739655754203665, "learning_rate": 1.1254622943058363e-05, "loss": 0.2808, "step": 3405 }, { "epoch": 1.099439041123532, "grad_norm": 0.45803225685048093, "learning_rate": 1.1252311709235544e-05, "loss": 0.3361, "step": 3406 }, { "epoch": 1.0997618951531538, "grad_norm": 0.44962660128979554, "learning_rate": 1.125e-05, "loss": 0.3161, "step": 3407 }, { "epoch": 1.1000847491827757, "grad_norm": 0.3689771426831328, "learning_rate": 1.124768781564462e-05, "loss": 0.3268, "step": 3408 }, { "epoch": 1.1004076032123975, "grad_norm": 0.38691767429404866, "learning_rate": 1.1245375156462356e-05, "loss": 0.3552, "step": 3409 }, { "epoch": 1.1007304572420193, "grad_norm": 0.479031641828172, "learning_rate": 1.1243062022746216e-05, "loss": 0.2534, "step": 3410 }, { "epoch": 1.1010533112716414, "grad_norm": 0.4455811685549658, "learning_rate": 1.124074841478927e-05, "loss": 0.2532, "step": 3411 }, { "epoch": 1.1013761653012633, "grad_norm": 0.38673615766121183, "learning_rate": 1.1238434332884653e-05, "loss": 0.3201, "step": 3412 }, { "epoch": 1.101699019330885, "grad_norm": 0.5621012402363115, "learning_rate": 1.1236119777325555e-05, "loss": 0.2849, "step": 3413 }, { "epoch": 1.102021873360507, "grad_norm": 0.46602241529201216, "learning_rate": 1.1233804748405227e-05, "loss": 0.3142, "step": 3414 }, { "epoch": 1.1023447273901288, "grad_norm": 0.4471418257333207, "learning_rate": 1.1231489246416976e-05, "loss": 0.2962, "step": 3415 }, { "epoch": 1.1026675814197506, "grad_norm": 0.4600824748837668, "learning_rate": 1.122917327165418e-05, "loss": 0.2852, "step": 3416 }, { "epoch": 1.1029904354493725, "grad_norm": 0.4340663683565431, "learning_rate": 1.1226856824410261e-05, "loss": 0.2949, "step": 3417 }, { "epoch": 1.1033132894789943, "grad_norm": 0.5054434733643556, "learning_rate": 1.1224539904978718e-05, "loss": 0.3185, "step": 3418 }, { "epoch": 1.1036361435086162, "grad_norm": 0.39146338963916094, "learning_rate": 1.1222222513653096e-05, "loss": 0.3248, "step": 3419 }, { "epoch": 1.103958997538238, "grad_norm": 0.5169466975153844, "learning_rate": 1.121990465072701e-05, "loss": 0.2623, "step": 3420 }, { "epoch": 1.1042818515678599, "grad_norm": 0.43007833958549324, "learning_rate": 1.1217586316494126e-05, "loss": 0.2797, "step": 3421 }, { "epoch": 1.1046047055974817, "grad_norm": 0.4327765014248168, "learning_rate": 1.1215267511248176e-05, "loss": 0.2827, "step": 3422 }, { "epoch": 1.1049275596271035, "grad_norm": 0.3943935423482067, "learning_rate": 1.1212948235282946e-05, "loss": 0.303, "step": 3423 }, { "epoch": 1.1052504136567254, "grad_norm": 0.39612293562766226, "learning_rate": 1.121062848889229e-05, "loss": 0.2932, "step": 3424 }, { "epoch": 1.1055732676863472, "grad_norm": 0.4526233619661319, "learning_rate": 1.1208308272370115e-05, "loss": 0.3666, "step": 3425 }, { "epoch": 1.105896121715969, "grad_norm": 0.34765833762252324, "learning_rate": 1.1205987586010385e-05, "loss": 0.2573, "step": 3426 }, { "epoch": 1.1062189757455911, "grad_norm": 0.48458684365896276, "learning_rate": 1.1203666430107133e-05, "loss": 0.3454, "step": 3427 }, { "epoch": 1.106541829775213, "grad_norm": 0.455534063086837, "learning_rate": 1.1201344804954445e-05, "loss": 0.2657, "step": 3428 }, { "epoch": 1.1068646838048348, "grad_norm": 0.37360171173704926, "learning_rate": 1.1199022710846466e-05, "loss": 0.3583, "step": 3429 }, { "epoch": 1.1071875378344567, "grad_norm": 0.3807032154780368, "learning_rate": 1.1196700148077404e-05, "loss": 0.3087, "step": 3430 }, { "epoch": 1.1075103918640785, "grad_norm": 0.40882507075026175, "learning_rate": 1.1194377116941526e-05, "loss": 0.3445, "step": 3431 }, { "epoch": 1.1078332458937004, "grad_norm": 0.3562521081113106, "learning_rate": 1.1192053617733153e-05, "loss": 0.3007, "step": 3432 }, { "epoch": 1.1081560999233222, "grad_norm": 0.49142495758964055, "learning_rate": 1.1189729650746672e-05, "loss": 0.3046, "step": 3433 }, { "epoch": 1.108478953952944, "grad_norm": 0.41142516673657875, "learning_rate": 1.1187405216276523e-05, "loss": 0.3195, "step": 3434 }, { "epoch": 1.108801807982566, "grad_norm": 0.39758236119087087, "learning_rate": 1.1185080314617213e-05, "loss": 0.2424, "step": 3435 }, { "epoch": 1.1091246620121877, "grad_norm": 0.4063096705870214, "learning_rate": 1.1182754946063301e-05, "loss": 0.2724, "step": 3436 }, { "epoch": 1.1094475160418096, "grad_norm": 0.38506812129230944, "learning_rate": 1.1180429110909408e-05, "loss": 0.3363, "step": 3437 }, { "epoch": 1.1097703700714314, "grad_norm": 0.3827059242954524, "learning_rate": 1.1178102809450218e-05, "loss": 0.3186, "step": 3438 }, { "epoch": 1.1100932241010533, "grad_norm": 0.4347540638873324, "learning_rate": 1.1175776041980464e-05, "loss": 0.3175, "step": 3439 }, { "epoch": 1.110416078130675, "grad_norm": 0.3923123162242907, "learning_rate": 1.117344880879495e-05, "loss": 0.2652, "step": 3440 }, { "epoch": 1.110738932160297, "grad_norm": 0.44626130923885526, "learning_rate": 1.1171121110188527e-05, "loss": 0.3161, "step": 3441 }, { "epoch": 1.1110617861899188, "grad_norm": 0.4704806696902176, "learning_rate": 1.1168792946456116e-05, "loss": 0.2842, "step": 3442 }, { "epoch": 1.1113846402195406, "grad_norm": 0.4088079118123721, "learning_rate": 1.1166464317892692e-05, "loss": 0.2935, "step": 3443 }, { "epoch": 1.1117074942491625, "grad_norm": 0.37882188381754517, "learning_rate": 1.1164135224793287e-05, "loss": 0.3045, "step": 3444 }, { "epoch": 1.1120303482787846, "grad_norm": 0.43171239603614275, "learning_rate": 1.1161805667452995e-05, "loss": 0.2926, "step": 3445 }, { "epoch": 1.1123532023084064, "grad_norm": 0.4291264100523131, "learning_rate": 1.115947564616697e-05, "loss": 0.2722, "step": 3446 }, { "epoch": 1.1126760563380282, "grad_norm": 0.3937878020716653, "learning_rate": 1.1157145161230416e-05, "loss": 0.3126, "step": 3447 }, { "epoch": 1.11299891036765, "grad_norm": 0.4632498776388868, "learning_rate": 1.1154814212938608e-05, "loss": 0.2945, "step": 3448 }, { "epoch": 1.113321764397272, "grad_norm": 0.4128412297116511, "learning_rate": 1.1152482801586873e-05, "loss": 0.372, "step": 3449 }, { "epoch": 1.1136446184268938, "grad_norm": 0.4006447504172888, "learning_rate": 1.1150150927470596e-05, "loss": 0.3578, "step": 3450 }, { "epoch": 1.1139674724565156, "grad_norm": 0.4334742078229356, "learning_rate": 1.1147818590885222e-05, "loss": 0.3241, "step": 3451 }, { "epoch": 1.1142903264861375, "grad_norm": 0.4484486141442535, "learning_rate": 1.1145485792126258e-05, "loss": 0.2797, "step": 3452 }, { "epoch": 1.1146131805157593, "grad_norm": 0.4772730937396866, "learning_rate": 1.1143152531489263e-05, "loss": 0.2864, "step": 3453 }, { "epoch": 1.1149360345453811, "grad_norm": 0.4025993366954776, "learning_rate": 1.114081880926986e-05, "loss": 0.2639, "step": 3454 }, { "epoch": 1.115258888575003, "grad_norm": 0.4481891925982167, "learning_rate": 1.1138484625763726e-05, "loss": 0.3225, "step": 3455 }, { "epoch": 1.1155817426046248, "grad_norm": 0.38402801217574895, "learning_rate": 1.1136149981266603e-05, "loss": 0.2464, "step": 3456 }, { "epoch": 1.1159045966342467, "grad_norm": 0.41165533488857997, "learning_rate": 1.1133814876074285e-05, "loss": 0.2963, "step": 3457 }, { "epoch": 1.1162274506638685, "grad_norm": 0.42267029965530595, "learning_rate": 1.1131479310482621e-05, "loss": 0.2728, "step": 3458 }, { "epoch": 1.1165503046934904, "grad_norm": 0.4264784316459077, "learning_rate": 1.1129143284787533e-05, "loss": 0.3024, "step": 3459 }, { "epoch": 1.1168731587231122, "grad_norm": 0.4785117711355052, "learning_rate": 1.1126806799284988e-05, "loss": 0.3239, "step": 3460 }, { "epoch": 1.1171960127527343, "grad_norm": 0.39770060448812705, "learning_rate": 1.1124469854271018e-05, "loss": 0.2707, "step": 3461 }, { "epoch": 1.1175188667823561, "grad_norm": 0.5216746206160969, "learning_rate": 1.1122132450041706e-05, "loss": 0.3114, "step": 3462 }, { "epoch": 1.117841720811978, "grad_norm": 0.41747196837094735, "learning_rate": 1.11197945868932e-05, "loss": 0.2898, "step": 3463 }, { "epoch": 1.1181645748415998, "grad_norm": 0.44451208193500785, "learning_rate": 1.1117456265121708e-05, "loss": 0.273, "step": 3464 }, { "epoch": 1.1184874288712217, "grad_norm": 0.43213823448996697, "learning_rate": 1.1115117485023484e-05, "loss": 0.3361, "step": 3465 }, { "epoch": 1.1188102829008435, "grad_norm": 0.38898793580546986, "learning_rate": 1.1112778246894857e-05, "loss": 0.2954, "step": 3466 }, { "epoch": 1.1191331369304653, "grad_norm": 0.33866620449769014, "learning_rate": 1.1110438551032198e-05, "loss": 0.3097, "step": 3467 }, { "epoch": 1.1194559909600872, "grad_norm": 0.3849843070292859, "learning_rate": 1.1108098397731948e-05, "loss": 0.2985, "step": 3468 }, { "epoch": 1.119778844989709, "grad_norm": 0.38864979631933844, "learning_rate": 1.11057577872906e-05, "loss": 0.2339, "step": 3469 }, { "epoch": 1.1201016990193309, "grad_norm": 0.4140154499309351, "learning_rate": 1.1103416720004707e-05, "loss": 0.3151, "step": 3470 }, { "epoch": 1.1204245530489527, "grad_norm": 0.43791916705768263, "learning_rate": 1.1101075196170875e-05, "loss": 0.3387, "step": 3471 }, { "epoch": 1.1207474070785746, "grad_norm": 0.3416401860519256, "learning_rate": 1.1098733216085777e-05, "loss": 0.294, "step": 3472 }, { "epoch": 1.1210702611081964, "grad_norm": 0.4128214430714466, "learning_rate": 1.1096390780046136e-05, "loss": 0.3042, "step": 3473 }, { "epoch": 1.1213931151378183, "grad_norm": 0.4527881544220701, "learning_rate": 1.1094047888348734e-05, "loss": 0.3748, "step": 3474 }, { "epoch": 1.12171596916744, "grad_norm": 0.383418777421038, "learning_rate": 1.1091704541290415e-05, "loss": 0.2683, "step": 3475 }, { "epoch": 1.122038823197062, "grad_norm": 0.37835220619280324, "learning_rate": 1.1089360739168075e-05, "loss": 0.274, "step": 3476 }, { "epoch": 1.1223616772266838, "grad_norm": 0.394463099315158, "learning_rate": 1.1087016482278676e-05, "loss": 0.3123, "step": 3477 }, { "epoch": 1.1226845312563056, "grad_norm": 0.35443030801831554, "learning_rate": 1.1084671770919225e-05, "loss": 0.3618, "step": 3478 }, { "epoch": 1.1230073852859277, "grad_norm": 0.3641544627450112, "learning_rate": 1.1082326605386797e-05, "loss": 0.3554, "step": 3479 }, { "epoch": 1.1233302393155495, "grad_norm": 0.3900485754313955, "learning_rate": 1.1079980985978524e-05, "loss": 0.3081, "step": 3480 }, { "epoch": 1.1236530933451714, "grad_norm": 0.6094549414497121, "learning_rate": 1.1077634912991588e-05, "loss": 0.305, "step": 3481 }, { "epoch": 1.1239759473747932, "grad_norm": 0.3486050601909293, "learning_rate": 1.1075288386723235e-05, "loss": 0.3075, "step": 3482 }, { "epoch": 1.124298801404415, "grad_norm": 0.4877868554151437, "learning_rate": 1.1072941407470768e-05, "loss": 0.268, "step": 3483 }, { "epoch": 1.124621655434037, "grad_norm": 0.40757625661637326, "learning_rate": 1.1070593975531541e-05, "loss": 0.3084, "step": 3484 }, { "epoch": 1.1249445094636588, "grad_norm": 0.3769710250485501, "learning_rate": 1.106824609120298e-05, "loss": 0.2326, "step": 3485 }, { "epoch": 1.1252673634932806, "grad_norm": 0.46544415188452054, "learning_rate": 1.106589775478255e-05, "loss": 0.3629, "step": 3486 }, { "epoch": 1.1255902175229024, "grad_norm": 0.4652325088742244, "learning_rate": 1.1063548966567785e-05, "loss": 0.3008, "step": 3487 }, { "epoch": 1.1259130715525243, "grad_norm": 0.49613920413899254, "learning_rate": 1.1061199726856274e-05, "loss": 0.292, "step": 3488 }, { "epoch": 1.1262359255821461, "grad_norm": 0.3959312447710953, "learning_rate": 1.1058850035945663e-05, "loss": 0.2683, "step": 3489 }, { "epoch": 1.126558779611768, "grad_norm": 0.4435095997238483, "learning_rate": 1.105649989413365e-05, "loss": 0.3312, "step": 3490 }, { "epoch": 1.1268816336413898, "grad_norm": 0.34201340903905275, "learning_rate": 1.1054149301718e-05, "loss": 0.3219, "step": 3491 }, { "epoch": 1.1272044876710117, "grad_norm": 0.39822530619261687, "learning_rate": 1.1051798258996528e-05, "loss": 0.3005, "step": 3492 }, { "epoch": 1.1275273417006335, "grad_norm": 0.40306198184564374, "learning_rate": 1.1049446766267108e-05, "loss": 0.3261, "step": 3493 }, { "epoch": 1.1278501957302556, "grad_norm": 0.38141748345056486, "learning_rate": 1.104709482382767e-05, "loss": 0.315, "step": 3494 }, { "epoch": 1.1281730497598774, "grad_norm": 0.381553347345685, "learning_rate": 1.1044742431976203e-05, "loss": 0.3098, "step": 3495 }, { "epoch": 1.1284959037894993, "grad_norm": 0.5334530102265275, "learning_rate": 1.104238959101075e-05, "loss": 0.3186, "step": 3496 }, { "epoch": 1.128818757819121, "grad_norm": 0.3979462589579792, "learning_rate": 1.1040036301229412e-05, "loss": 0.3345, "step": 3497 }, { "epoch": 1.129141611848743, "grad_norm": 0.44979451423041195, "learning_rate": 1.1037682562930353e-05, "loss": 0.3088, "step": 3498 }, { "epoch": 1.1294644658783648, "grad_norm": 0.4595377244813872, "learning_rate": 1.1035328376411785e-05, "loss": 0.3504, "step": 3499 }, { "epoch": 1.1297873199079866, "grad_norm": 0.4065259883519845, "learning_rate": 1.1032973741971978e-05, "loss": 0.2609, "step": 3500 }, { "epoch": 1.1301101739376085, "grad_norm": 0.4623499952516531, "learning_rate": 1.1030618659909266e-05, "loss": 0.2814, "step": 3501 }, { "epoch": 1.1304330279672303, "grad_norm": 0.4102341576794453, "learning_rate": 1.1028263130522029e-05, "loss": 0.3506, "step": 3502 }, { "epoch": 1.1307558819968522, "grad_norm": 0.4169327710583793, "learning_rate": 1.1025907154108713e-05, "loss": 0.3668, "step": 3503 }, { "epoch": 1.131078736026474, "grad_norm": 0.40150035423036123, "learning_rate": 1.1023550730967816e-05, "loss": 0.3065, "step": 3504 }, { "epoch": 1.1314015900560959, "grad_norm": 0.37265003220991716, "learning_rate": 1.1021193861397896e-05, "loss": 0.3084, "step": 3505 }, { "epoch": 1.1317244440857177, "grad_norm": 0.5226406429671188, "learning_rate": 1.1018836545697558e-05, "loss": 0.2669, "step": 3506 }, { "epoch": 1.1320472981153396, "grad_norm": 0.4279257332760106, "learning_rate": 1.101647878416548e-05, "loss": 0.3538, "step": 3507 }, { "epoch": 1.1323701521449614, "grad_norm": 0.3666763910486776, "learning_rate": 1.1014120577100379e-05, "loss": 0.2952, "step": 3508 }, { "epoch": 1.1326930061745832, "grad_norm": 0.3600159910228635, "learning_rate": 1.1011761924801041e-05, "loss": 0.2722, "step": 3509 }, { "epoch": 1.133015860204205, "grad_norm": 0.3247642898099795, "learning_rate": 1.1009402827566304e-05, "loss": 0.3061, "step": 3510 }, { "epoch": 1.133338714233827, "grad_norm": 0.42443409335239585, "learning_rate": 1.1007043285695063e-05, "loss": 0.287, "step": 3511 }, { "epoch": 1.1336615682634488, "grad_norm": 0.44430780425333377, "learning_rate": 1.1004683299486264e-05, "loss": 0.3057, "step": 3512 }, { "epoch": 1.1339844222930708, "grad_norm": 0.45564975871071045, "learning_rate": 1.1002322869238917e-05, "loss": 0.2549, "step": 3513 }, { "epoch": 1.1343072763226927, "grad_norm": 0.4222920426568446, "learning_rate": 1.0999961995252088e-05, "loss": 0.3381, "step": 3514 }, { "epoch": 1.1346301303523145, "grad_norm": 0.43334507682899753, "learning_rate": 1.0997600677824893e-05, "loss": 0.3106, "step": 3515 }, { "epoch": 1.1349529843819364, "grad_norm": 0.4382704527716008, "learning_rate": 1.0995238917256508e-05, "loss": 0.3428, "step": 3516 }, { "epoch": 1.1352758384115582, "grad_norm": 0.3915896374667341, "learning_rate": 1.0992876713846166e-05, "loss": 0.3376, "step": 3517 }, { "epoch": 1.13559869244118, "grad_norm": 0.6346454552037837, "learning_rate": 1.0990514067893156e-05, "loss": 0.3333, "step": 3518 }, { "epoch": 1.135921546470802, "grad_norm": 0.4494160880981193, "learning_rate": 1.0988150979696819e-05, "loss": 0.3114, "step": 3519 }, { "epoch": 1.1362444005004237, "grad_norm": 0.4620091122342146, "learning_rate": 1.0985787449556554e-05, "loss": 0.2387, "step": 3520 }, { "epoch": 1.1365672545300456, "grad_norm": 0.3953599002568697, "learning_rate": 1.0983423477771822e-05, "loss": 0.2644, "step": 3521 }, { "epoch": 1.1368901085596674, "grad_norm": 0.47645623425100064, "learning_rate": 1.0981059064642131e-05, "loss": 0.3538, "step": 3522 }, { "epoch": 1.1372129625892893, "grad_norm": 0.3954396259068807, "learning_rate": 1.097869421046705e-05, "loss": 0.3462, "step": 3523 }, { "epoch": 1.1375358166189111, "grad_norm": 0.4038649358891756, "learning_rate": 1.0976328915546203e-05, "loss": 0.2904, "step": 3524 }, { "epoch": 1.137858670648533, "grad_norm": 0.4207945906956224, "learning_rate": 1.0973963180179272e-05, "loss": 0.3232, "step": 3525 }, { "epoch": 1.1381815246781548, "grad_norm": 0.9301012991010427, "learning_rate": 1.0971597004665988e-05, "loss": 0.3029, "step": 3526 }, { "epoch": 1.1385043787077767, "grad_norm": 0.5004262529108382, "learning_rate": 1.0969230389306146e-05, "loss": 0.324, "step": 3527 }, { "epoch": 1.1388272327373987, "grad_norm": 0.49875027196858357, "learning_rate": 1.0966863334399589e-05, "loss": 0.289, "step": 3528 }, { "epoch": 1.1391500867670206, "grad_norm": 0.3714460305301214, "learning_rate": 1.096449584024622e-05, "loss": 0.303, "step": 3529 }, { "epoch": 1.1394729407966424, "grad_norm": 0.4192094234772128, "learning_rate": 1.0962127907146003e-05, "loss": 0.3356, "step": 3530 }, { "epoch": 1.1397957948262643, "grad_norm": 0.3936783934435432, "learning_rate": 1.0959759535398946e-05, "loss": 0.2981, "step": 3531 }, { "epoch": 1.140118648855886, "grad_norm": 0.44907020162572836, "learning_rate": 1.095739072530512e-05, "loss": 0.2944, "step": 3532 }, { "epoch": 1.140441502885508, "grad_norm": 0.47435616795623625, "learning_rate": 1.095502147716465e-05, "loss": 0.3577, "step": 3533 }, { "epoch": 1.1407643569151298, "grad_norm": 0.4688559614412702, "learning_rate": 1.0952651791277719e-05, "loss": 0.3027, "step": 3534 }, { "epoch": 1.1410872109447516, "grad_norm": 0.375174304626177, "learning_rate": 1.0950281667944558e-05, "loss": 0.2943, "step": 3535 }, { "epoch": 1.1414100649743735, "grad_norm": 0.41148615020480067, "learning_rate": 1.0947911107465463e-05, "loss": 0.3582, "step": 3536 }, { "epoch": 1.1417329190039953, "grad_norm": 0.45818696845044365, "learning_rate": 1.0945540110140777e-05, "loss": 0.3373, "step": 3537 }, { "epoch": 1.1420557730336172, "grad_norm": 0.4031046728181577, "learning_rate": 1.0943168676270905e-05, "loss": 0.3314, "step": 3538 }, { "epoch": 1.142378627063239, "grad_norm": 0.4035264735420256, "learning_rate": 1.0940796806156302e-05, "loss": 0.3348, "step": 3539 }, { "epoch": 1.1427014810928608, "grad_norm": 0.4404334893003483, "learning_rate": 1.0938424500097482e-05, "loss": 0.3043, "step": 3540 }, { "epoch": 1.1430243351224827, "grad_norm": 0.45885030972291774, "learning_rate": 1.0936051758395013e-05, "loss": 0.3516, "step": 3541 }, { "epoch": 1.1433471891521045, "grad_norm": 0.3948882834638873, "learning_rate": 1.093367858134952e-05, "loss": 0.2719, "step": 3542 }, { "epoch": 1.1436700431817264, "grad_norm": 0.39462391004268804, "learning_rate": 1.0931304969261676e-05, "loss": 0.26, "step": 3543 }, { "epoch": 1.1439928972113482, "grad_norm": 0.4431086470326863, "learning_rate": 1.092893092243222e-05, "loss": 0.2831, "step": 3544 }, { "epoch": 1.14431575124097, "grad_norm": 0.40347157552070667, "learning_rate": 1.0926556441161936e-05, "loss": 0.299, "step": 3545 }, { "epoch": 1.144638605270592, "grad_norm": 0.43590429564526956, "learning_rate": 1.092418152575167e-05, "loss": 0.3173, "step": 3546 }, { "epoch": 1.144961459300214, "grad_norm": 0.41842451213516285, "learning_rate": 1.092180617650232e-05, "loss": 0.3438, "step": 3547 }, { "epoch": 1.1452843133298358, "grad_norm": 0.45542486607755495, "learning_rate": 1.0919430393714842e-05, "loss": 0.3426, "step": 3548 }, { "epoch": 1.1456071673594577, "grad_norm": 0.46670573391907566, "learning_rate": 1.0917054177690241e-05, "loss": 0.3173, "step": 3549 }, { "epoch": 1.1459300213890795, "grad_norm": 0.4322142062537602, "learning_rate": 1.0914677528729581e-05, "loss": 0.3016, "step": 3550 }, { "epoch": 1.1462528754187014, "grad_norm": 0.4410445139419054, "learning_rate": 1.0912300447133979e-05, "loss": 0.2986, "step": 3551 }, { "epoch": 1.1465757294483232, "grad_norm": 0.3646023967441369, "learning_rate": 1.0909922933204611e-05, "loss": 0.3308, "step": 3552 }, { "epoch": 1.146898583477945, "grad_norm": 0.3888617062413724, "learning_rate": 1.0907544987242703e-05, "loss": 0.368, "step": 3553 }, { "epoch": 1.1472214375075669, "grad_norm": 0.3758912144172966, "learning_rate": 1.0905166609549536e-05, "loss": 0.3089, "step": 3554 }, { "epoch": 1.1475442915371887, "grad_norm": 0.41949082122281517, "learning_rate": 1.0902787800426451e-05, "loss": 0.3091, "step": 3555 }, { "epoch": 1.1478671455668106, "grad_norm": 0.39431233471678234, "learning_rate": 1.0900408560174835e-05, "loss": 0.2944, "step": 3556 }, { "epoch": 1.1481899995964324, "grad_norm": 0.4187093230005943, "learning_rate": 1.0898028889096139e-05, "loss": 0.3586, "step": 3557 }, { "epoch": 1.1485128536260543, "grad_norm": 0.3474515598170988, "learning_rate": 1.0895648787491861e-05, "loss": 0.295, "step": 3558 }, { "epoch": 1.148835707655676, "grad_norm": 0.37006004779126184, "learning_rate": 1.0893268255663558e-05, "loss": 0.3041, "step": 3559 }, { "epoch": 1.149158561685298, "grad_norm": 0.3988794144494854, "learning_rate": 1.089088729391284e-05, "loss": 0.3073, "step": 3560 }, { "epoch": 1.1494814157149198, "grad_norm": 0.4684727809298982, "learning_rate": 1.088850590254137e-05, "loss": 0.2698, "step": 3561 }, { "epoch": 1.1498042697445419, "grad_norm": 0.4817731972232316, "learning_rate": 1.0886124081850872e-05, "loss": 0.2558, "step": 3562 }, { "epoch": 1.1501271237741637, "grad_norm": 0.3838318356609732, "learning_rate": 1.0883741832143114e-05, "loss": 0.2959, "step": 3563 }, { "epoch": 1.1504499778037856, "grad_norm": 0.45386153401941415, "learning_rate": 1.0881359153719926e-05, "loss": 0.2524, "step": 3564 }, { "epoch": 1.1507728318334074, "grad_norm": 0.4307269251751989, "learning_rate": 1.0878976046883189e-05, "loss": 0.2769, "step": 3565 }, { "epoch": 1.1510956858630292, "grad_norm": 0.3721682615436501, "learning_rate": 1.087659251193484e-05, "loss": 0.2455, "step": 3566 }, { "epoch": 1.151418539892651, "grad_norm": 0.5453245555549184, "learning_rate": 1.087420854917687e-05, "loss": 0.3361, "step": 3567 }, { "epoch": 1.151741393922273, "grad_norm": 0.3828216868368545, "learning_rate": 1.087182415891132e-05, "loss": 0.2715, "step": 3568 }, { "epoch": 1.1520642479518948, "grad_norm": 0.3489269359883982, "learning_rate": 1.0869439341440295e-05, "loss": 0.3177, "step": 3569 }, { "epoch": 1.1523871019815166, "grad_norm": 0.5031638523203492, "learning_rate": 1.0867054097065946e-05, "loss": 0.3622, "step": 3570 }, { "epoch": 1.1527099560111385, "grad_norm": 0.4180855884932528, "learning_rate": 1.086466842609048e-05, "loss": 0.3377, "step": 3571 }, { "epoch": 1.1530328100407603, "grad_norm": 0.4439409814536043, "learning_rate": 1.0862282328816157e-05, "loss": 0.3486, "step": 3572 }, { "epoch": 1.1533556640703821, "grad_norm": 0.3625585805232526, "learning_rate": 1.0859895805545294e-05, "loss": 0.2622, "step": 3573 }, { "epoch": 1.153678518100004, "grad_norm": 0.4265405661701351, "learning_rate": 1.085750885658026e-05, "loss": 0.313, "step": 3574 }, { "epoch": 1.1540013721296258, "grad_norm": 0.3884406366039814, "learning_rate": 1.0855121482223473e-05, "loss": 0.3523, "step": 3575 }, { "epoch": 1.1543242261592477, "grad_norm": 0.42690460047345324, "learning_rate": 1.0852733682777418e-05, "loss": 0.3523, "step": 3576 }, { "epoch": 1.1546470801888695, "grad_norm": 0.4524429709091521, "learning_rate": 1.085034545854462e-05, "loss": 0.3037, "step": 3577 }, { "epoch": 1.1549699342184914, "grad_norm": 0.37333096457152487, "learning_rate": 1.0847956809827667e-05, "loss": 0.3069, "step": 3578 }, { "epoch": 1.1552927882481132, "grad_norm": 0.3603182972152532, "learning_rate": 1.08455677369292e-05, "loss": 0.3072, "step": 3579 }, { "epoch": 1.155615642277735, "grad_norm": 0.38569639352922286, "learning_rate": 1.0843178240151904e-05, "loss": 0.2843, "step": 3580 }, { "epoch": 1.1559384963073571, "grad_norm": 0.4330389905408382, "learning_rate": 1.0840788319798531e-05, "loss": 0.281, "step": 3581 }, { "epoch": 1.156261350336979, "grad_norm": 0.36445441608225465, "learning_rate": 1.0838397976171876e-05, "loss": 0.3113, "step": 3582 }, { "epoch": 1.1565842043666008, "grad_norm": 0.5080032944784483, "learning_rate": 1.0836007209574798e-05, "loss": 0.3643, "step": 3583 }, { "epoch": 1.1569070583962227, "grad_norm": 0.37544726931129085, "learning_rate": 1.0833616020310196e-05, "loss": 0.3333, "step": 3584 }, { "epoch": 1.1572299124258445, "grad_norm": 0.4229823766506469, "learning_rate": 1.0831224408681038e-05, "loss": 0.368, "step": 3585 }, { "epoch": 1.1575527664554663, "grad_norm": 0.35403109015566026, "learning_rate": 1.0828832374990335e-05, "loss": 0.2283, "step": 3586 }, { "epoch": 1.1578756204850882, "grad_norm": 0.4245796023654468, "learning_rate": 1.0826439919541155e-05, "loss": 0.3398, "step": 3587 }, { "epoch": 1.15819847451471, "grad_norm": 0.38934794661068295, "learning_rate": 1.0824047042636617e-05, "loss": 0.3693, "step": 3588 }, { "epoch": 1.1585213285443319, "grad_norm": 0.4513482617624975, "learning_rate": 1.0821653744579894e-05, "loss": 0.338, "step": 3589 }, { "epoch": 1.1588441825739537, "grad_norm": 0.3802131980337234, "learning_rate": 1.0819260025674218e-05, "loss": 0.327, "step": 3590 }, { "epoch": 1.1591670366035756, "grad_norm": 0.3799933407409748, "learning_rate": 1.0816865886222867e-05, "loss": 0.264, "step": 3591 }, { "epoch": 1.1594898906331974, "grad_norm": 0.3738128906667822, "learning_rate": 1.0814471326529175e-05, "loss": 0.2936, "step": 3592 }, { "epoch": 1.1598127446628193, "grad_norm": 0.5369470658378726, "learning_rate": 1.0812076346896533e-05, "loss": 0.3604, "step": 3593 }, { "epoch": 1.160135598692441, "grad_norm": 0.464168719119532, "learning_rate": 1.0809680947628377e-05, "loss": 0.2764, "step": 3594 }, { "epoch": 1.1604584527220632, "grad_norm": 0.47546823703278646, "learning_rate": 1.08072851290282e-05, "loss": 0.3411, "step": 3595 }, { "epoch": 1.160781306751685, "grad_norm": 0.3747240461442349, "learning_rate": 1.0804888891399557e-05, "loss": 0.2833, "step": 3596 }, { "epoch": 1.1611041607813068, "grad_norm": 0.43420021323829106, "learning_rate": 1.0802492235046037e-05, "loss": 0.3101, "step": 3597 }, { "epoch": 1.1614270148109287, "grad_norm": 0.4911153768946056, "learning_rate": 1.0800095160271303e-05, "loss": 0.3328, "step": 3598 }, { "epoch": 1.1617498688405505, "grad_norm": 0.4004849626903769, "learning_rate": 1.0797697667379055e-05, "loss": 0.2995, "step": 3599 }, { "epoch": 1.1620727228701724, "grad_norm": 0.41022630844701713, "learning_rate": 1.0795299756673052e-05, "loss": 0.3089, "step": 3600 }, { "epoch": 1.1623955768997942, "grad_norm": 0.5446155107470874, "learning_rate": 1.079290142845711e-05, "loss": 0.3112, "step": 3601 }, { "epoch": 1.162718430929416, "grad_norm": 0.3770932744030746, "learning_rate": 1.0790502683035092e-05, "loss": 0.306, "step": 3602 }, { "epoch": 1.163041284959038, "grad_norm": 0.40497113514179295, "learning_rate": 1.0788103520710915e-05, "loss": 0.3405, "step": 3603 }, { "epoch": 1.1633641389886598, "grad_norm": 0.4204213971631819, "learning_rate": 1.0785703941788548e-05, "loss": 0.3015, "step": 3604 }, { "epoch": 1.1636869930182816, "grad_norm": 0.4937499823585635, "learning_rate": 1.078330394657202e-05, "loss": 0.2537, "step": 3605 }, { "epoch": 1.1640098470479034, "grad_norm": 0.3918229362596138, "learning_rate": 1.0780903535365401e-05, "loss": 0.311, "step": 3606 }, { "epoch": 1.1643327010775253, "grad_norm": 0.38325436474133495, "learning_rate": 1.077850270847282e-05, "loss": 0.2671, "step": 3607 }, { "epoch": 1.1646555551071471, "grad_norm": 0.44264144717747744, "learning_rate": 1.0776101466198463e-05, "loss": 0.3733, "step": 3608 }, { "epoch": 1.164978409136769, "grad_norm": 0.5344250019463035, "learning_rate": 1.0773699808846561e-05, "loss": 0.2971, "step": 3609 }, { "epoch": 1.1653012631663908, "grad_norm": 0.33975328467717886, "learning_rate": 1.0771297736721404e-05, "loss": 0.3296, "step": 3610 }, { "epoch": 1.1656241171960127, "grad_norm": 0.4966568781086854, "learning_rate": 1.0768895250127326e-05, "loss": 0.2777, "step": 3611 }, { "epoch": 1.1659469712256345, "grad_norm": 0.40041423637604096, "learning_rate": 1.0766492349368721e-05, "loss": 0.3141, "step": 3612 }, { "epoch": 1.1662698252552564, "grad_norm": 0.3730368157872748, "learning_rate": 1.0764089034750036e-05, "loss": 0.3283, "step": 3613 }, { "epoch": 1.1665926792848784, "grad_norm": 0.3873075239685321, "learning_rate": 1.0761685306575762e-05, "loss": 0.3087, "step": 3614 }, { "epoch": 1.1669155333145003, "grad_norm": 0.40959523653849894, "learning_rate": 1.0759281165150452e-05, "loss": 0.3122, "step": 3615 }, { "epoch": 1.167238387344122, "grad_norm": 0.3930129049983321, "learning_rate": 1.0756876610778707e-05, "loss": 0.3524, "step": 3616 }, { "epoch": 1.167561241373744, "grad_norm": 0.3785939846745167, "learning_rate": 1.0754471643765181e-05, "loss": 0.2674, "step": 3617 }, { "epoch": 1.1678840954033658, "grad_norm": 0.4390936077423951, "learning_rate": 1.0752066264414579e-05, "loss": 0.3067, "step": 3618 }, { "epoch": 1.1682069494329876, "grad_norm": 0.40971079952803474, "learning_rate": 1.074966047303166e-05, "loss": 0.3589, "step": 3619 }, { "epoch": 1.1685298034626095, "grad_norm": 0.4028214834137019, "learning_rate": 1.0747254269921233e-05, "loss": 0.3026, "step": 3620 }, { "epoch": 1.1688526574922313, "grad_norm": 0.4182975782346912, "learning_rate": 1.0744847655388164e-05, "loss": 0.3601, "step": 3621 }, { "epoch": 1.1691755115218532, "grad_norm": 0.3766487329362992, "learning_rate": 1.0742440629737365e-05, "loss": 0.3048, "step": 3622 }, { "epoch": 1.169498365551475, "grad_norm": 0.3978088069505552, "learning_rate": 1.0740033193273802e-05, "loss": 0.2924, "step": 3623 }, { "epoch": 1.1698212195810969, "grad_norm": 0.44104232042899216, "learning_rate": 1.0737625346302496e-05, "loss": 0.2822, "step": 3624 }, { "epoch": 1.1701440736107187, "grad_norm": 0.5094030061414687, "learning_rate": 1.0735217089128517e-05, "loss": 0.3026, "step": 3625 }, { "epoch": 1.1704669276403405, "grad_norm": 0.36375710924745647, "learning_rate": 1.0732808422056987e-05, "loss": 0.2863, "step": 3626 }, { "epoch": 1.1707897816699624, "grad_norm": 0.4425659000720202, "learning_rate": 1.0730399345393085e-05, "loss": 0.3166, "step": 3627 }, { "epoch": 1.1711126356995842, "grad_norm": 1.0666684713796515, "learning_rate": 1.0727989859442032e-05, "loss": 0.3214, "step": 3628 }, { "epoch": 1.1714354897292063, "grad_norm": 0.40314082652982924, "learning_rate": 1.072557996450911e-05, "loss": 0.3116, "step": 3629 }, { "epoch": 1.1717583437588281, "grad_norm": 0.368632486996878, "learning_rate": 1.0723169660899644e-05, "loss": 0.3028, "step": 3630 }, { "epoch": 1.17208119778845, "grad_norm": 0.36773346818701613, "learning_rate": 1.0720758948919026e-05, "loss": 0.324, "step": 3631 }, { "epoch": 1.1724040518180718, "grad_norm": 0.36583043623200096, "learning_rate": 1.071834782887268e-05, "loss": 0.2926, "step": 3632 }, { "epoch": 1.1727269058476937, "grad_norm": 0.4189471463174111, "learning_rate": 1.0715936301066098e-05, "loss": 0.3167, "step": 3633 }, { "epoch": 1.1730497598773155, "grad_norm": 0.41327057980761883, "learning_rate": 1.0713524365804816e-05, "loss": 0.2783, "step": 3634 }, { "epoch": 1.1733726139069374, "grad_norm": 0.41035241722891697, "learning_rate": 1.0711112023394424e-05, "loss": 0.2828, "step": 3635 }, { "epoch": 1.1736954679365592, "grad_norm": 0.3820905159886558, "learning_rate": 1.0708699274140559e-05, "loss": 0.3571, "step": 3636 }, { "epoch": 1.174018321966181, "grad_norm": 0.42196524020311577, "learning_rate": 1.0706286118348913e-05, "loss": 0.3528, "step": 3637 }, { "epoch": 1.174341175995803, "grad_norm": 0.35365549157302356, "learning_rate": 1.0703872556325232e-05, "loss": 0.2759, "step": 3638 }, { "epoch": 1.1746640300254247, "grad_norm": 0.4335732456908914, "learning_rate": 1.070145858837531e-05, "loss": 0.2728, "step": 3639 }, { "epoch": 1.1749868840550466, "grad_norm": 0.4768649313789605, "learning_rate": 1.0699044214804993e-05, "loss": 0.332, "step": 3640 }, { "epoch": 1.1753097380846684, "grad_norm": 0.7230159851027851, "learning_rate": 1.0696629435920181e-05, "loss": 0.2642, "step": 3641 }, { "epoch": 1.1756325921142903, "grad_norm": 0.3831023259672771, "learning_rate": 1.0694214252026822e-05, "loss": 0.2893, "step": 3642 }, { "epoch": 1.1759554461439121, "grad_norm": 0.4636507138051754, "learning_rate": 1.0691798663430913e-05, "loss": 0.2727, "step": 3643 }, { "epoch": 1.176278300173534, "grad_norm": 0.37290797083152216, "learning_rate": 1.0689382670438513e-05, "loss": 0.2471, "step": 3644 }, { "epoch": 1.1766011542031558, "grad_norm": 0.49255346497066604, "learning_rate": 1.0686966273355716e-05, "loss": 0.2543, "step": 3645 }, { "epoch": 1.1769240082327777, "grad_norm": 0.40523950191376595, "learning_rate": 1.0684549472488681e-05, "loss": 0.2746, "step": 3646 }, { "epoch": 1.1772468622623995, "grad_norm": 0.3370344275676944, "learning_rate": 1.0682132268143617e-05, "loss": 0.2893, "step": 3647 }, { "epoch": 1.1775697162920216, "grad_norm": 0.38773483346046006, "learning_rate": 1.0679714660626775e-05, "loss": 0.304, "step": 3648 }, { "epoch": 1.1778925703216434, "grad_norm": 0.41099797713342934, "learning_rate": 1.0677296650244464e-05, "loss": 0.2673, "step": 3649 }, { "epoch": 1.1782154243512653, "grad_norm": 0.39196771737428787, "learning_rate": 1.0674878237303043e-05, "loss": 0.3364, "step": 3650 }, { "epoch": 1.178538278380887, "grad_norm": 0.507120879432741, "learning_rate": 1.067245942210892e-05, "loss": 0.2916, "step": 3651 }, { "epoch": 1.178861132410509, "grad_norm": 0.45092073024505713, "learning_rate": 1.0670040204968559e-05, "loss": 0.3025, "step": 3652 }, { "epoch": 1.1791839864401308, "grad_norm": 0.3910918730653184, "learning_rate": 1.0667620586188469e-05, "loss": 0.2521, "step": 3653 }, { "epoch": 1.1795068404697526, "grad_norm": 0.410997247362455, "learning_rate": 1.0665200566075211e-05, "loss": 0.2548, "step": 3654 }, { "epoch": 1.1798296944993745, "grad_norm": 0.36405001428208367, "learning_rate": 1.0662780144935402e-05, "loss": 0.3148, "step": 3655 }, { "epoch": 1.1801525485289963, "grad_norm": 0.40863574338194725, "learning_rate": 1.0660359323075703e-05, "loss": 0.3287, "step": 3656 }, { "epoch": 1.1804754025586182, "grad_norm": 0.4659359347640222, "learning_rate": 1.065793810080283e-05, "loss": 0.255, "step": 3657 }, { "epoch": 1.18079825658824, "grad_norm": 0.4647505770488467, "learning_rate": 1.065551647842355e-05, "loss": 0.3227, "step": 3658 }, { "epoch": 1.1811211106178618, "grad_norm": 0.6259471339522805, "learning_rate": 1.0653094456244675e-05, "loss": 0.3363, "step": 3659 }, { "epoch": 1.1814439646474837, "grad_norm": 0.3472106866467765, "learning_rate": 1.0650672034573077e-05, "loss": 0.2644, "step": 3660 }, { "epoch": 1.1817668186771055, "grad_norm": 0.5476542668883504, "learning_rate": 1.064824921371567e-05, "loss": 0.2967, "step": 3661 }, { "epoch": 1.1820896727067274, "grad_norm": 0.390265944928813, "learning_rate": 1.064582599397942e-05, "loss": 0.359, "step": 3662 }, { "epoch": 1.1824125267363494, "grad_norm": 0.45083801938733925, "learning_rate": 1.0643402375671353e-05, "loss": 0.3241, "step": 3663 }, { "epoch": 1.1827353807659713, "grad_norm": 0.40003968664535916, "learning_rate": 1.0640978359098534e-05, "loss": 0.3007, "step": 3664 }, { "epoch": 1.1830582347955931, "grad_norm": 0.447675730952194, "learning_rate": 1.063855394456808e-05, "loss": 0.314, "step": 3665 }, { "epoch": 1.183381088825215, "grad_norm": 0.44858568369004714, "learning_rate": 1.0636129132387163e-05, "loss": 0.3077, "step": 3666 }, { "epoch": 1.1837039428548368, "grad_norm": 0.4570091265699777, "learning_rate": 1.0633703922863004e-05, "loss": 0.2814, "step": 3667 }, { "epoch": 1.1840267968844587, "grad_norm": 0.42065718538283114, "learning_rate": 1.0631278316302874e-05, "loss": 0.2973, "step": 3668 }, { "epoch": 1.1843496509140805, "grad_norm": 0.356156082954899, "learning_rate": 1.0628852313014091e-05, "loss": 0.3322, "step": 3669 }, { "epoch": 1.1846725049437024, "grad_norm": 0.38121458656498897, "learning_rate": 1.0626425913304031e-05, "loss": 0.34, "step": 3670 }, { "epoch": 1.1849953589733242, "grad_norm": 0.9535497493553553, "learning_rate": 1.0623999117480111e-05, "loss": 0.2691, "step": 3671 }, { "epoch": 1.185318213002946, "grad_norm": 0.6360386570671493, "learning_rate": 1.062157192584981e-05, "loss": 0.2799, "step": 3672 }, { "epoch": 1.1856410670325679, "grad_norm": 0.37059416316919647, "learning_rate": 1.0619144338720639e-05, "loss": 0.314, "step": 3673 }, { "epoch": 1.1859639210621897, "grad_norm": 0.4813228200205165, "learning_rate": 1.061671635640018e-05, "loss": 0.2741, "step": 3674 }, { "epoch": 1.1862867750918116, "grad_norm": 0.3776503008166239, "learning_rate": 1.0614287979196047e-05, "loss": 0.2756, "step": 3675 }, { "epoch": 1.1866096291214334, "grad_norm": 0.3649254383908565, "learning_rate": 1.0611859207415918e-05, "loss": 0.3041, "step": 3676 }, { "epoch": 1.1869324831510553, "grad_norm": 0.4308324801059457, "learning_rate": 1.0609430041367513e-05, "loss": 0.2795, "step": 3677 }, { "epoch": 1.187255337180677, "grad_norm": 0.9005913474113723, "learning_rate": 1.06070004813586e-05, "loss": 0.328, "step": 3678 }, { "epoch": 1.187578191210299, "grad_norm": 0.3763200369352914, "learning_rate": 1.0604570527697008e-05, "loss": 0.325, "step": 3679 }, { "epoch": 1.1879010452399208, "grad_norm": 0.39458936238757114, "learning_rate": 1.0602140180690604e-05, "loss": 0.2869, "step": 3680 }, { "epoch": 1.1882238992695426, "grad_norm": 0.3555714179416301, "learning_rate": 1.0599709440647313e-05, "loss": 0.3223, "step": 3681 }, { "epoch": 1.1885467532991647, "grad_norm": 0.3377377863735592, "learning_rate": 1.0597278307875101e-05, "loss": 0.2338, "step": 3682 }, { "epoch": 1.1888696073287865, "grad_norm": 0.40487622730121864, "learning_rate": 1.0594846782681996e-05, "loss": 0.338, "step": 3683 }, { "epoch": 1.1891924613584084, "grad_norm": 0.4308474678244712, "learning_rate": 1.0592414865376064e-05, "loss": 0.2847, "step": 3684 }, { "epoch": 1.1895153153880302, "grad_norm": 0.4651810269021247, "learning_rate": 1.0589982556265426e-05, "loss": 0.3219, "step": 3685 }, { "epoch": 1.189838169417652, "grad_norm": 0.48853768676708714, "learning_rate": 1.0587549855658256e-05, "loss": 0.2754, "step": 3686 }, { "epoch": 1.190161023447274, "grad_norm": 0.3788749147583104, "learning_rate": 1.0585116763862769e-05, "loss": 0.2826, "step": 3687 }, { "epoch": 1.1904838774768958, "grad_norm": 0.46475204380966495, "learning_rate": 1.0582683281187235e-05, "loss": 0.2758, "step": 3688 }, { "epoch": 1.1908067315065176, "grad_norm": 0.3772365355143002, "learning_rate": 1.0580249407939976e-05, "loss": 0.2371, "step": 3689 }, { "epoch": 1.1911295855361395, "grad_norm": 0.37938381169927865, "learning_rate": 1.057781514442936e-05, "loss": 0.2903, "step": 3690 }, { "epoch": 1.1914524395657613, "grad_norm": 0.3985913700708402, "learning_rate": 1.0575380490963802e-05, "loss": 0.3338, "step": 3691 }, { "epoch": 1.1917752935953831, "grad_norm": 0.5134776731262976, "learning_rate": 1.0572945447851772e-05, "loss": 0.3279, "step": 3692 }, { "epoch": 1.192098147625005, "grad_norm": 0.3717984626142267, "learning_rate": 1.0570510015401785e-05, "loss": 0.2233, "step": 3693 }, { "epoch": 1.1924210016546268, "grad_norm": 0.41154932958053075, "learning_rate": 1.0568074193922408e-05, "loss": 0.2226, "step": 3694 }, { "epoch": 1.1927438556842487, "grad_norm": 0.38636714788824694, "learning_rate": 1.0565637983722254e-05, "loss": 0.3011, "step": 3695 }, { "epoch": 1.1930667097138705, "grad_norm": 0.4438456729409396, "learning_rate": 1.056320138510999e-05, "loss": 0.4261, "step": 3696 }, { "epoch": 1.1933895637434926, "grad_norm": 0.44929627826285656, "learning_rate": 1.0560764398394329e-05, "loss": 0.2917, "step": 3697 }, { "epoch": 1.1937124177731144, "grad_norm": 0.4483223155435828, "learning_rate": 1.0558327023884034e-05, "loss": 0.2854, "step": 3698 }, { "epoch": 1.1940352718027363, "grad_norm": 0.3828299674350223, "learning_rate": 1.0555889261887917e-05, "loss": 0.2958, "step": 3699 }, { "epoch": 1.1943581258323581, "grad_norm": 0.42677089341601054, "learning_rate": 1.0553451112714838e-05, "loss": 0.3191, "step": 3700 }, { "epoch": 1.19468097986198, "grad_norm": 0.47851812708864006, "learning_rate": 1.0551012576673706e-05, "loss": 0.2861, "step": 3701 }, { "epoch": 1.1950038338916018, "grad_norm": 0.46294508294464165, "learning_rate": 1.0548573654073484e-05, "loss": 0.3076, "step": 3702 }, { "epoch": 1.1953266879212237, "grad_norm": 0.40324313958146774, "learning_rate": 1.0546134345223179e-05, "loss": 0.3123, "step": 3703 }, { "epoch": 1.1956495419508455, "grad_norm": 0.4701567536918203, "learning_rate": 1.0543694650431846e-05, "loss": 0.3101, "step": 3704 }, { "epoch": 1.1959723959804673, "grad_norm": 0.45666541030414753, "learning_rate": 1.0541254570008594e-05, "loss": 0.3146, "step": 3705 }, { "epoch": 1.1962952500100892, "grad_norm": 0.54691098313876, "learning_rate": 1.0538814104262571e-05, "loss": 0.274, "step": 3706 }, { "epoch": 1.196618104039711, "grad_norm": 0.39297823431905815, "learning_rate": 1.0536373253502991e-05, "loss": 0.2748, "step": 3707 }, { "epoch": 1.1969409580693329, "grad_norm": 0.46365142052821684, "learning_rate": 1.0533932018039098e-05, "loss": 0.2403, "step": 3708 }, { "epoch": 1.1972638120989547, "grad_norm": 0.5629228084723547, "learning_rate": 1.0531490398180197e-05, "loss": 0.2598, "step": 3709 }, { "epoch": 1.1975866661285766, "grad_norm": 0.41147957785661665, "learning_rate": 1.0529048394235639e-05, "loss": 0.3154, "step": 3710 }, { "epoch": 1.1979095201581984, "grad_norm": 0.351735215589172, "learning_rate": 1.052660600651482e-05, "loss": 0.3212, "step": 3711 }, { "epoch": 1.1982323741878202, "grad_norm": 0.41845405936453456, "learning_rate": 1.0524163235327185e-05, "loss": 0.2848, "step": 3712 }, { "epoch": 1.198555228217442, "grad_norm": 0.4428135681325655, "learning_rate": 1.0521720080982236e-05, "loss": 0.3201, "step": 3713 }, { "epoch": 1.198878082247064, "grad_norm": 0.3192774829678618, "learning_rate": 1.0519276543789515e-05, "loss": 0.2909, "step": 3714 }, { "epoch": 1.1992009362766858, "grad_norm": 0.33781667211552036, "learning_rate": 1.0516832624058612e-05, "loss": 0.2765, "step": 3715 }, { "epoch": 1.1995237903063078, "grad_norm": 0.4807236271709828, "learning_rate": 1.0514388322099167e-05, "loss": 0.3161, "step": 3716 }, { "epoch": 1.1998466443359297, "grad_norm": 0.3931906753127619, "learning_rate": 1.0511943638220875e-05, "loss": 0.2872, "step": 3717 }, { "epoch": 1.2001694983655515, "grad_norm": 0.4680379092423867, "learning_rate": 1.0509498572733471e-05, "loss": 0.3374, "step": 3718 }, { "epoch": 1.2004923523951734, "grad_norm": 0.3618539720677018, "learning_rate": 1.0507053125946745e-05, "loss": 0.2949, "step": 3719 }, { "epoch": 1.2008152064247952, "grad_norm": 0.4487138596524558, "learning_rate": 1.0504607298170526e-05, "loss": 0.3191, "step": 3720 }, { "epoch": 1.201138060454417, "grad_norm": 0.43719197276784577, "learning_rate": 1.05021610897147e-05, "loss": 0.2827, "step": 3721 }, { "epoch": 1.201460914484039, "grad_norm": 0.5487130804984393, "learning_rate": 1.04997145008892e-05, "loss": 0.2824, "step": 3722 }, { "epoch": 1.2017837685136608, "grad_norm": 0.3518746729879456, "learning_rate": 1.0497267532004003e-05, "loss": 0.2831, "step": 3723 }, { "epoch": 1.2021066225432826, "grad_norm": 0.4957012120071717, "learning_rate": 1.0494820183369134e-05, "loss": 0.2924, "step": 3724 }, { "epoch": 1.2024294765729044, "grad_norm": 0.4570227872121174, "learning_rate": 1.0492372455294675e-05, "loss": 0.3352, "step": 3725 }, { "epoch": 1.2027523306025263, "grad_norm": 0.36536682980791885, "learning_rate": 1.0489924348090748e-05, "loss": 0.2913, "step": 3726 }, { "epoch": 1.2030751846321481, "grad_norm": 0.5012222724836851, "learning_rate": 1.048747586206752e-05, "loss": 0.2977, "step": 3727 }, { "epoch": 1.20339803866177, "grad_norm": 0.43693513980122484, "learning_rate": 1.0485026997535215e-05, "loss": 0.359, "step": 3728 }, { "epoch": 1.2037208926913918, "grad_norm": 0.4804819195576936, "learning_rate": 1.0482577754804104e-05, "loss": 0.2798, "step": 3729 }, { "epoch": 1.2040437467210139, "grad_norm": 0.38132978837285514, "learning_rate": 1.0480128134184494e-05, "loss": 0.3127, "step": 3730 }, { "epoch": 1.2043666007506357, "grad_norm": 0.44450084972344195, "learning_rate": 1.0477678135986757e-05, "loss": 0.3011, "step": 3731 }, { "epoch": 1.2046894547802576, "grad_norm": 0.4277368616866988, "learning_rate": 1.0475227760521295e-05, "loss": 0.2744, "step": 3732 }, { "epoch": 1.2050123088098794, "grad_norm": 0.4539765235708334, "learning_rate": 1.0472777008098578e-05, "loss": 0.3502, "step": 3733 }, { "epoch": 1.2053351628395013, "grad_norm": 0.44380539626261795, "learning_rate": 1.0470325879029106e-05, "loss": 0.31, "step": 3734 }, { "epoch": 1.205658016869123, "grad_norm": 0.39773289082499336, "learning_rate": 1.0467874373623437e-05, "loss": 0.3119, "step": 3735 }, { "epoch": 1.205980870898745, "grad_norm": 0.31958252443201773, "learning_rate": 1.0465422492192172e-05, "loss": 0.2597, "step": 3736 }, { "epoch": 1.2063037249283668, "grad_norm": 0.4699405862935696, "learning_rate": 1.0462970235045958e-05, "loss": 0.285, "step": 3737 }, { "epoch": 1.2066265789579886, "grad_norm": 0.4213623452477965, "learning_rate": 1.0460517602495498e-05, "loss": 0.3092, "step": 3738 }, { "epoch": 1.2069494329876105, "grad_norm": 0.39827660110348534, "learning_rate": 1.0458064594851534e-05, "loss": 0.3032, "step": 3739 }, { "epoch": 1.2072722870172323, "grad_norm": 0.382980945622799, "learning_rate": 1.0455611212424857e-05, "loss": 0.2528, "step": 3740 }, { "epoch": 1.2075951410468542, "grad_norm": 0.4357164584749138, "learning_rate": 1.0453157455526312e-05, "loss": 0.3466, "step": 3741 }, { "epoch": 1.207917995076476, "grad_norm": 0.4334646163937073, "learning_rate": 1.0450703324466785e-05, "loss": 0.2752, "step": 3742 }, { "epoch": 1.2082408491060979, "grad_norm": 0.36147000914478405, "learning_rate": 1.0448248819557207e-05, "loss": 0.3093, "step": 3743 }, { "epoch": 1.2085637031357197, "grad_norm": 0.3858852862047989, "learning_rate": 1.0445793941108565e-05, "loss": 0.2867, "step": 3744 }, { "epoch": 1.2088865571653415, "grad_norm": 0.4523892933686237, "learning_rate": 1.0443338689431889e-05, "loss": 0.3065, "step": 3745 }, { "epoch": 1.2092094111949634, "grad_norm": 0.49827365621292735, "learning_rate": 1.0440883064838253e-05, "loss": 0.298, "step": 3746 }, { "epoch": 1.2095322652245852, "grad_norm": 0.3805814854364435, "learning_rate": 1.0438427067638783e-05, "loss": 0.3211, "step": 3747 }, { "epoch": 1.209855119254207, "grad_norm": 0.4511323133357903, "learning_rate": 1.0435970698144648e-05, "loss": 0.2797, "step": 3748 }, { "epoch": 1.2101779732838291, "grad_norm": 0.38410768588349614, "learning_rate": 1.0433513956667071e-05, "loss": 0.2806, "step": 3749 }, { "epoch": 1.210500827313451, "grad_norm": 0.5110206552287725, "learning_rate": 1.0431056843517314e-05, "loss": 0.2694, "step": 3750 }, { "epoch": 1.2108236813430728, "grad_norm": 0.4920482394732382, "learning_rate": 1.0428599359006691e-05, "loss": 0.3662, "step": 3751 }, { "epoch": 1.2111465353726947, "grad_norm": 0.34455194743542156, "learning_rate": 1.0426141503446565e-05, "loss": 0.3014, "step": 3752 }, { "epoch": 1.2114693894023165, "grad_norm": 0.3758247832392931, "learning_rate": 1.0423683277148338e-05, "loss": 0.3396, "step": 3753 }, { "epoch": 1.2117922434319384, "grad_norm": 0.40614106734175404, "learning_rate": 1.0421224680423464e-05, "loss": 0.3559, "step": 3754 }, { "epoch": 1.2121150974615602, "grad_norm": 0.535363497059571, "learning_rate": 1.0418765713583449e-05, "loss": 0.2999, "step": 3755 }, { "epoch": 1.212437951491182, "grad_norm": 0.4249471532883419, "learning_rate": 1.0416306376939833e-05, "loss": 0.2741, "step": 3756 }, { "epoch": 1.212760805520804, "grad_norm": 0.5886377794113882, "learning_rate": 1.0413846670804218e-05, "loss": 0.2846, "step": 3757 }, { "epoch": 1.2130836595504257, "grad_norm": 0.39812788858066145, "learning_rate": 1.0411386595488245e-05, "loss": 0.3092, "step": 3758 }, { "epoch": 1.2134065135800476, "grad_norm": 0.3566118431318245, "learning_rate": 1.0408926151303597e-05, "loss": 0.3051, "step": 3759 }, { "epoch": 1.2137293676096694, "grad_norm": 0.4348939477111965, "learning_rate": 1.0406465338562012e-05, "loss": 0.2685, "step": 3760 }, { "epoch": 1.2140522216392913, "grad_norm": 0.4035874561171378, "learning_rate": 1.0404004157575269e-05, "loss": 0.2796, "step": 3761 }, { "epoch": 1.2143750756689131, "grad_norm": 0.41391296152152846, "learning_rate": 1.04015426086552e-05, "loss": 0.3296, "step": 3762 }, { "epoch": 1.214697929698535, "grad_norm": 0.49574338955064806, "learning_rate": 1.0399080692113678e-05, "loss": 0.253, "step": 3763 }, { "epoch": 1.215020783728157, "grad_norm": 0.38469612198898245, "learning_rate": 1.0396618408262624e-05, "loss": 0.2719, "step": 3764 }, { "epoch": 1.2153436377577789, "grad_norm": 0.366041185350003, "learning_rate": 1.0394155757414008e-05, "loss": 0.2892, "step": 3765 }, { "epoch": 1.2156664917874007, "grad_norm": 0.42819022560100034, "learning_rate": 1.0391692739879841e-05, "loss": 0.3502, "step": 3766 }, { "epoch": 1.2159893458170226, "grad_norm": 0.3874963712392295, "learning_rate": 1.0389229355972187e-05, "loss": 0.3053, "step": 3767 }, { "epoch": 1.2163121998466444, "grad_norm": 0.4594568091388609, "learning_rate": 1.0386765606003154e-05, "loss": 0.2455, "step": 3768 }, { "epoch": 1.2166350538762662, "grad_norm": 0.3035600493712886, "learning_rate": 1.0384301490284892e-05, "loss": 0.2814, "step": 3769 }, { "epoch": 1.216957907905888, "grad_norm": 0.38498884144251827, "learning_rate": 1.0381837009129605e-05, "loss": 0.2723, "step": 3770 }, { "epoch": 1.21728076193551, "grad_norm": 0.5024812696615067, "learning_rate": 1.0379372162849534e-05, "loss": 0.316, "step": 3771 }, { "epoch": 1.2176036159651318, "grad_norm": 0.48300387674324574, "learning_rate": 1.0376906951756977e-05, "loss": 0.3415, "step": 3772 }, { "epoch": 1.2179264699947536, "grad_norm": 0.43549667870543785, "learning_rate": 1.0374441376164271e-05, "loss": 0.3753, "step": 3773 }, { "epoch": 1.2182493240243755, "grad_norm": 0.4595301249266777, "learning_rate": 1.03719754363838e-05, "loss": 0.2766, "step": 3774 }, { "epoch": 1.2185721780539973, "grad_norm": 0.430255206231442, "learning_rate": 1.0369509132728e-05, "loss": 0.3635, "step": 3775 }, { "epoch": 1.2188950320836192, "grad_norm": 0.37950747416289116, "learning_rate": 1.036704246550934e-05, "loss": 0.287, "step": 3776 }, { "epoch": 1.219217886113241, "grad_norm": 0.4630390489082917, "learning_rate": 1.036457543504035e-05, "loss": 0.2933, "step": 3777 }, { "epoch": 1.2195407401428628, "grad_norm": 0.4144545597535042, "learning_rate": 1.0362108041633596e-05, "loss": 0.2899, "step": 3778 }, { "epoch": 1.2198635941724847, "grad_norm": 0.3731610290231008, "learning_rate": 1.0359640285601696e-05, "loss": 0.3294, "step": 3779 }, { "epoch": 1.2201864482021065, "grad_norm": 0.39276129437096424, "learning_rate": 1.0357172167257306e-05, "loss": 0.3087, "step": 3780 }, { "epoch": 1.2205093022317284, "grad_norm": 0.3346370670371464, "learning_rate": 1.035470368691314e-05, "loss": 0.2547, "step": 3781 }, { "epoch": 1.2208321562613502, "grad_norm": 0.4001444169091056, "learning_rate": 1.0352234844881948e-05, "loss": 0.3126, "step": 3782 }, { "epoch": 1.2211550102909723, "grad_norm": 0.38016738280561824, "learning_rate": 1.0349765641476528e-05, "loss": 0.2777, "step": 3783 }, { "epoch": 1.2214778643205941, "grad_norm": 0.4562656826748666, "learning_rate": 1.0347296077009727e-05, "loss": 0.3024, "step": 3784 }, { "epoch": 1.221800718350216, "grad_norm": 0.32575712163126197, "learning_rate": 1.0344826151794436e-05, "loss": 0.2601, "step": 3785 }, { "epoch": 1.2221235723798378, "grad_norm": 0.5350131276423573, "learning_rate": 1.0342355866143586e-05, "loss": 0.3146, "step": 3786 }, { "epoch": 1.2224464264094597, "grad_norm": 0.3831119359149656, "learning_rate": 1.033988522037016e-05, "loss": 0.2838, "step": 3787 }, { "epoch": 1.2227692804390815, "grad_norm": 0.37238579501565927, "learning_rate": 1.0337414214787191e-05, "loss": 0.3158, "step": 3788 }, { "epoch": 1.2230921344687034, "grad_norm": 0.3798054982679121, "learning_rate": 1.0334942849707747e-05, "loss": 0.2978, "step": 3789 }, { "epoch": 1.2234149884983252, "grad_norm": 0.3290806210973058, "learning_rate": 1.0332471125444948e-05, "loss": 0.2918, "step": 3790 }, { "epoch": 1.223737842527947, "grad_norm": 0.3887267355539374, "learning_rate": 1.0329999042311957e-05, "loss": 0.2663, "step": 3791 }, { "epoch": 1.2240606965575689, "grad_norm": 0.8604042887384173, "learning_rate": 1.0327526600621988e-05, "loss": 0.3229, "step": 3792 }, { "epoch": 1.2243835505871907, "grad_norm": 0.4008716663224963, "learning_rate": 1.0325053800688289e-05, "loss": 0.2412, "step": 3793 }, { "epoch": 1.2247064046168126, "grad_norm": 0.4026820212278839, "learning_rate": 1.0322580642824163e-05, "loss": 0.2944, "step": 3794 }, { "epoch": 1.2250292586464344, "grad_norm": 0.412511937008816, "learning_rate": 1.0320107127342958e-05, "loss": 0.2736, "step": 3795 }, { "epoch": 1.2253521126760563, "grad_norm": 0.36216499618117326, "learning_rate": 1.0317633254558064e-05, "loss": 0.2724, "step": 3796 }, { "epoch": 1.225674966705678, "grad_norm": 0.4107679270171651, "learning_rate": 1.0315159024782915e-05, "loss": 0.3226, "step": 3797 }, { "epoch": 1.2259978207353002, "grad_norm": 0.4157734012210629, "learning_rate": 1.0312684438330995e-05, "loss": 0.2789, "step": 3798 }, { "epoch": 1.226320674764922, "grad_norm": 0.34559294095212917, "learning_rate": 1.0310209495515832e-05, "loss": 0.319, "step": 3799 }, { "epoch": 1.2266435287945439, "grad_norm": 0.4210988046307596, "learning_rate": 1.0307734196650993e-05, "loss": 0.2948, "step": 3800 }, { "epoch": 1.2269663828241657, "grad_norm": 0.49610162922401546, "learning_rate": 1.03052585420501e-05, "loss": 0.3272, "step": 3801 }, { "epoch": 1.2272892368537875, "grad_norm": 0.3691539515811394, "learning_rate": 1.0302782532026812e-05, "loss": 0.3498, "step": 3802 }, { "epoch": 1.2276120908834094, "grad_norm": 0.35141635048474196, "learning_rate": 1.0300306166894835e-05, "loss": 0.2989, "step": 3803 }, { "epoch": 1.2279349449130312, "grad_norm": 0.5009963281153861, "learning_rate": 1.0297829446967925e-05, "loss": 0.3069, "step": 3804 }, { "epoch": 1.228257798942653, "grad_norm": 0.36747382712909027, "learning_rate": 1.0295352372559876e-05, "loss": 0.3281, "step": 3805 }, { "epoch": 1.228580652972275, "grad_norm": 0.4206114048407405, "learning_rate": 1.0292874943984533e-05, "loss": 0.2847, "step": 3806 }, { "epoch": 1.2289035070018968, "grad_norm": 0.42106427449235606, "learning_rate": 1.0290397161555779e-05, "loss": 0.2187, "step": 3807 }, { "epoch": 1.2292263610315186, "grad_norm": 0.4048982278016847, "learning_rate": 1.0287919025587548e-05, "loss": 0.3423, "step": 3808 }, { "epoch": 1.2295492150611405, "grad_norm": 0.419041084254791, "learning_rate": 1.0285440536393815e-05, "loss": 0.3452, "step": 3809 }, { "epoch": 1.2298720690907623, "grad_norm": 0.5155107039515859, "learning_rate": 1.02829616942886e-05, "loss": 0.2948, "step": 3810 }, { "epoch": 1.2301949231203841, "grad_norm": 0.47810641778214996, "learning_rate": 1.0280482499585972e-05, "loss": 0.3389, "step": 3811 }, { "epoch": 1.230517777150006, "grad_norm": 0.41383674808490856, "learning_rate": 1.0278002952600042e-05, "loss": 0.3005, "step": 3812 }, { "epoch": 1.2308406311796278, "grad_norm": 0.4122188649772925, "learning_rate": 1.0275523053644963e-05, "loss": 0.2846, "step": 3813 }, { "epoch": 1.2311634852092497, "grad_norm": 0.4041005205195224, "learning_rate": 1.0273042803034935e-05, "loss": 0.2818, "step": 3814 }, { "epoch": 1.2314863392388715, "grad_norm": 0.42530672764855226, "learning_rate": 1.0270562201084201e-05, "loss": 0.3158, "step": 3815 }, { "epoch": 1.2318091932684934, "grad_norm": 0.36648198074305305, "learning_rate": 1.0268081248107052e-05, "loss": 0.3183, "step": 3816 }, { "epoch": 1.2321320472981154, "grad_norm": 0.44198349009482424, "learning_rate": 1.026559994441782e-05, "loss": 0.3313, "step": 3817 }, { "epoch": 1.2324549013277373, "grad_norm": 0.4405218784864066, "learning_rate": 1.0263118290330884e-05, "loss": 0.2804, "step": 3818 }, { "epoch": 1.2327777553573591, "grad_norm": 0.40863656652394986, "learning_rate": 1.0260636286160665e-05, "loss": 0.256, "step": 3819 }, { "epoch": 1.233100609386981, "grad_norm": 0.5149223062015048, "learning_rate": 1.0258153932221632e-05, "loss": 0.2892, "step": 3820 }, { "epoch": 1.2334234634166028, "grad_norm": 0.39934400366217754, "learning_rate": 1.0255671228828293e-05, "loss": 0.3403, "step": 3821 }, { "epoch": 1.2337463174462246, "grad_norm": 0.6389853702906143, "learning_rate": 1.0253188176295206e-05, "loss": 0.2737, "step": 3822 }, { "epoch": 1.2340691714758465, "grad_norm": 0.3586316895731791, "learning_rate": 1.0250704774936966e-05, "loss": 0.2231, "step": 3823 }, { "epoch": 1.2343920255054683, "grad_norm": 0.37216125194055827, "learning_rate": 1.024822102506822e-05, "loss": 0.2612, "step": 3824 }, { "epoch": 1.2347148795350902, "grad_norm": 0.3589750476302135, "learning_rate": 1.0245736927003655e-05, "loss": 0.313, "step": 3825 }, { "epoch": 1.235037733564712, "grad_norm": 0.4426995142977027, "learning_rate": 1.0243252481058e-05, "loss": 0.263, "step": 3826 }, { "epoch": 1.2353605875943339, "grad_norm": 0.4244612628287083, "learning_rate": 1.0240767687546035e-05, "loss": 0.2908, "step": 3827 }, { "epoch": 1.2356834416239557, "grad_norm": 0.44571101936686985, "learning_rate": 1.0238282546782579e-05, "loss": 0.2985, "step": 3828 }, { "epoch": 1.2360062956535776, "grad_norm": 0.42128048884195624, "learning_rate": 1.0235797059082496e-05, "loss": 0.2698, "step": 3829 }, { "epoch": 1.2363291496831994, "grad_norm": 0.5945422704405511, "learning_rate": 1.0233311224760691e-05, "loss": 0.3285, "step": 3830 }, { "epoch": 1.2366520037128212, "grad_norm": 0.45379524706463925, "learning_rate": 1.023082504413212e-05, "loss": 0.3192, "step": 3831 }, { "epoch": 1.2369748577424433, "grad_norm": 0.5036099176286768, "learning_rate": 1.0228338517511775e-05, "loss": 0.2576, "step": 3832 }, { "epoch": 1.2372977117720652, "grad_norm": 0.5258331459751812, "learning_rate": 1.0225851645214697e-05, "loss": 0.3222, "step": 3833 }, { "epoch": 1.237620565801687, "grad_norm": 0.37292182419548886, "learning_rate": 1.022336442755597e-05, "loss": 0.3001, "step": 3834 }, { "epoch": 1.2379434198313088, "grad_norm": 0.3842648312670213, "learning_rate": 1.0220876864850717e-05, "loss": 0.2983, "step": 3835 }, { "epoch": 1.2382662738609307, "grad_norm": 0.6665367981886307, "learning_rate": 1.0218388957414114e-05, "loss": 0.3341, "step": 3836 }, { "epoch": 1.2385891278905525, "grad_norm": 0.6562704913908246, "learning_rate": 1.0215900705561374e-05, "loss": 0.3186, "step": 3837 }, { "epoch": 1.2389119819201744, "grad_norm": 0.514223896264516, "learning_rate": 1.0213412109607755e-05, "loss": 0.2514, "step": 3838 }, { "epoch": 1.2392348359497962, "grad_norm": 0.40611754141568207, "learning_rate": 1.021092316986856e-05, "loss": 0.246, "step": 3839 }, { "epoch": 1.239557689979418, "grad_norm": 0.3638623986723835, "learning_rate": 1.020843388665913e-05, "loss": 0.3129, "step": 3840 }, { "epoch": 1.23988054400904, "grad_norm": 0.4175857958131577, "learning_rate": 1.0205944260294856e-05, "loss": 0.2728, "step": 3841 }, { "epoch": 1.2402033980386618, "grad_norm": 0.43364045406167506, "learning_rate": 1.020345429109117e-05, "loss": 0.3236, "step": 3842 }, { "epoch": 1.2405262520682836, "grad_norm": 0.3914468113116586, "learning_rate": 1.0200963979363553e-05, "loss": 0.297, "step": 3843 }, { "epoch": 1.2408491060979054, "grad_norm": 0.346482731752617, "learning_rate": 1.0198473325427516e-05, "loss": 0.2949, "step": 3844 }, { "epoch": 1.2411719601275273, "grad_norm": 0.38201413108238785, "learning_rate": 1.0195982329598624e-05, "loss": 0.3317, "step": 3845 }, { "epoch": 1.2414948141571491, "grad_norm": 0.4363881601405493, "learning_rate": 1.0193490992192484e-05, "loss": 0.3107, "step": 3846 }, { "epoch": 1.241817668186771, "grad_norm": 0.4462168112640075, "learning_rate": 1.0190999313524744e-05, "loss": 0.314, "step": 3847 }, { "epoch": 1.2421405222163928, "grad_norm": 0.4044994294683553, "learning_rate": 1.0188507293911098e-05, "loss": 0.2845, "step": 3848 }, { "epoch": 1.2424633762460147, "grad_norm": 0.3966794373851005, "learning_rate": 1.018601493366728e-05, "loss": 0.2868, "step": 3849 }, { "epoch": 1.2427862302756365, "grad_norm": 0.3672177972553997, "learning_rate": 1.0183522233109066e-05, "loss": 0.3432, "step": 3850 }, { "epoch": 1.2431090843052586, "grad_norm": 0.41553405844382585, "learning_rate": 1.0181029192552284e-05, "loss": 0.2881, "step": 3851 }, { "epoch": 1.2434319383348804, "grad_norm": 0.4276622769098822, "learning_rate": 1.0178535812312794e-05, "loss": 0.3346, "step": 3852 }, { "epoch": 1.2437547923645023, "grad_norm": 0.3695107490146571, "learning_rate": 1.0176042092706506e-05, "loss": 0.3187, "step": 3853 }, { "epoch": 1.244077646394124, "grad_norm": 0.4039154336241109, "learning_rate": 1.0173548034049367e-05, "loss": 0.2851, "step": 3854 }, { "epoch": 1.244400500423746, "grad_norm": 0.3818209837935908, "learning_rate": 1.0171053636657376e-05, "loss": 0.2288, "step": 3855 }, { "epoch": 1.2447233544533678, "grad_norm": 0.4947789342550218, "learning_rate": 1.0168558900846568e-05, "loss": 0.3105, "step": 3856 }, { "epoch": 1.2450462084829896, "grad_norm": 0.3753447648442251, "learning_rate": 1.0166063826933023e-05, "loss": 0.3041, "step": 3857 }, { "epoch": 1.2453690625126115, "grad_norm": 0.4589312872449624, "learning_rate": 1.0163568415232857e-05, "loss": 0.3313, "step": 3858 }, { "epoch": 1.2456919165422333, "grad_norm": 0.4120318217680097, "learning_rate": 1.0161072666062246e-05, "loss": 0.3256, "step": 3859 }, { "epoch": 1.2460147705718552, "grad_norm": 0.38581729965127, "learning_rate": 1.0158576579737392e-05, "loss": 0.3239, "step": 3860 }, { "epoch": 1.246337624601477, "grad_norm": 0.40728377350115275, "learning_rate": 1.0156080156574545e-05, "loss": 0.3362, "step": 3861 }, { "epoch": 1.2466604786310989, "grad_norm": 0.3968539653891613, "learning_rate": 1.0153583396890002e-05, "loss": 0.3105, "step": 3862 }, { "epoch": 1.2469833326607207, "grad_norm": 0.4737270590630093, "learning_rate": 1.0151086301000093e-05, "loss": 0.2832, "step": 3863 }, { "epoch": 1.2473061866903425, "grad_norm": 0.3985451730666994, "learning_rate": 1.0148588869221206e-05, "loss": 0.312, "step": 3864 }, { "epoch": 1.2476290407199646, "grad_norm": 0.4266760935557804, "learning_rate": 1.0146091101869752e-05, "loss": 0.2992, "step": 3865 }, { "epoch": 1.2479518947495865, "grad_norm": 0.6250949541034673, "learning_rate": 1.01435929992622e-05, "loss": 0.3778, "step": 3866 }, { "epoch": 1.2482747487792083, "grad_norm": 0.4372261421762013, "learning_rate": 1.0141094561715058e-05, "loss": 0.3354, "step": 3867 }, { "epoch": 1.2485976028088301, "grad_norm": 0.46463487630168854, "learning_rate": 1.013859578954487e-05, "loss": 0.3539, "step": 3868 }, { "epoch": 1.248920456838452, "grad_norm": 0.40782741224073954, "learning_rate": 1.013609668306823e-05, "loss": 0.2856, "step": 3869 }, { "epoch": 1.2492433108680738, "grad_norm": 0.3900693871954252, "learning_rate": 1.0133597242601768e-05, "loss": 0.2507, "step": 3870 }, { "epoch": 1.2495661648976957, "grad_norm": 0.3871209231660493, "learning_rate": 1.0131097468462165e-05, "loss": 0.293, "step": 3871 }, { "epoch": 1.2498890189273175, "grad_norm": 0.41020713718373036, "learning_rate": 1.0128597360966133e-05, "loss": 0.2645, "step": 3872 }, { "epoch": 1.2502118729569394, "grad_norm": 0.34156343989882704, "learning_rate": 1.0126096920430438e-05, "loss": 0.3352, "step": 3873 }, { "epoch": 1.2505347269865612, "grad_norm": 0.45583929650522614, "learning_rate": 1.0123596147171876e-05, "loss": 0.3527, "step": 3874 }, { "epoch": 1.250857581016183, "grad_norm": 0.40236835341091587, "learning_rate": 1.01210950415073e-05, "loss": 0.3, "step": 3875 }, { "epoch": 1.251180435045805, "grad_norm": 0.4084779917818216, "learning_rate": 1.0118593603753586e-05, "loss": 0.3074, "step": 3876 }, { "epoch": 1.2515032890754267, "grad_norm": 0.3922134755196153, "learning_rate": 1.0116091834227673e-05, "loss": 0.2822, "step": 3877 }, { "epoch": 1.2518261431050486, "grad_norm": 0.376390652840398, "learning_rate": 1.0113589733246525e-05, "loss": 0.3115, "step": 3878 }, { "epoch": 1.2521489971346704, "grad_norm": 0.375297381848034, "learning_rate": 1.011108730112716e-05, "loss": 0.2918, "step": 3879 }, { "epoch": 1.2524718511642923, "grad_norm": 0.48639492242538457, "learning_rate": 1.0108584538186626e-05, "loss": 0.2824, "step": 3880 }, { "epoch": 1.2527947051939141, "grad_norm": 0.39971823549186125, "learning_rate": 1.0106081444742024e-05, "loss": 0.3134, "step": 3881 }, { "epoch": 1.253117559223536, "grad_norm": 0.43124925891667, "learning_rate": 1.0103578021110493e-05, "loss": 0.3141, "step": 3882 }, { "epoch": 1.2534404132531578, "grad_norm": 0.5496672930337038, "learning_rate": 1.010107426760921e-05, "loss": 0.304, "step": 3883 }, { "epoch": 1.2537632672827796, "grad_norm": 0.36246863074358643, "learning_rate": 1.00985701845554e-05, "loss": 0.2752, "step": 3884 }, { "epoch": 1.2540861213124015, "grad_norm": 0.3492773046896468, "learning_rate": 1.0096065772266327e-05, "loss": 0.3234, "step": 3885 }, { "epoch": 1.2544089753420236, "grad_norm": 0.40546195776731847, "learning_rate": 1.0093561031059298e-05, "loss": 0.3659, "step": 3886 }, { "epoch": 1.2547318293716454, "grad_norm": 0.3955530294941626, "learning_rate": 1.0091055961251655e-05, "loss": 0.253, "step": 3887 }, { "epoch": 1.2550546834012672, "grad_norm": 0.3735566363262567, "learning_rate": 1.0088550563160789e-05, "loss": 0.3458, "step": 3888 }, { "epoch": 1.255377537430889, "grad_norm": 0.4261155214172522, "learning_rate": 1.0086044837104132e-05, "loss": 0.3323, "step": 3889 }, { "epoch": 1.255700391460511, "grad_norm": 0.39973984375259847, "learning_rate": 1.0083538783399155e-05, "loss": 0.3102, "step": 3890 }, { "epoch": 1.2560232454901328, "grad_norm": 0.36838571997677044, "learning_rate": 1.0081032402363372e-05, "loss": 0.3491, "step": 3891 }, { "epoch": 1.2563460995197546, "grad_norm": 0.42964397795285114, "learning_rate": 1.0078525694314339e-05, "loss": 0.3017, "step": 3892 }, { "epoch": 1.2566689535493765, "grad_norm": 0.390047959883531, "learning_rate": 1.0076018659569649e-05, "loss": 0.3005, "step": 3893 }, { "epoch": 1.2569918075789983, "grad_norm": 0.4439880343589355, "learning_rate": 1.0073511298446944e-05, "loss": 0.2249, "step": 3894 }, { "epoch": 1.2573146616086202, "grad_norm": 0.37587749005503357, "learning_rate": 1.0071003611263904e-05, "loss": 0.2848, "step": 3895 }, { "epoch": 1.257637515638242, "grad_norm": 0.3826871902167496, "learning_rate": 1.0068495598338245e-05, "loss": 0.2438, "step": 3896 }, { "epoch": 1.2579603696678638, "grad_norm": 0.39920954060223546, "learning_rate": 1.0065987259987728e-05, "loss": 0.2722, "step": 3897 }, { "epoch": 1.258283223697486, "grad_norm": 0.39553541475046994, "learning_rate": 1.0063478596530159e-05, "loss": 0.3309, "step": 3898 }, { "epoch": 1.2586060777271078, "grad_norm": 0.4984330539530697, "learning_rate": 1.0060969608283385e-05, "loss": 0.3114, "step": 3899 }, { "epoch": 1.2589289317567296, "grad_norm": 0.4237229189040011, "learning_rate": 1.0058460295565289e-05, "loss": 0.3361, "step": 3900 }, { "epoch": 1.2592517857863514, "grad_norm": 0.4545566455871988, "learning_rate": 1.0055950658693794e-05, "loss": 0.2929, "step": 3901 }, { "epoch": 1.2595746398159733, "grad_norm": 0.41096940127335935, "learning_rate": 1.005344069798687e-05, "loss": 0.3126, "step": 3902 }, { "epoch": 1.2598974938455951, "grad_norm": 0.42960946897899105, "learning_rate": 1.0050930413762526e-05, "loss": 0.3973, "step": 3903 }, { "epoch": 1.260220347875217, "grad_norm": 0.46157927996455367, "learning_rate": 1.0048419806338813e-05, "loss": 0.2837, "step": 3904 }, { "epoch": 1.2605432019048388, "grad_norm": 0.4467533934306536, "learning_rate": 1.004590887603382e-05, "loss": 0.3309, "step": 3905 }, { "epoch": 1.2608660559344607, "grad_norm": 0.48576297841031896, "learning_rate": 1.0043397623165676e-05, "loss": 0.2355, "step": 3906 }, { "epoch": 1.2611889099640825, "grad_norm": 0.4016602657129394, "learning_rate": 1.004088604805256e-05, "loss": 0.3583, "step": 3907 }, { "epoch": 1.2615117639937043, "grad_norm": 0.4251170187522968, "learning_rate": 1.0038374151012677e-05, "loss": 0.3107, "step": 3908 }, { "epoch": 1.2618346180233262, "grad_norm": 0.4021314553275014, "learning_rate": 1.0035861932364285e-05, "loss": 0.2709, "step": 3909 }, { "epoch": 1.262157472052948, "grad_norm": 0.40737897849153676, "learning_rate": 1.0033349392425679e-05, "loss": 0.2995, "step": 3910 }, { "epoch": 1.2624803260825699, "grad_norm": 0.4480556478725673, "learning_rate": 1.0030836531515193e-05, "loss": 0.2759, "step": 3911 }, { "epoch": 1.2628031801121917, "grad_norm": 0.4179675129577791, "learning_rate": 1.0028323349951205e-05, "loss": 0.251, "step": 3912 }, { "epoch": 1.2631260341418136, "grad_norm": 0.5083474648962544, "learning_rate": 1.0025809848052127e-05, "loss": 0.3105, "step": 3913 }, { "epoch": 1.2634488881714354, "grad_norm": 0.5601959140908792, "learning_rate": 1.0023296026136423e-05, "loss": 0.333, "step": 3914 }, { "epoch": 1.2637717422010573, "grad_norm": 0.3535153332760817, "learning_rate": 1.0020781884522587e-05, "loss": 0.2769, "step": 3915 }, { "epoch": 1.264094596230679, "grad_norm": 0.4937691126943971, "learning_rate": 1.0018267423529155e-05, "loss": 0.3253, "step": 3916 }, { "epoch": 1.264417450260301, "grad_norm": 0.4785542531969651, "learning_rate": 1.001575264347471e-05, "loss": 0.3181, "step": 3917 }, { "epoch": 1.2647403042899228, "grad_norm": 0.4265515552881296, "learning_rate": 1.0013237544677871e-05, "loss": 0.2846, "step": 3918 }, { "epoch": 1.2650631583195449, "grad_norm": 0.3741543751631275, "learning_rate": 1.0010722127457295e-05, "loss": 0.2637, "step": 3919 }, { "epoch": 1.2653860123491667, "grad_norm": 0.3679073738094856, "learning_rate": 1.0008206392131679e-05, "loss": 0.2873, "step": 3920 }, { "epoch": 1.2657088663787885, "grad_norm": 0.4037296665889679, "learning_rate": 1.0005690339019771e-05, "loss": 0.341, "step": 3921 }, { "epoch": 1.2660317204084104, "grad_norm": 0.394008362906941, "learning_rate": 1.0003173968440349e-05, "loss": 0.2884, "step": 3922 }, { "epoch": 1.2663545744380322, "grad_norm": 0.43676404617741277, "learning_rate": 1.000065728071223e-05, "loss": 0.3042, "step": 3923 }, { "epoch": 1.266677428467654, "grad_norm": 0.41510371958044445, "learning_rate": 9.998140276154277e-06, "loss": 0.2622, "step": 3924 }, { "epoch": 1.267000282497276, "grad_norm": 0.38629706484159687, "learning_rate": 9.995622955085391e-06, "loss": 0.295, "step": 3925 }, { "epoch": 1.2673231365268978, "grad_norm": 0.3572407327720498, "learning_rate": 9.993105317824515e-06, "loss": 0.3043, "step": 3926 }, { "epoch": 1.2676459905565196, "grad_norm": 0.44275627752810287, "learning_rate": 9.990587364690627e-06, "loss": 0.3076, "step": 3927 }, { "epoch": 1.2679688445861415, "grad_norm": 0.36194594053670487, "learning_rate": 9.988069096002748e-06, "loss": 0.3296, "step": 3928 }, { "epoch": 1.2682916986157633, "grad_norm": 0.43468645095594655, "learning_rate": 9.985550512079943e-06, "loss": 0.3583, "step": 3929 }, { "epoch": 1.2686145526453851, "grad_norm": 0.4298566091261411, "learning_rate": 9.98303161324131e-06, "loss": 0.2553, "step": 3930 }, { "epoch": 1.268937406675007, "grad_norm": 0.4750297653166133, "learning_rate": 9.980512399805992e-06, "loss": 0.2886, "step": 3931 }, { "epoch": 1.269260260704629, "grad_norm": 0.3626204333760063, "learning_rate": 9.977992872093169e-06, "loss": 0.2942, "step": 3932 }, { "epoch": 1.269583114734251, "grad_norm": 0.36668092838962063, "learning_rate": 9.975473030422063e-06, "loss": 0.2902, "step": 3933 }, { "epoch": 1.2699059687638727, "grad_norm": 0.38526019787455273, "learning_rate": 9.972952875111931e-06, "loss": 0.3077, "step": 3934 }, { "epoch": 1.2702288227934946, "grad_norm": 0.376259609973643, "learning_rate": 9.970432406482075e-06, "loss": 0.3401, "step": 3935 }, { "epoch": 1.2705516768231164, "grad_norm": 0.40224687725512376, "learning_rate": 9.967911624851834e-06, "loss": 0.3145, "step": 3936 }, { "epoch": 1.2708745308527383, "grad_norm": 0.37620996921637334, "learning_rate": 9.96539053054059e-06, "loss": 0.2951, "step": 3937 }, { "epoch": 1.2711973848823601, "grad_norm": 0.4177764530954863, "learning_rate": 9.96286912386776e-06, "loss": 0.3494, "step": 3938 }, { "epoch": 1.271520238911982, "grad_norm": 0.42507206636478057, "learning_rate": 9.960347405152805e-06, "loss": 0.3192, "step": 3939 }, { "epoch": 1.2718430929416038, "grad_norm": 0.39636308490992045, "learning_rate": 9.957825374715222e-06, "loss": 0.3042, "step": 3940 }, { "epoch": 1.2721659469712256, "grad_norm": 0.44049134983812205, "learning_rate": 9.955303032874547e-06, "loss": 0.3066, "step": 3941 }, { "epoch": 1.2724888010008475, "grad_norm": 0.5355907748922837, "learning_rate": 9.95278037995036e-06, "loss": 0.296, "step": 3942 }, { "epoch": 1.2728116550304693, "grad_norm": 0.3787895398670541, "learning_rate": 9.950257416262276e-06, "loss": 0.2627, "step": 3943 }, { "epoch": 1.2731345090600912, "grad_norm": 0.417539165306836, "learning_rate": 9.947734142129951e-06, "loss": 0.3379, "step": 3944 }, { "epoch": 1.273457363089713, "grad_norm": 0.3754679532834841, "learning_rate": 9.945210557873084e-06, "loss": 0.2937, "step": 3945 }, { "epoch": 1.2737802171193349, "grad_norm": 0.487821650080927, "learning_rate": 9.942686663811402e-06, "loss": 0.2778, "step": 3946 }, { "epoch": 1.2741030711489567, "grad_norm": 0.3797125770608437, "learning_rate": 9.940162460264686e-06, "loss": 0.2714, "step": 3947 }, { "epoch": 1.2744259251785786, "grad_norm": 0.39842794016378646, "learning_rate": 9.937637947552745e-06, "loss": 0.3233, "step": 3948 }, { "epoch": 1.2747487792082004, "grad_norm": 0.41169263972831355, "learning_rate": 9.935113125995436e-06, "loss": 0.2984, "step": 3949 }, { "epoch": 1.2750716332378222, "grad_norm": 0.3749926522455994, "learning_rate": 9.932587995912645e-06, "loss": 0.2891, "step": 3950 }, { "epoch": 1.275394487267444, "grad_norm": 0.3328572696807036, "learning_rate": 9.930062557624306e-06, "loss": 0.2531, "step": 3951 }, { "epoch": 1.275717341297066, "grad_norm": 0.6272486788581606, "learning_rate": 9.927536811450385e-06, "loss": 0.2772, "step": 3952 }, { "epoch": 1.276040195326688, "grad_norm": 0.46410403705627273, "learning_rate": 9.925010757710895e-06, "loss": 0.3474, "step": 3953 }, { "epoch": 1.2763630493563098, "grad_norm": 0.35940951559137524, "learning_rate": 9.92248439672588e-06, "loss": 0.3136, "step": 3954 }, { "epoch": 1.2766859033859317, "grad_norm": 0.44639869220019157, "learning_rate": 9.919957728815431e-06, "loss": 0.3182, "step": 3955 }, { "epoch": 1.2770087574155535, "grad_norm": 0.47185928672331934, "learning_rate": 9.917430754299668e-06, "loss": 0.2826, "step": 3956 }, { "epoch": 1.2773316114451754, "grad_norm": 0.48589013821223664, "learning_rate": 9.914903473498758e-06, "loss": 0.3256, "step": 3957 }, { "epoch": 1.2776544654747972, "grad_norm": 0.36212397996155493, "learning_rate": 9.912375886732902e-06, "loss": 0.2936, "step": 3958 }, { "epoch": 1.277977319504419, "grad_norm": 0.4295753345513036, "learning_rate": 9.909847994322342e-06, "loss": 0.3362, "step": 3959 }, { "epoch": 1.278300173534041, "grad_norm": 0.5677616368288942, "learning_rate": 9.907319796587363e-06, "loss": 0.2746, "step": 3960 }, { "epoch": 1.2786230275636628, "grad_norm": 0.4688389519749365, "learning_rate": 9.904791293848282e-06, "loss": 0.3238, "step": 3961 }, { "epoch": 1.2789458815932846, "grad_norm": 0.3560557704870712, "learning_rate": 9.902262486425452e-06, "loss": 0.2817, "step": 3962 }, { "epoch": 1.2792687356229064, "grad_norm": 0.3657234745763159, "learning_rate": 9.899733374639276e-06, "loss": 0.3037, "step": 3963 }, { "epoch": 1.2795915896525283, "grad_norm": 0.4031256380481748, "learning_rate": 9.897203958810184e-06, "loss": 0.2983, "step": 3964 }, { "epoch": 1.2799144436821501, "grad_norm": 0.4135671392997112, "learning_rate": 9.894674239258651e-06, "loss": 0.3111, "step": 3965 }, { "epoch": 1.2802372977117722, "grad_norm": 0.37692436942504365, "learning_rate": 9.892144216305191e-06, "loss": 0.278, "step": 3966 }, { "epoch": 1.280560151741394, "grad_norm": 0.38850335845662204, "learning_rate": 9.889613890270352e-06, "loss": 0.3507, "step": 3967 }, { "epoch": 1.2808830057710159, "grad_norm": 0.42516162711111455, "learning_rate": 9.887083261474723e-06, "loss": 0.295, "step": 3968 }, { "epoch": 1.2812058598006377, "grad_norm": 0.534425799134805, "learning_rate": 9.884552330238934e-06, "loss": 0.2619, "step": 3969 }, { "epoch": 1.2815287138302596, "grad_norm": 0.7526389829805419, "learning_rate": 9.882021096883648e-06, "loss": 0.292, "step": 3970 }, { "epoch": 1.2818515678598814, "grad_norm": 0.4646485380244275, "learning_rate": 9.87948956172957e-06, "loss": 0.3186, "step": 3971 }, { "epoch": 1.2821744218895033, "grad_norm": 0.4037279995932879, "learning_rate": 9.87695772509744e-06, "loss": 0.3398, "step": 3972 }, { "epoch": 1.282497275919125, "grad_norm": 0.46417509193273643, "learning_rate": 9.874425587308041e-06, "loss": 0.3297, "step": 3973 }, { "epoch": 1.282820129948747, "grad_norm": 0.468346465023644, "learning_rate": 9.871893148682189e-06, "loss": 0.2642, "step": 3974 }, { "epoch": 1.2831429839783688, "grad_norm": 0.42705226834728616, "learning_rate": 9.86936040954074e-06, "loss": 0.2991, "step": 3975 }, { "epoch": 1.2834658380079906, "grad_norm": 0.34461109888243296, "learning_rate": 9.866827370204592e-06, "loss": 0.3013, "step": 3976 }, { "epoch": 1.2837886920376125, "grad_norm": 0.3862065818219458, "learning_rate": 9.864294030994674e-06, "loss": 0.2796, "step": 3977 }, { "epoch": 1.2841115460672343, "grad_norm": 0.45330428152231034, "learning_rate": 9.861760392231959e-06, "loss": 0.3288, "step": 3978 }, { "epoch": 1.2844344000968562, "grad_norm": 0.37828064020180635, "learning_rate": 9.859226454237454e-06, "loss": 0.3403, "step": 3979 }, { "epoch": 1.284757254126478, "grad_norm": 0.3564882340989661, "learning_rate": 9.856692217332206e-06, "loss": 0.2268, "step": 3980 }, { "epoch": 1.2850801081560999, "grad_norm": 0.49051882463279245, "learning_rate": 9.854157681837301e-06, "loss": 0.3237, "step": 3981 }, { "epoch": 1.2854029621857217, "grad_norm": 0.3909531336565797, "learning_rate": 9.851622848073857e-06, "loss": 0.297, "step": 3982 }, { "epoch": 1.2857258162153435, "grad_norm": 0.39644305980891226, "learning_rate": 9.849087716363036e-06, "loss": 0.3312, "step": 3983 }, { "epoch": 1.2860486702449654, "grad_norm": 0.41487145561256866, "learning_rate": 9.846552287026039e-06, "loss": 0.3557, "step": 3984 }, { "epoch": 1.2863715242745872, "grad_norm": 0.439165719651219, "learning_rate": 9.844016560384097e-06, "loss": 0.3377, "step": 3985 }, { "epoch": 1.286694378304209, "grad_norm": 0.4644747693552964, "learning_rate": 9.841480536758488e-06, "loss": 0.3397, "step": 3986 }, { "epoch": 1.2870172323338311, "grad_norm": 0.40795995034687366, "learning_rate": 9.838944216470515e-06, "loss": 0.2959, "step": 3987 }, { "epoch": 1.287340086363453, "grad_norm": 0.4418205922243598, "learning_rate": 9.836407599841534e-06, "loss": 0.3329, "step": 3988 }, { "epoch": 1.2876629403930748, "grad_norm": 0.4073290193892014, "learning_rate": 9.833870687192926e-06, "loss": 0.3121, "step": 3989 }, { "epoch": 1.2879857944226967, "grad_norm": 0.4080397459105725, "learning_rate": 9.831333478846116e-06, "loss": 0.3364, "step": 3990 }, { "epoch": 1.2883086484523185, "grad_norm": 0.4767705127543823, "learning_rate": 9.828795975122563e-06, "loss": 0.3252, "step": 3991 }, { "epoch": 1.2886315024819404, "grad_norm": 0.41124212380295555, "learning_rate": 9.826258176343768e-06, "loss": 0.3152, "step": 3992 }, { "epoch": 1.2889543565115622, "grad_norm": 0.4447436086580914, "learning_rate": 9.823720082831267e-06, "loss": 0.2522, "step": 3993 }, { "epoch": 1.289277210541184, "grad_norm": 0.34489294670004544, "learning_rate": 9.821181694906631e-06, "loss": 0.3025, "step": 3994 }, { "epoch": 1.289600064570806, "grad_norm": 0.37107179079472535, "learning_rate": 9.818643012891471e-06, "loss": 0.3031, "step": 3995 }, { "epoch": 1.2899229186004277, "grad_norm": 0.4411516302865084, "learning_rate": 9.816104037107433e-06, "loss": 0.3459, "step": 3996 }, { "epoch": 1.2902457726300496, "grad_norm": 0.734407412809885, "learning_rate": 9.813564767876204e-06, "loss": 0.2977, "step": 3997 }, { "epoch": 1.2905686266596714, "grad_norm": 0.5600162765951964, "learning_rate": 9.811025205519502e-06, "loss": 0.2564, "step": 3998 }, { "epoch": 1.2908914806892935, "grad_norm": 0.35852643687306757, "learning_rate": 9.808485350359092e-06, "loss": 0.2957, "step": 3999 }, { "epoch": 1.2912143347189153, "grad_norm": 0.35625738228929416, "learning_rate": 9.805945202716767e-06, "loss": 0.3307, "step": 4000 }, { "epoch": 1.2915371887485372, "grad_norm": 0.3440877993050829, "learning_rate": 9.803404762914359e-06, "loss": 0.2996, "step": 4001 }, { "epoch": 1.291860042778159, "grad_norm": 0.42091442366380327, "learning_rate": 9.800864031273738e-06, "loss": 0.2469, "step": 4002 }, { "epoch": 1.2921828968077809, "grad_norm": 0.3297118965288488, "learning_rate": 9.798323008116815e-06, "loss": 0.2897, "step": 4003 }, { "epoch": 1.2925057508374027, "grad_norm": 0.3466858536660386, "learning_rate": 9.795781693765532e-06, "loss": 0.3433, "step": 4004 }, { "epoch": 1.2928286048670246, "grad_norm": 0.32364591111952484, "learning_rate": 9.793240088541871e-06, "loss": 0.3088, "step": 4005 }, { "epoch": 1.2931514588966464, "grad_norm": 0.3321422694629475, "learning_rate": 9.790698192767845e-06, "loss": 0.2867, "step": 4006 }, { "epoch": 1.2934743129262682, "grad_norm": 0.41081623519289684, "learning_rate": 9.788156006765515e-06, "loss": 0.325, "step": 4007 }, { "epoch": 1.29379716695589, "grad_norm": 0.3944396250238895, "learning_rate": 9.78561353085697e-06, "loss": 0.2892, "step": 4008 }, { "epoch": 1.294120020985512, "grad_norm": 0.44821070663111, "learning_rate": 9.783070765364339e-06, "loss": 0.2825, "step": 4009 }, { "epoch": 1.2944428750151338, "grad_norm": 0.3463945719562571, "learning_rate": 9.780527710609785e-06, "loss": 0.3012, "step": 4010 }, { "epoch": 1.2947657290447556, "grad_norm": 0.3959546452804246, "learning_rate": 9.777984366915511e-06, "loss": 0.3159, "step": 4011 }, { "epoch": 1.2950885830743775, "grad_norm": 0.4102827790763738, "learning_rate": 9.775440734603754e-06, "loss": 0.3003, "step": 4012 }, { "epoch": 1.2954114371039993, "grad_norm": 0.3648993251304458, "learning_rate": 9.772896813996789e-06, "loss": 0.2511, "step": 4013 }, { "epoch": 1.2957342911336212, "grad_norm": 0.5452369075231936, "learning_rate": 9.770352605416929e-06, "loss": 0.2661, "step": 4014 }, { "epoch": 1.296057145163243, "grad_norm": 0.41418067819370985, "learning_rate": 9.767808109186518e-06, "loss": 0.2706, "step": 4015 }, { "epoch": 1.2963799991928648, "grad_norm": 0.4117925896024724, "learning_rate": 9.765263325627944e-06, "loss": 0.2564, "step": 4016 }, { "epoch": 1.2967028532224867, "grad_norm": 0.3368447622840997, "learning_rate": 9.762718255063626e-06, "loss": 0.3146, "step": 4017 }, { "epoch": 1.2970257072521085, "grad_norm": 0.41347832246625615, "learning_rate": 9.760172897816022e-06, "loss": 0.3204, "step": 4018 }, { "epoch": 1.2973485612817304, "grad_norm": 0.3602882168806283, "learning_rate": 9.757627254207623e-06, "loss": 0.3385, "step": 4019 }, { "epoch": 1.2976714153113522, "grad_norm": 0.4112181163815936, "learning_rate": 9.755081324560959e-06, "loss": 0.3078, "step": 4020 }, { "epoch": 1.2979942693409743, "grad_norm": 0.41057205828535703, "learning_rate": 9.752535109198597e-06, "loss": 0.2778, "step": 4021 }, { "epoch": 1.2983171233705961, "grad_norm": 0.43334379132575407, "learning_rate": 9.749988608443138e-06, "loss": 0.2417, "step": 4022 }, { "epoch": 1.298639977400218, "grad_norm": 0.43158771343756064, "learning_rate": 9.747441822617222e-06, "loss": 0.2788, "step": 4023 }, { "epoch": 1.2989628314298398, "grad_norm": 0.45100298582526444, "learning_rate": 9.744894752043522e-06, "loss": 0.2753, "step": 4024 }, { "epoch": 1.2992856854594617, "grad_norm": 0.3669257324637715, "learning_rate": 9.742347397044751e-06, "loss": 0.2951, "step": 4025 }, { "epoch": 1.2996085394890835, "grad_norm": 0.4896184036115751, "learning_rate": 9.73979975794365e-06, "loss": 0.2759, "step": 4026 }, { "epoch": 1.2999313935187053, "grad_norm": 0.38899972687522805, "learning_rate": 9.737251835063003e-06, "loss": 0.3024, "step": 4027 }, { "epoch": 1.3002542475483272, "grad_norm": 0.42473588164505705, "learning_rate": 9.734703628725631e-06, "loss": 0.3199, "step": 4028 }, { "epoch": 1.300577101577949, "grad_norm": 0.3480482731010906, "learning_rate": 9.732155139254387e-06, "loss": 0.3016, "step": 4029 }, { "epoch": 1.3008999556075709, "grad_norm": 0.37855469371442513, "learning_rate": 9.729606366972158e-06, "loss": 0.3319, "step": 4030 }, { "epoch": 1.3012228096371927, "grad_norm": 0.7032738545604162, "learning_rate": 9.727057312201876e-06, "loss": 0.3204, "step": 4031 }, { "epoch": 1.3015456636668146, "grad_norm": 0.36773517900468694, "learning_rate": 9.724507975266497e-06, "loss": 0.3226, "step": 4032 }, { "epoch": 1.3018685176964366, "grad_norm": 0.48700869251628454, "learning_rate": 9.721958356489023e-06, "loss": 0.2855, "step": 4033 }, { "epoch": 1.3021913717260585, "grad_norm": 0.4756959742109556, "learning_rate": 9.719408456192485e-06, "loss": 0.3137, "step": 4034 }, { "epoch": 1.3025142257556803, "grad_norm": 0.4386003470859883, "learning_rate": 9.716858274699951e-06, "loss": 0.2676, "step": 4035 }, { "epoch": 1.3028370797853022, "grad_norm": 0.42030950242673976, "learning_rate": 9.714307812334524e-06, "loss": 0.333, "step": 4036 }, { "epoch": 1.303159933814924, "grad_norm": 0.3797236117575117, "learning_rate": 9.711757069419347e-06, "loss": 0.3499, "step": 4037 }, { "epoch": 1.3034827878445459, "grad_norm": 0.42158907728671774, "learning_rate": 9.709206046277596e-06, "loss": 0.276, "step": 4038 }, { "epoch": 1.3038056418741677, "grad_norm": 0.3968026569672756, "learning_rate": 9.706654743232482e-06, "loss": 0.3309, "step": 4039 }, { "epoch": 1.3041284959037895, "grad_norm": 0.36495499787727664, "learning_rate": 9.704103160607251e-06, "loss": 0.3216, "step": 4040 }, { "epoch": 1.3044513499334114, "grad_norm": 0.3651791946734064, "learning_rate": 9.701551298725186e-06, "loss": 0.2916, "step": 4041 }, { "epoch": 1.3047742039630332, "grad_norm": 0.361630833368451, "learning_rate": 9.6989991579096e-06, "loss": 0.2915, "step": 4042 }, { "epoch": 1.305097057992655, "grad_norm": 0.4296235843819704, "learning_rate": 9.69644673848385e-06, "loss": 0.3646, "step": 4043 }, { "epoch": 1.305419912022277, "grad_norm": 0.5508123605721263, "learning_rate": 9.693894040771325e-06, "loss": 0.2608, "step": 4044 }, { "epoch": 1.3057427660518988, "grad_norm": 0.43968736077976706, "learning_rate": 9.691341065095443e-06, "loss": 0.3106, "step": 4045 }, { "epoch": 1.3060656200815206, "grad_norm": 0.44382820923157945, "learning_rate": 9.688787811779669e-06, "loss": 0.3579, "step": 4046 }, { "epoch": 1.3063884741111425, "grad_norm": 0.4023085391742305, "learning_rate": 9.686234281147491e-06, "loss": 0.2808, "step": 4047 }, { "epoch": 1.3067113281407643, "grad_norm": 0.49137611409260135, "learning_rate": 9.683680473522441e-06, "loss": 0.2931, "step": 4048 }, { "epoch": 1.3070341821703861, "grad_norm": 0.38129196569922164, "learning_rate": 9.681126389228083e-06, "loss": 0.273, "step": 4049 }, { "epoch": 1.307357036200008, "grad_norm": 0.47254160889615937, "learning_rate": 9.678572028588013e-06, "loss": 0.3681, "step": 4050 }, { "epoch": 1.3076798902296298, "grad_norm": 0.3734505203778906, "learning_rate": 9.67601739192587e-06, "loss": 0.3646, "step": 4051 }, { "epoch": 1.3080027442592517, "grad_norm": 0.39294020171225086, "learning_rate": 9.673462479565319e-06, "loss": 0.3431, "step": 4052 }, { "epoch": 1.3083255982888735, "grad_norm": 0.506133025562762, "learning_rate": 9.67090729183006e-06, "loss": 0.3376, "step": 4053 }, { "epoch": 1.3086484523184956, "grad_norm": 0.32699055540087324, "learning_rate": 9.66835182904384e-06, "loss": 0.2558, "step": 4054 }, { "epoch": 1.3089713063481174, "grad_norm": 0.37773234919711357, "learning_rate": 9.665796091530432e-06, "loss": 0.2919, "step": 4055 }, { "epoch": 1.3092941603777393, "grad_norm": 0.3810973492689309, "learning_rate": 9.66324007961364e-06, "loss": 0.3131, "step": 4056 }, { "epoch": 1.3096170144073611, "grad_norm": 0.38999125802332457, "learning_rate": 9.660683793617307e-06, "loss": 0.285, "step": 4057 }, { "epoch": 1.309939868436983, "grad_norm": 0.3533732224014281, "learning_rate": 9.658127233865314e-06, "loss": 0.3219, "step": 4058 }, { "epoch": 1.3102627224666048, "grad_norm": 0.6454728535439341, "learning_rate": 9.655570400681574e-06, "loss": 0.3172, "step": 4059 }, { "epoch": 1.3105855764962266, "grad_norm": 0.4536692521023451, "learning_rate": 9.653013294390032e-06, "loss": 0.3136, "step": 4060 }, { "epoch": 1.3109084305258485, "grad_norm": 0.3922237019523464, "learning_rate": 9.650455915314667e-06, "loss": 0.2612, "step": 4061 }, { "epoch": 1.3112312845554703, "grad_norm": 0.4103047947226342, "learning_rate": 9.647898263779503e-06, "loss": 0.355, "step": 4062 }, { "epoch": 1.3115541385850922, "grad_norm": 0.4354542046204427, "learning_rate": 9.645340340108584e-06, "loss": 0.3877, "step": 4063 }, { "epoch": 1.311876992614714, "grad_norm": 0.4579072038660712, "learning_rate": 9.642782144626e-06, "loss": 0.316, "step": 4064 }, { "epoch": 1.3121998466443359, "grad_norm": 0.43174796516914105, "learning_rate": 9.64022367765587e-06, "loss": 0.2867, "step": 4065 }, { "epoch": 1.3125227006739577, "grad_norm": 0.41889653582368436, "learning_rate": 9.637664939522346e-06, "loss": 0.2871, "step": 4066 }, { "epoch": 1.3128455547035798, "grad_norm": 0.39696440552037393, "learning_rate": 9.635105930549619e-06, "loss": 0.2883, "step": 4067 }, { "epoch": 1.3131684087332016, "grad_norm": 0.4093575806776495, "learning_rate": 9.63254665106191e-06, "loss": 0.3215, "step": 4068 }, { "epoch": 1.3134912627628235, "grad_norm": 0.4158498560362145, "learning_rate": 9.629987101383475e-06, "loss": 0.2619, "step": 4069 }, { "epoch": 1.3138141167924453, "grad_norm": 0.5021898451476989, "learning_rate": 9.62742728183861e-06, "loss": 0.2554, "step": 4070 }, { "epoch": 1.3141369708220672, "grad_norm": 0.33786242691626056, "learning_rate": 9.624867192751637e-06, "loss": 0.3086, "step": 4071 }, { "epoch": 1.314459824851689, "grad_norm": 0.36658313286771405, "learning_rate": 9.622306834446916e-06, "loss": 0.28, "step": 4072 }, { "epoch": 1.3147826788813108, "grad_norm": 0.3617931200983803, "learning_rate": 9.619746207248843e-06, "loss": 0.2994, "step": 4073 }, { "epoch": 1.3151055329109327, "grad_norm": 0.42526293676927723, "learning_rate": 9.617185311481845e-06, "loss": 0.2953, "step": 4074 }, { "epoch": 1.3154283869405545, "grad_norm": 0.3564631681570876, "learning_rate": 9.614624147470381e-06, "loss": 0.285, "step": 4075 }, { "epoch": 1.3157512409701764, "grad_norm": 0.3341419579942676, "learning_rate": 9.612062715538953e-06, "loss": 0.267, "step": 4076 }, { "epoch": 1.3160740949997982, "grad_norm": 0.43584673274993274, "learning_rate": 9.609501016012083e-06, "loss": 0.3489, "step": 4077 }, { "epoch": 1.31639694902942, "grad_norm": 0.40675688134006244, "learning_rate": 9.606939049214343e-06, "loss": 0.2796, "step": 4078 }, { "epoch": 1.316719803059042, "grad_norm": 0.42624438051228103, "learning_rate": 9.604376815470326e-06, "loss": 0.289, "step": 4079 }, { "epoch": 1.3170426570886637, "grad_norm": 0.3617050105698492, "learning_rate": 9.601814315104666e-06, "loss": 0.3635, "step": 4080 }, { "epoch": 1.3173655111182856, "grad_norm": 0.4075264212005966, "learning_rate": 9.599251548442023e-06, "loss": 0.2821, "step": 4081 }, { "epoch": 1.3176883651479074, "grad_norm": 0.3946783506012375, "learning_rate": 9.596688515807103e-06, "loss": 0.333, "step": 4082 }, { "epoch": 1.3180112191775293, "grad_norm": 0.4171830342442324, "learning_rate": 9.594125217524634e-06, "loss": 0.3633, "step": 4083 }, { "epoch": 1.3183340732071511, "grad_norm": 0.37300695554948776, "learning_rate": 9.591561653919383e-06, "loss": 0.2926, "step": 4084 }, { "epoch": 1.318656927236773, "grad_norm": 0.3737251606200803, "learning_rate": 9.58899782531615e-06, "loss": 0.2769, "step": 4085 }, { "epoch": 1.3189797812663948, "grad_norm": 0.38781915749656987, "learning_rate": 9.586433732039772e-06, "loss": 0.2914, "step": 4086 }, { "epoch": 1.3193026352960167, "grad_norm": 0.4418892158339284, "learning_rate": 9.583869374415112e-06, "loss": 0.3022, "step": 4087 }, { "epoch": 1.3196254893256387, "grad_norm": 0.3660371211951909, "learning_rate": 9.581304752767073e-06, "loss": 0.2647, "step": 4088 }, { "epoch": 1.3199483433552606, "grad_norm": 0.33879223108322665, "learning_rate": 9.578739867420585e-06, "loss": 0.3097, "step": 4089 }, { "epoch": 1.3202711973848824, "grad_norm": 0.4297549058051992, "learning_rate": 9.576174718700622e-06, "loss": 0.3035, "step": 4090 }, { "epoch": 1.3205940514145043, "grad_norm": 0.42954137119667524, "learning_rate": 9.573609306932179e-06, "loss": 0.2721, "step": 4091 }, { "epoch": 1.320916905444126, "grad_norm": 0.40778289865030104, "learning_rate": 9.571043632440292e-06, "loss": 0.3254, "step": 4092 }, { "epoch": 1.321239759473748, "grad_norm": 0.35765412555087184, "learning_rate": 9.56847769555003e-06, "loss": 0.3004, "step": 4093 }, { "epoch": 1.3215626135033698, "grad_norm": 0.4245118361386417, "learning_rate": 9.56591149658649e-06, "loss": 0.2955, "step": 4094 }, { "epoch": 1.3218854675329916, "grad_norm": 0.4390824425384509, "learning_rate": 9.56334503587481e-06, "loss": 0.2924, "step": 4095 }, { "epoch": 1.3222083215626135, "grad_norm": 0.34496396253457157, "learning_rate": 9.560778313740156e-06, "loss": 0.3278, "step": 4096 }, { "epoch": 1.3225311755922353, "grad_norm": 0.44122178883979624, "learning_rate": 9.558211330507729e-06, "loss": 0.3181, "step": 4097 }, { "epoch": 1.3228540296218572, "grad_norm": 0.3875123620495912, "learning_rate": 9.555644086502758e-06, "loss": 0.3458, "step": 4098 }, { "epoch": 1.323176883651479, "grad_norm": 0.5007545041112581, "learning_rate": 9.553076582050514e-06, "loss": 0.2968, "step": 4099 }, { "epoch": 1.3234997376811009, "grad_norm": 0.4125368852067124, "learning_rate": 9.550508817476291e-06, "loss": 0.3228, "step": 4100 }, { "epoch": 1.323822591710723, "grad_norm": 0.44442418343754403, "learning_rate": 9.547940793105427e-06, "loss": 0.3727, "step": 4101 }, { "epoch": 1.3241454457403448, "grad_norm": 0.4510945676800566, "learning_rate": 9.545372509263286e-06, "loss": 0.3041, "step": 4102 }, { "epoch": 1.3244682997699666, "grad_norm": 0.4129571527890083, "learning_rate": 9.542803966275262e-06, "loss": 0.2918, "step": 4103 }, { "epoch": 1.3247911537995885, "grad_norm": 0.387627628448923, "learning_rate": 9.54023516446679e-06, "loss": 0.3241, "step": 4104 }, { "epoch": 1.3251140078292103, "grad_norm": 0.40799519109367516, "learning_rate": 9.537666104163333e-06, "loss": 0.3306, "step": 4105 }, { "epoch": 1.3254368618588321, "grad_norm": 0.3876657982499184, "learning_rate": 9.535096785690384e-06, "loss": 0.2871, "step": 4106 }, { "epoch": 1.325759715888454, "grad_norm": 0.5444679258600718, "learning_rate": 9.532527209373478e-06, "loss": 0.2649, "step": 4107 }, { "epoch": 1.3260825699180758, "grad_norm": 0.4781462069767693, "learning_rate": 9.529957375538172e-06, "loss": 0.287, "step": 4108 }, { "epoch": 1.3264054239476977, "grad_norm": 0.37314188694727746, "learning_rate": 9.527387284510062e-06, "loss": 0.3573, "step": 4109 }, { "epoch": 1.3267282779773195, "grad_norm": 0.45889245993779526, "learning_rate": 9.524816936614775e-06, "loss": 0.2912, "step": 4110 }, { "epoch": 1.3270511320069414, "grad_norm": 0.5617246635454443, "learning_rate": 9.52224633217797e-06, "loss": 0.3389, "step": 4111 }, { "epoch": 1.3273739860365632, "grad_norm": 0.46987521089138906, "learning_rate": 9.519675471525337e-06, "loss": 0.3434, "step": 4112 }, { "epoch": 1.327696840066185, "grad_norm": 0.44191846462768636, "learning_rate": 9.517104354982606e-06, "loss": 0.3109, "step": 4113 }, { "epoch": 1.328019694095807, "grad_norm": 0.40206671353875995, "learning_rate": 9.514532982875532e-06, "loss": 0.2941, "step": 4114 }, { "epoch": 1.3283425481254287, "grad_norm": 0.410761779798601, "learning_rate": 9.511961355529902e-06, "loss": 0.3025, "step": 4115 }, { "epoch": 1.3286654021550506, "grad_norm": 0.3753134278691215, "learning_rate": 9.509389473271537e-06, "loss": 0.3446, "step": 4116 }, { "epoch": 1.3289882561846724, "grad_norm": 0.3747571056832815, "learning_rate": 9.506817336426296e-06, "loss": 0.3032, "step": 4117 }, { "epoch": 1.3293111102142943, "grad_norm": 0.33683760177260663, "learning_rate": 9.504244945320062e-06, "loss": 0.2848, "step": 4118 }, { "epoch": 1.3296339642439161, "grad_norm": 0.40635922892613724, "learning_rate": 9.50167230027875e-06, "loss": 0.3111, "step": 4119 }, { "epoch": 1.329956818273538, "grad_norm": 0.4335755761320625, "learning_rate": 9.499099401628318e-06, "loss": 0.3144, "step": 4120 }, { "epoch": 1.3302796723031598, "grad_norm": 0.446617882219543, "learning_rate": 9.496526249694743e-06, "loss": 0.2625, "step": 4121 }, { "epoch": 1.3306025263327819, "grad_norm": 0.36749168311336733, "learning_rate": 9.493952844804042e-06, "loss": 0.329, "step": 4122 }, { "epoch": 1.3309253803624037, "grad_norm": 0.4496085832268971, "learning_rate": 9.491379187282259e-06, "loss": 0.3329, "step": 4123 }, { "epoch": 1.3312482343920256, "grad_norm": 0.41795974673844377, "learning_rate": 9.488805277455478e-06, "loss": 0.3054, "step": 4124 }, { "epoch": 1.3315710884216474, "grad_norm": 0.40060565715926594, "learning_rate": 9.486231115649807e-06, "loss": 0.2885, "step": 4125 }, { "epoch": 1.3318939424512692, "grad_norm": 0.48790476233824526, "learning_rate": 9.483656702191387e-06, "loss": 0.3198, "step": 4126 }, { "epoch": 1.332216796480891, "grad_norm": 0.48204188357730654, "learning_rate": 9.481082037406397e-06, "loss": 0.3647, "step": 4127 }, { "epoch": 1.332539650510513, "grad_norm": 0.3858569601745114, "learning_rate": 9.478507121621038e-06, "loss": 0.2863, "step": 4128 }, { "epoch": 1.3328625045401348, "grad_norm": 0.4664346511482936, "learning_rate": 9.475931955161553e-06, "loss": 0.2788, "step": 4129 }, { "epoch": 1.3331853585697566, "grad_norm": 0.3419258540534685, "learning_rate": 9.47335653835421e-06, "loss": 0.2841, "step": 4130 }, { "epoch": 1.3335082125993785, "grad_norm": 0.34591403285112904, "learning_rate": 9.470780871525311e-06, "loss": 0.2902, "step": 4131 }, { "epoch": 1.3338310666290003, "grad_norm": 0.33949016559604184, "learning_rate": 9.468204955001189e-06, "loss": 0.2784, "step": 4132 }, { "epoch": 1.3341539206586221, "grad_norm": 0.34083520486839874, "learning_rate": 9.46562878910821e-06, "loss": 0.2949, "step": 4133 }, { "epoch": 1.3344767746882442, "grad_norm": 0.34953943267260856, "learning_rate": 9.46305237417277e-06, "loss": 0.3033, "step": 4134 }, { "epoch": 1.334799628717866, "grad_norm": 0.4219996461800626, "learning_rate": 9.4604757105213e-06, "loss": 0.3392, "step": 4135 }, { "epoch": 1.335122482747488, "grad_norm": 0.3987086043034209, "learning_rate": 9.457898798480256e-06, "loss": 0.283, "step": 4136 }, { "epoch": 1.3354453367771097, "grad_norm": 0.38464662448696824, "learning_rate": 9.455321638376128e-06, "loss": 0.2947, "step": 4137 }, { "epoch": 1.3357681908067316, "grad_norm": 0.4136751387057622, "learning_rate": 9.452744230535443e-06, "loss": 0.2907, "step": 4138 }, { "epoch": 1.3360910448363534, "grad_norm": 0.4106694836190524, "learning_rate": 9.45016657528475e-06, "loss": 0.2993, "step": 4139 }, { "epoch": 1.3364138988659753, "grad_norm": 0.5487437326489072, "learning_rate": 9.44758867295064e-06, "loss": 0.2695, "step": 4140 }, { "epoch": 1.3367367528955971, "grad_norm": 0.3816674247036338, "learning_rate": 9.44501052385973e-06, "loss": 0.3215, "step": 4141 }, { "epoch": 1.337059606925219, "grad_norm": 0.3316099241177053, "learning_rate": 9.442432128338662e-06, "loss": 0.2725, "step": 4142 }, { "epoch": 1.3373824609548408, "grad_norm": 0.3729654900774521, "learning_rate": 9.439853486714119e-06, "loss": 0.3221, "step": 4143 }, { "epoch": 1.3377053149844627, "grad_norm": 0.370439793363727, "learning_rate": 9.43727459931281e-06, "loss": 0.279, "step": 4144 }, { "epoch": 1.3380281690140845, "grad_norm": 0.5736435998698658, "learning_rate": 9.434695466461479e-06, "loss": 0.2835, "step": 4145 }, { "epoch": 1.3383510230437063, "grad_norm": 0.4938281035877988, "learning_rate": 9.432116088486892e-06, "loss": 0.3107, "step": 4146 }, { "epoch": 1.3386738770733282, "grad_norm": 0.41790980454689075, "learning_rate": 9.429536465715862e-06, "loss": 0.3011, "step": 4147 }, { "epoch": 1.33899673110295, "grad_norm": 0.3536337687339252, "learning_rate": 9.426956598475218e-06, "loss": 0.3232, "step": 4148 }, { "epoch": 1.3393195851325719, "grad_norm": 0.41832237408723716, "learning_rate": 9.424376487091824e-06, "loss": 0.3064, "step": 4149 }, { "epoch": 1.3396424391621937, "grad_norm": 0.3813085430874868, "learning_rate": 9.421796131892579e-06, "loss": 0.3414, "step": 4150 }, { "epoch": 1.3399652931918156, "grad_norm": 0.43190570191809774, "learning_rate": 9.419215533204413e-06, "loss": 0.3121, "step": 4151 }, { "epoch": 1.3402881472214374, "grad_norm": 0.4558391888159078, "learning_rate": 9.416634691354282e-06, "loss": 0.2832, "step": 4152 }, { "epoch": 1.3406110012510593, "grad_norm": 0.4662734109466188, "learning_rate": 9.414053606669175e-06, "loss": 0.3602, "step": 4153 }, { "epoch": 1.340933855280681, "grad_norm": 0.4547509158114245, "learning_rate": 9.41147227947611e-06, "loss": 0.2336, "step": 4154 }, { "epoch": 1.341256709310303, "grad_norm": 0.40238193634659003, "learning_rate": 9.40889071010214e-06, "loss": 0.2835, "step": 4155 }, { "epoch": 1.341579563339925, "grad_norm": 0.3595534352785594, "learning_rate": 9.406308898874345e-06, "loss": 0.332, "step": 4156 }, { "epoch": 1.3419024173695469, "grad_norm": 0.43667429651065964, "learning_rate": 9.40372684611984e-06, "loss": 0.2764, "step": 4157 }, { "epoch": 1.3422252713991687, "grad_norm": 0.45611081505353795, "learning_rate": 9.401144552165765e-06, "loss": 0.3101, "step": 4158 }, { "epoch": 1.3425481254287905, "grad_norm": 0.41329759783636233, "learning_rate": 9.398562017339292e-06, "loss": 0.2552, "step": 4159 }, { "epoch": 1.3428709794584124, "grad_norm": 0.4468859703199162, "learning_rate": 9.395979241967625e-06, "loss": 0.322, "step": 4160 }, { "epoch": 1.3431938334880342, "grad_norm": 0.49428593461502796, "learning_rate": 9.393396226378001e-06, "loss": 0.269, "step": 4161 }, { "epoch": 1.343516687517656, "grad_norm": 0.3436544503105802, "learning_rate": 9.39081297089768e-06, "loss": 0.3182, "step": 4162 }, { "epoch": 1.343839541547278, "grad_norm": 0.3878481427995147, "learning_rate": 9.38822947585396e-06, "loss": 0.2636, "step": 4163 }, { "epoch": 1.3441623955768998, "grad_norm": 0.40182864959650627, "learning_rate": 9.385645741574167e-06, "loss": 0.3262, "step": 4164 }, { "epoch": 1.3444852496065216, "grad_norm": 0.3983966426238038, "learning_rate": 9.383061768385657e-06, "loss": 0.3077, "step": 4165 }, { "epoch": 1.3448081036361434, "grad_norm": 0.4825928900479603, "learning_rate": 9.380477556615812e-06, "loss": 0.3353, "step": 4166 }, { "epoch": 1.3451309576657653, "grad_norm": 0.4391603662712502, "learning_rate": 9.37789310659205e-06, "loss": 0.274, "step": 4167 }, { "epoch": 1.3454538116953874, "grad_norm": 0.5014253514053475, "learning_rate": 9.375308418641816e-06, "loss": 0.3167, "step": 4168 }, { "epoch": 1.3457766657250092, "grad_norm": 0.3611903849481267, "learning_rate": 9.37272349309259e-06, "loss": 0.2872, "step": 4169 }, { "epoch": 1.346099519754631, "grad_norm": 0.40502771790468817, "learning_rate": 9.370138330271876e-06, "loss": 0.2911, "step": 4170 }, { "epoch": 1.346422373784253, "grad_norm": 0.3561343416663645, "learning_rate": 9.367552930507208e-06, "loss": 0.3163, "step": 4171 }, { "epoch": 1.3467452278138747, "grad_norm": 0.39630712095700343, "learning_rate": 9.364967294126159e-06, "loss": 0.2861, "step": 4172 }, { "epoch": 1.3470680818434966, "grad_norm": 0.4477591126459777, "learning_rate": 9.36238142145632e-06, "loss": 0.295, "step": 4173 }, { "epoch": 1.3473909358731184, "grad_norm": 0.47193851795375497, "learning_rate": 9.359795312825321e-06, "loss": 0.2926, "step": 4174 }, { "epoch": 1.3477137899027403, "grad_norm": 0.440315581050006, "learning_rate": 9.357208968560817e-06, "loss": 0.29, "step": 4175 }, { "epoch": 1.348036643932362, "grad_norm": 0.3726689833169158, "learning_rate": 9.354622388990495e-06, "loss": 0.3103, "step": 4176 }, { "epoch": 1.348359497961984, "grad_norm": 0.37073798710845407, "learning_rate": 9.35203557444207e-06, "loss": 0.3328, "step": 4177 }, { "epoch": 1.3486823519916058, "grad_norm": 0.3615379234417217, "learning_rate": 9.349448525243286e-06, "loss": 0.3024, "step": 4178 }, { "epoch": 1.3490052060212276, "grad_norm": 0.44239288507455454, "learning_rate": 9.34686124172192e-06, "loss": 0.2952, "step": 4179 }, { "epoch": 1.3493280600508495, "grad_norm": 0.424215371826306, "learning_rate": 9.344273724205781e-06, "loss": 0.3069, "step": 4180 }, { "epoch": 1.3496509140804713, "grad_norm": 0.418215286130865, "learning_rate": 9.341685973022701e-06, "loss": 0.3044, "step": 4181 }, { "epoch": 1.3499737681100932, "grad_norm": 0.4533500823560545, "learning_rate": 9.339097988500543e-06, "loss": 0.261, "step": 4182 }, { "epoch": 1.350296622139715, "grad_norm": 0.3982212116237538, "learning_rate": 9.336509770967202e-06, "loss": 0.3175, "step": 4183 }, { "epoch": 1.3506194761693369, "grad_norm": 0.4647010374017341, "learning_rate": 9.333921320750602e-06, "loss": 0.3352, "step": 4184 }, { "epoch": 1.3509423301989587, "grad_norm": 0.34676439669471165, "learning_rate": 9.331332638178696e-06, "loss": 0.2969, "step": 4185 }, { "epoch": 1.3512651842285806, "grad_norm": 0.3695434998761941, "learning_rate": 9.328743723579468e-06, "loss": 0.2936, "step": 4186 }, { "epoch": 1.3515880382582024, "grad_norm": 0.42895657257395203, "learning_rate": 9.326154577280926e-06, "loss": 0.3416, "step": 4187 }, { "epoch": 1.3519108922878242, "grad_norm": 0.41782737776730444, "learning_rate": 9.323565199611116e-06, "loss": 0.3091, "step": 4188 }, { "epoch": 1.3522337463174463, "grad_norm": 0.4756253046504519, "learning_rate": 9.320975590898106e-06, "loss": 0.2988, "step": 4189 }, { "epoch": 1.3525566003470681, "grad_norm": 0.3967458599151208, "learning_rate": 9.318385751469997e-06, "loss": 0.3285, "step": 4190 }, { "epoch": 1.35287945437669, "grad_norm": 0.38417470146603167, "learning_rate": 9.315795681654917e-06, "loss": 0.3116, "step": 4191 }, { "epoch": 1.3532023084063118, "grad_norm": 0.3097778659686604, "learning_rate": 9.313205381781024e-06, "loss": 0.2615, "step": 4192 }, { "epoch": 1.3535251624359337, "grad_norm": 0.3770727212139519, "learning_rate": 9.310614852176506e-06, "loss": 0.2913, "step": 4193 }, { "epoch": 1.3538480164655555, "grad_norm": 0.4765029332057367, "learning_rate": 9.30802409316958e-06, "loss": 0.3128, "step": 4194 }, { "epoch": 1.3541708704951774, "grad_norm": 0.37619533810721, "learning_rate": 9.30543310508849e-06, "loss": 0.264, "step": 4195 }, { "epoch": 1.3544937245247992, "grad_norm": 0.37239907806817646, "learning_rate": 9.302841888261514e-06, "loss": 0.3525, "step": 4196 }, { "epoch": 1.354816578554421, "grad_norm": 0.3572212846895684, "learning_rate": 9.300250443016954e-06, "loss": 0.2898, "step": 4197 }, { "epoch": 1.355139432584043, "grad_norm": 0.3929881577352556, "learning_rate": 9.29765876968314e-06, "loss": 0.3244, "step": 4198 }, { "epoch": 1.3554622866136647, "grad_norm": 0.3825050969134004, "learning_rate": 9.295066868588435e-06, "loss": 0.2668, "step": 4199 }, { "epoch": 1.3557851406432866, "grad_norm": 0.4201566650204854, "learning_rate": 9.292474740061233e-06, "loss": 0.3433, "step": 4200 }, { "epoch": 1.3561079946729084, "grad_norm": 0.4754411750143791, "learning_rate": 9.289882384429945e-06, "loss": 0.3299, "step": 4201 }, { "epoch": 1.3564308487025305, "grad_norm": 0.35834949980482195, "learning_rate": 9.287289802023026e-06, "loss": 0.3132, "step": 4202 }, { "epoch": 1.3567537027321523, "grad_norm": 0.3990260019182314, "learning_rate": 9.284696993168952e-06, "loss": 0.2983, "step": 4203 }, { "epoch": 1.3570765567617742, "grad_norm": 0.3811167544518815, "learning_rate": 9.282103958196223e-06, "loss": 0.263, "step": 4204 }, { "epoch": 1.357399410791396, "grad_norm": 0.48371916037809926, "learning_rate": 9.279510697433378e-06, "loss": 0.3022, "step": 4205 }, { "epoch": 1.3577222648210179, "grad_norm": 0.39647380283164324, "learning_rate": 9.276917211208979e-06, "loss": 0.3091, "step": 4206 }, { "epoch": 1.3580451188506397, "grad_norm": 0.4676078625135591, "learning_rate": 9.274323499851616e-06, "loss": 0.2965, "step": 4207 }, { "epoch": 1.3583679728802616, "grad_norm": 0.39398758412966295, "learning_rate": 9.271729563689909e-06, "loss": 0.2972, "step": 4208 }, { "epoch": 1.3586908269098834, "grad_norm": 0.46726288613113814, "learning_rate": 9.269135403052506e-06, "loss": 0.3253, "step": 4209 }, { "epoch": 1.3590136809395053, "grad_norm": 0.4400023490833067, "learning_rate": 9.266541018268082e-06, "loss": 0.2631, "step": 4210 }, { "epoch": 1.359336534969127, "grad_norm": 0.4181828889641979, "learning_rate": 9.263946409665347e-06, "loss": 0.2563, "step": 4211 }, { "epoch": 1.359659388998749, "grad_norm": 0.3463334297154797, "learning_rate": 9.261351577573029e-06, "loss": 0.2998, "step": 4212 }, { "epoch": 1.3599822430283708, "grad_norm": 0.44451085143534463, "learning_rate": 9.258756522319892e-06, "loss": 0.3009, "step": 4213 }, { "epoch": 1.3603050970579926, "grad_norm": 0.36310375672695144, "learning_rate": 9.256161244234728e-06, "loss": 0.2557, "step": 4214 }, { "epoch": 1.3606279510876145, "grad_norm": 0.37328488907686597, "learning_rate": 9.253565743646349e-06, "loss": 0.2963, "step": 4215 }, { "epoch": 1.3609508051172363, "grad_norm": 0.36291650636714795, "learning_rate": 9.250970020883607e-06, "loss": 0.2895, "step": 4216 }, { "epoch": 1.3612736591468582, "grad_norm": 0.41431491929684444, "learning_rate": 9.248374076275373e-06, "loss": 0.2937, "step": 4217 }, { "epoch": 1.36159651317648, "grad_norm": 0.4050911727664202, "learning_rate": 9.245777910150552e-06, "loss": 0.308, "step": 4218 }, { "epoch": 1.3619193672061018, "grad_norm": 0.39518186040725506, "learning_rate": 9.243181522838076e-06, "loss": 0.2826, "step": 4219 }, { "epoch": 1.3622422212357237, "grad_norm": 0.4009522063147643, "learning_rate": 9.240584914666897e-06, "loss": 0.3102, "step": 4220 }, { "epoch": 1.3625650752653455, "grad_norm": 0.4628434097864514, "learning_rate": 9.237988085966012e-06, "loss": 0.2884, "step": 4221 }, { "epoch": 1.3628879292949674, "grad_norm": 0.5371263103772133, "learning_rate": 9.235391037064425e-06, "loss": 0.2808, "step": 4222 }, { "epoch": 1.3632107833245894, "grad_norm": 0.373328322447055, "learning_rate": 9.232793768291184e-06, "loss": 0.3092, "step": 4223 }, { "epoch": 1.3635336373542113, "grad_norm": 0.39662315657283426, "learning_rate": 9.230196279975361e-06, "loss": 0.322, "step": 4224 }, { "epoch": 1.3638564913838331, "grad_norm": 0.36832526938678756, "learning_rate": 9.22759857244605e-06, "loss": 0.3095, "step": 4225 }, { "epoch": 1.364179345413455, "grad_norm": 0.4792600007079326, "learning_rate": 9.22500064603238e-06, "loss": 0.3105, "step": 4226 }, { "epoch": 1.3645021994430768, "grad_norm": 0.38139369578728194, "learning_rate": 9.222402501063503e-06, "loss": 0.3148, "step": 4227 }, { "epoch": 1.3648250534726987, "grad_norm": 0.36529767314167855, "learning_rate": 9.219804137868602e-06, "loss": 0.3448, "step": 4228 }, { "epoch": 1.3651479075023205, "grad_norm": 0.39821555928960084, "learning_rate": 9.217205556776885e-06, "loss": 0.3241, "step": 4229 }, { "epoch": 1.3654707615319424, "grad_norm": 0.37234543621990485, "learning_rate": 9.214606758117588e-06, "loss": 0.3246, "step": 4230 }, { "epoch": 1.3657936155615642, "grad_norm": 0.43541393975952086, "learning_rate": 9.212007742219977e-06, "loss": 0.2774, "step": 4231 }, { "epoch": 1.366116469591186, "grad_norm": 0.5918746885396388, "learning_rate": 9.20940850941334e-06, "loss": 0.2897, "step": 4232 }, { "epoch": 1.3664393236208079, "grad_norm": 0.7165684286651419, "learning_rate": 9.206809060027e-06, "loss": 0.3148, "step": 4233 }, { "epoch": 1.3667621776504297, "grad_norm": 0.41233199009156274, "learning_rate": 9.204209394390303e-06, "loss": 0.3467, "step": 4234 }, { "epoch": 1.3670850316800516, "grad_norm": 0.41638491621434875, "learning_rate": 9.201609512832623e-06, "loss": 0.2859, "step": 4235 }, { "epoch": 1.3674078857096736, "grad_norm": 0.4606722523150908, "learning_rate": 9.199009415683361e-06, "loss": 0.3364, "step": 4236 }, { "epoch": 1.3677307397392955, "grad_norm": 0.4302055561189858, "learning_rate": 9.196409103271946e-06, "loss": 0.3119, "step": 4237 }, { "epoch": 1.3680535937689173, "grad_norm": 0.5712318625054145, "learning_rate": 9.193808575927833e-06, "loss": 0.2586, "step": 4238 }, { "epoch": 1.3683764477985392, "grad_norm": 0.43982556089637304, "learning_rate": 9.191207833980507e-06, "loss": 0.3148, "step": 4239 }, { "epoch": 1.368699301828161, "grad_norm": 0.464254421407675, "learning_rate": 9.188606877759474e-06, "loss": 0.3343, "step": 4240 }, { "epoch": 1.3690221558577829, "grad_norm": 0.7068922315587297, "learning_rate": 9.186005707594277e-06, "loss": 0.2575, "step": 4241 }, { "epoch": 1.3693450098874047, "grad_norm": 0.40893593077961954, "learning_rate": 9.183404323814477e-06, "loss": 0.2804, "step": 4242 }, { "epoch": 1.3696678639170266, "grad_norm": 0.4271060797757475, "learning_rate": 9.18080272674967e-06, "loss": 0.3084, "step": 4243 }, { "epoch": 1.3699907179466484, "grad_norm": 0.3576218979258643, "learning_rate": 9.178200916729472e-06, "loss": 0.3176, "step": 4244 }, { "epoch": 1.3703135719762702, "grad_norm": 0.6047939639213596, "learning_rate": 9.175598894083527e-06, "loss": 0.2678, "step": 4245 }, { "epoch": 1.370636426005892, "grad_norm": 0.4183377279300048, "learning_rate": 9.172996659141513e-06, "loss": 0.2802, "step": 4246 }, { "epoch": 1.370959280035514, "grad_norm": 0.40452338166270885, "learning_rate": 9.170394212233122e-06, "loss": 0.3163, "step": 4247 }, { "epoch": 1.3712821340651358, "grad_norm": 0.44298117858148245, "learning_rate": 9.167791553688088e-06, "loss": 0.3155, "step": 4248 }, { "epoch": 1.3716049880947576, "grad_norm": 0.43010630043248216, "learning_rate": 9.165188683836158e-06, "loss": 0.2925, "step": 4249 }, { "epoch": 1.3719278421243795, "grad_norm": 0.3656125707878657, "learning_rate": 9.162585603007116e-06, "loss": 0.2705, "step": 4250 }, { "epoch": 1.3722506961540013, "grad_norm": 0.40910144073375415, "learning_rate": 9.159982311530767e-06, "loss": 0.2718, "step": 4251 }, { "epoch": 1.3725735501836231, "grad_norm": 0.7273721834273214, "learning_rate": 9.157378809736948e-06, "loss": 0.305, "step": 4252 }, { "epoch": 1.372896404213245, "grad_norm": 0.36744358514145475, "learning_rate": 9.154775097955514e-06, "loss": 0.29, "step": 4253 }, { "epoch": 1.3732192582428668, "grad_norm": 0.4337432550880553, "learning_rate": 9.152171176516355e-06, "loss": 0.28, "step": 4254 }, { "epoch": 1.3735421122724887, "grad_norm": 0.43281929019229926, "learning_rate": 9.14956704574938e-06, "loss": 0.2969, "step": 4255 }, { "epoch": 1.3738649663021105, "grad_norm": 0.4401230107246821, "learning_rate": 9.146962705984533e-06, "loss": 0.3304, "step": 4256 }, { "epoch": 1.3741878203317326, "grad_norm": 0.49911068242196777, "learning_rate": 9.14435815755178e-06, "loss": 0.2853, "step": 4257 }, { "epoch": 1.3745106743613544, "grad_norm": 0.5458724200057784, "learning_rate": 9.141753400781114e-06, "loss": 0.3245, "step": 4258 }, { "epoch": 1.3748335283909763, "grad_norm": 0.4739891907972435, "learning_rate": 9.139148436002552e-06, "loss": 0.2657, "step": 4259 }, { "epoch": 1.3751563824205981, "grad_norm": 0.4374051312683464, "learning_rate": 9.136543263546138e-06, "loss": 0.2689, "step": 4260 }, { "epoch": 1.37547923645022, "grad_norm": 1.0990310265768923, "learning_rate": 9.13393788374195e-06, "loss": 0.2867, "step": 4261 }, { "epoch": 1.3758020904798418, "grad_norm": 0.39622578554871846, "learning_rate": 9.13133229692008e-06, "loss": 0.2588, "step": 4262 }, { "epoch": 1.3761249445094637, "grad_norm": 0.37412919473630907, "learning_rate": 9.128726503410655e-06, "loss": 0.3202, "step": 4263 }, { "epoch": 1.3764477985390855, "grad_norm": 0.325259249792721, "learning_rate": 9.126120503543827e-06, "loss": 0.3062, "step": 4264 }, { "epoch": 1.3767706525687073, "grad_norm": 0.3905353233209924, "learning_rate": 9.123514297649766e-06, "loss": 0.3251, "step": 4265 }, { "epoch": 1.3770935065983292, "grad_norm": 0.4301402439527536, "learning_rate": 9.120907886058683e-06, "loss": 0.2895, "step": 4266 }, { "epoch": 1.377416360627951, "grad_norm": 0.4538649813218356, "learning_rate": 9.1183012691008e-06, "loss": 0.3457, "step": 4267 }, { "epoch": 1.3777392146575729, "grad_norm": 0.3598577229159063, "learning_rate": 9.115694447106377e-06, "loss": 0.2928, "step": 4268 }, { "epoch": 1.3780620686871947, "grad_norm": 0.3090240849116393, "learning_rate": 9.113087420405691e-06, "loss": 0.2421, "step": 4269 }, { "epoch": 1.3783849227168168, "grad_norm": 0.46660067089619944, "learning_rate": 9.110480189329052e-06, "loss": 0.2802, "step": 4270 }, { "epoch": 1.3787077767464386, "grad_norm": 0.37387935706303105, "learning_rate": 9.107872754206791e-06, "loss": 0.3255, "step": 4271 }, { "epoch": 1.3790306307760605, "grad_norm": 0.4146219375427621, "learning_rate": 9.105265115369264e-06, "loss": 0.3515, "step": 4272 }, { "epoch": 1.3793534848056823, "grad_norm": 0.37930286346250175, "learning_rate": 9.102657273146857e-06, "loss": 0.3238, "step": 4273 }, { "epoch": 1.3796763388353042, "grad_norm": 0.31815773246539814, "learning_rate": 9.100049227869983e-06, "loss": 0.281, "step": 4274 }, { "epoch": 1.379999192864926, "grad_norm": 0.36833464761830637, "learning_rate": 9.097440979869075e-06, "loss": 0.2997, "step": 4275 }, { "epoch": 1.3803220468945478, "grad_norm": 0.4357820443274304, "learning_rate": 9.094832529474594e-06, "loss": 0.3759, "step": 4276 }, { "epoch": 1.3806449009241697, "grad_norm": 0.36180225056803245, "learning_rate": 9.092223877017028e-06, "loss": 0.303, "step": 4277 }, { "epoch": 1.3809677549537915, "grad_norm": 0.3532735963497924, "learning_rate": 9.089615022826887e-06, "loss": 0.2456, "step": 4278 }, { "epoch": 1.3812906089834134, "grad_norm": 0.3657426787810253, "learning_rate": 9.087005967234715e-06, "loss": 0.3613, "step": 4279 }, { "epoch": 1.3816134630130352, "grad_norm": 0.37683540431173457, "learning_rate": 9.08439671057107e-06, "loss": 0.2936, "step": 4280 }, { "epoch": 1.381936317042657, "grad_norm": 0.4900720909621998, "learning_rate": 9.081787253166545e-06, "loss": 0.3485, "step": 4281 }, { "epoch": 1.382259171072279, "grad_norm": 0.4917962787621494, "learning_rate": 9.07917759535175e-06, "loss": 0.3317, "step": 4282 }, { "epoch": 1.3825820251019008, "grad_norm": 0.38916664337657114, "learning_rate": 9.076567737457333e-06, "loss": 0.3062, "step": 4283 }, { "epoch": 1.3829048791315226, "grad_norm": 0.43779854910359056, "learning_rate": 9.073957679813953e-06, "loss": 0.3429, "step": 4284 }, { "epoch": 1.3832277331611444, "grad_norm": 0.5508947384409459, "learning_rate": 9.071347422752302e-06, "loss": 0.2746, "step": 4285 }, { "epoch": 1.3835505871907663, "grad_norm": 0.3802534478821231, "learning_rate": 9.068736966603096e-06, "loss": 0.2728, "step": 4286 }, { "epoch": 1.3838734412203881, "grad_norm": 0.4221048923128368, "learning_rate": 9.06612631169708e-06, "loss": 0.2361, "step": 4287 }, { "epoch": 1.38419629525001, "grad_norm": 0.33193476437212976, "learning_rate": 9.063515458365013e-06, "loss": 0.3501, "step": 4288 }, { "epoch": 1.3845191492796318, "grad_norm": 0.38015105288476275, "learning_rate": 9.060904406937692e-06, "loss": 0.3145, "step": 4289 }, { "epoch": 1.3848420033092537, "grad_norm": 0.40016104945447595, "learning_rate": 9.058293157745933e-06, "loss": 0.2851, "step": 4290 }, { "epoch": 1.3851648573388757, "grad_norm": 0.4035084578813799, "learning_rate": 9.055681711120577e-06, "loss": 0.2726, "step": 4291 }, { "epoch": 1.3854877113684976, "grad_norm": 0.3409560935426689, "learning_rate": 9.053070067392487e-06, "loss": 0.3085, "step": 4292 }, { "epoch": 1.3858105653981194, "grad_norm": 0.331684011177538, "learning_rate": 9.050458226892562e-06, "loss": 0.2571, "step": 4293 }, { "epoch": 1.3861334194277413, "grad_norm": 0.41174627261769353, "learning_rate": 9.047846189951713e-06, "loss": 0.314, "step": 4294 }, { "epoch": 1.386456273457363, "grad_norm": 0.35541895202127305, "learning_rate": 9.045233956900882e-06, "loss": 0.26, "step": 4295 }, { "epoch": 1.386779127486985, "grad_norm": 0.4265422206766371, "learning_rate": 9.042621528071039e-06, "loss": 0.3289, "step": 4296 }, { "epoch": 1.3871019815166068, "grad_norm": 0.46194308298022096, "learning_rate": 9.04000890379317e-06, "loss": 0.2771, "step": 4297 }, { "epoch": 1.3874248355462286, "grad_norm": 0.4108375039259963, "learning_rate": 9.037396084398295e-06, "loss": 0.2886, "step": 4298 }, { "epoch": 1.3877476895758505, "grad_norm": 0.4318046267317025, "learning_rate": 9.034783070217456e-06, "loss": 0.335, "step": 4299 }, { "epoch": 1.3880705436054723, "grad_norm": 0.4089463604700739, "learning_rate": 9.032169861581712e-06, "loss": 0.3464, "step": 4300 }, { "epoch": 1.3883933976350942, "grad_norm": 0.4013898944973582, "learning_rate": 9.029556458822159e-06, "loss": 0.2616, "step": 4301 }, { "epoch": 1.388716251664716, "grad_norm": 0.40628728946139925, "learning_rate": 9.026942862269908e-06, "loss": 0.3373, "step": 4302 }, { "epoch": 1.389039105694338, "grad_norm": 0.4336211201421499, "learning_rate": 9.024329072256098e-06, "loss": 0.3458, "step": 4303 }, { "epoch": 1.38936195972396, "grad_norm": 0.43660815756364146, "learning_rate": 9.021715089111894e-06, "loss": 0.3324, "step": 4304 }, { "epoch": 1.3896848137535818, "grad_norm": 0.44185907289382875, "learning_rate": 9.019100913168485e-06, "loss": 0.3034, "step": 4305 }, { "epoch": 1.3900076677832036, "grad_norm": 0.3991516053032082, "learning_rate": 9.01648654475708e-06, "loss": 0.3491, "step": 4306 }, { "epoch": 1.3903305218128255, "grad_norm": 0.4004964933192196, "learning_rate": 9.01387198420892e-06, "loss": 0.3497, "step": 4307 }, { "epoch": 1.3906533758424473, "grad_norm": 0.4620665148144194, "learning_rate": 9.011257231855263e-06, "loss": 0.3025, "step": 4308 }, { "epoch": 1.3909762298720691, "grad_norm": 0.3616721082960953, "learning_rate": 9.008642288027398e-06, "loss": 0.3103, "step": 4309 }, { "epoch": 1.391299083901691, "grad_norm": 0.41988678399783164, "learning_rate": 9.006027153056631e-06, "loss": 0.2835, "step": 4310 }, { "epoch": 1.3916219379313128, "grad_norm": 0.4957902873721203, "learning_rate": 9.003411827274297e-06, "loss": 0.2601, "step": 4311 }, { "epoch": 1.3919447919609347, "grad_norm": 0.4521499004320301, "learning_rate": 9.000796311011754e-06, "loss": 0.3054, "step": 4312 }, { "epoch": 1.3922676459905565, "grad_norm": 0.6993114317790826, "learning_rate": 8.998180604600388e-06, "loss": 0.3095, "step": 4313 }, { "epoch": 1.3925905000201784, "grad_norm": 0.3975878336062021, "learning_rate": 8.995564708371603e-06, "loss": 0.25, "step": 4314 }, { "epoch": 1.3929133540498002, "grad_norm": 0.3432805120429371, "learning_rate": 8.992948622656826e-06, "loss": 0.2823, "step": 4315 }, { "epoch": 1.393236208079422, "grad_norm": 0.43117458004728676, "learning_rate": 8.990332347787515e-06, "loss": 0.282, "step": 4316 }, { "epoch": 1.393559062109044, "grad_norm": 0.4072279843988659, "learning_rate": 8.987715884095148e-06, "loss": 0.3511, "step": 4317 }, { "epoch": 1.3938819161386657, "grad_norm": 0.3748478280156689, "learning_rate": 8.98509923191123e-06, "loss": 0.2967, "step": 4318 }, { "epoch": 1.3942047701682876, "grad_norm": 0.3683054260617161, "learning_rate": 8.982482391567282e-06, "loss": 0.3137, "step": 4319 }, { "epoch": 1.3945276241979094, "grad_norm": 0.3952257008953015, "learning_rate": 8.979865363394854e-06, "loss": 0.3241, "step": 4320 }, { "epoch": 1.3948504782275313, "grad_norm": 0.37029373260284093, "learning_rate": 8.977248147725525e-06, "loss": 0.2803, "step": 4321 }, { "epoch": 1.3951733322571531, "grad_norm": 0.4447935790504474, "learning_rate": 8.974630744890892e-06, "loss": 0.3101, "step": 4322 }, { "epoch": 1.395496186286775, "grad_norm": 0.37075664820776955, "learning_rate": 8.972013155222573e-06, "loss": 0.3368, "step": 4323 }, { "epoch": 1.395819040316397, "grad_norm": 0.3737516350912731, "learning_rate": 8.969395379052213e-06, "loss": 0.3445, "step": 4324 }, { "epoch": 1.3961418943460189, "grad_norm": 0.40220892560746746, "learning_rate": 8.966777416711483e-06, "loss": 0.3474, "step": 4325 }, { "epoch": 1.3964647483756407, "grad_norm": 0.5522320478261815, "learning_rate": 8.964159268532073e-06, "loss": 0.2805, "step": 4326 }, { "epoch": 1.3967876024052626, "grad_norm": 0.3512168957764429, "learning_rate": 8.961540934845698e-06, "loss": 0.3412, "step": 4327 }, { "epoch": 1.3971104564348844, "grad_norm": 0.36794723142821933, "learning_rate": 8.958922415984104e-06, "loss": 0.3131, "step": 4328 }, { "epoch": 1.3974333104645063, "grad_norm": 0.3911158966986543, "learning_rate": 8.956303712279047e-06, "loss": 0.2908, "step": 4329 }, { "epoch": 1.397756164494128, "grad_norm": 0.4083907350849082, "learning_rate": 8.953684824062315e-06, "loss": 0.2805, "step": 4330 }, { "epoch": 1.39807901852375, "grad_norm": 0.42014221052187684, "learning_rate": 8.951065751665717e-06, "loss": 0.2851, "step": 4331 }, { "epoch": 1.3984018725533718, "grad_norm": 0.3554197511302988, "learning_rate": 8.948446495421088e-06, "loss": 0.3026, "step": 4332 }, { "epoch": 1.3987247265829936, "grad_norm": 0.5334154631725755, "learning_rate": 8.945827055660282e-06, "loss": 0.2991, "step": 4333 }, { "epoch": 1.3990475806126155, "grad_norm": 0.39306240207392346, "learning_rate": 8.943207432715177e-06, "loss": 0.2905, "step": 4334 }, { "epoch": 1.3993704346422373, "grad_norm": 0.44047536260671916, "learning_rate": 8.940587626917678e-06, "loss": 0.3339, "step": 4335 }, { "epoch": 1.3996932886718592, "grad_norm": 0.4360369276560364, "learning_rate": 8.937967638599714e-06, "loss": 0.3151, "step": 4336 }, { "epoch": 1.4000161427014812, "grad_norm": 0.41080698571038765, "learning_rate": 8.935347468093228e-06, "loss": 0.3251, "step": 4337 }, { "epoch": 1.400338996731103, "grad_norm": 0.41137903029336303, "learning_rate": 8.932727115730195e-06, "loss": 0.3007, "step": 4338 }, { "epoch": 1.400661850760725, "grad_norm": 0.404491064995332, "learning_rate": 8.93010658184261e-06, "loss": 0.3163, "step": 4339 }, { "epoch": 1.4009847047903468, "grad_norm": 0.3819529972660678, "learning_rate": 8.927485866762489e-06, "loss": 0.3033, "step": 4340 }, { "epoch": 1.4013075588199686, "grad_norm": 0.37482631050017007, "learning_rate": 8.924864970821876e-06, "loss": 0.294, "step": 4341 }, { "epoch": 1.4016304128495904, "grad_norm": 0.42658037981709146, "learning_rate": 8.922243894352833e-06, "loss": 0.3089, "step": 4342 }, { "epoch": 1.4019532668792123, "grad_norm": 0.4375604269995555, "learning_rate": 8.919622637687443e-06, "loss": 0.2727, "step": 4343 }, { "epoch": 1.4022761209088341, "grad_norm": 0.4371925689036011, "learning_rate": 8.917001201157825e-06, "loss": 0.3266, "step": 4344 }, { "epoch": 1.402598974938456, "grad_norm": 0.4208620586324842, "learning_rate": 8.914379585096104e-06, "loss": 0.3259, "step": 4345 }, { "epoch": 1.4029218289680778, "grad_norm": 0.4340356398434998, "learning_rate": 8.91175778983444e-06, "loss": 0.2762, "step": 4346 }, { "epoch": 1.4032446829976997, "grad_norm": 0.45197284272568905, "learning_rate": 8.909135815705003e-06, "loss": 0.2603, "step": 4347 }, { "epoch": 1.4035675370273215, "grad_norm": 0.3278533817307056, "learning_rate": 8.906513663040002e-06, "loss": 0.2924, "step": 4348 }, { "epoch": 1.4038903910569434, "grad_norm": 0.4060691804323152, "learning_rate": 8.903891332171658e-06, "loss": 0.2913, "step": 4349 }, { "epoch": 1.4042132450865652, "grad_norm": 0.4321828474475437, "learning_rate": 8.901268823432212e-06, "loss": 0.2992, "step": 4350 }, { "epoch": 1.404536099116187, "grad_norm": 0.3764094624673385, "learning_rate": 8.898646137153937e-06, "loss": 0.3182, "step": 4351 }, { "epoch": 1.4048589531458089, "grad_norm": 0.455963294691049, "learning_rate": 8.896023273669124e-06, "loss": 0.3851, "step": 4352 }, { "epoch": 1.4051818071754307, "grad_norm": 0.46725755974016253, "learning_rate": 8.893400233310082e-06, "loss": 0.2547, "step": 4353 }, { "epoch": 1.4055046612050526, "grad_norm": 0.35641988872882313, "learning_rate": 8.890777016409151e-06, "loss": 0.3573, "step": 4354 }, { "epoch": 1.4058275152346744, "grad_norm": 0.5143986930130281, "learning_rate": 8.888153623298688e-06, "loss": 0.2969, "step": 4355 }, { "epoch": 1.4061503692642963, "grad_norm": 0.4586949459480777, "learning_rate": 8.885530054311072e-06, "loss": 0.3423, "step": 4356 }, { "epoch": 1.406473223293918, "grad_norm": 0.42595798442307203, "learning_rate": 8.882906309778706e-06, "loss": 0.3054, "step": 4357 }, { "epoch": 1.4067960773235402, "grad_norm": 0.41122008339276706, "learning_rate": 8.880282390034013e-06, "loss": 0.3287, "step": 4358 }, { "epoch": 1.407118931353162, "grad_norm": 0.3804756722108746, "learning_rate": 8.877658295409445e-06, "loss": 0.2991, "step": 4359 }, { "epoch": 1.4074417853827839, "grad_norm": 0.3710110353780652, "learning_rate": 8.875034026237466e-06, "loss": 0.312, "step": 4360 }, { "epoch": 1.4077646394124057, "grad_norm": 0.42518718149016027, "learning_rate": 8.872409582850572e-06, "loss": 0.3029, "step": 4361 }, { "epoch": 1.4080874934420275, "grad_norm": 0.3608280359512845, "learning_rate": 8.869784965581272e-06, "loss": 0.3382, "step": 4362 }, { "epoch": 1.4084103474716494, "grad_norm": 0.44052280585224113, "learning_rate": 8.867160174762104e-06, "loss": 0.3361, "step": 4363 }, { "epoch": 1.4087332015012712, "grad_norm": 0.3396782177656894, "learning_rate": 8.864535210725621e-06, "loss": 0.3012, "step": 4364 }, { "epoch": 1.409056055530893, "grad_norm": 0.37674192400048767, "learning_rate": 8.86191007380441e-06, "loss": 0.3211, "step": 4365 }, { "epoch": 1.409378909560515, "grad_norm": 0.3813046599594592, "learning_rate": 8.859284764331066e-06, "loss": 0.262, "step": 4366 }, { "epoch": 1.4097017635901368, "grad_norm": 0.41027499220906555, "learning_rate": 8.856659282638215e-06, "loss": 0.29, "step": 4367 }, { "epoch": 1.4100246176197586, "grad_norm": 0.7552367062420079, "learning_rate": 8.8540336290585e-06, "loss": 0.2833, "step": 4368 }, { "epoch": 1.4103474716493805, "grad_norm": 0.3859311769454016, "learning_rate": 8.85140780392459e-06, "loss": 0.3327, "step": 4369 }, { "epoch": 1.4106703256790023, "grad_norm": 0.4562351020140495, "learning_rate": 8.848781807569173e-06, "loss": 0.3208, "step": 4370 }, { "epoch": 1.4109931797086244, "grad_norm": 0.3692034116451449, "learning_rate": 8.846155640324957e-06, "loss": 0.3318, "step": 4371 }, { "epoch": 1.4113160337382462, "grad_norm": 0.38101627163478086, "learning_rate": 8.843529302524675e-06, "loss": 0.2666, "step": 4372 }, { "epoch": 1.411638887767868, "grad_norm": 0.3839500075741705, "learning_rate": 8.840902794501083e-06, "loss": 0.3681, "step": 4373 }, { "epoch": 1.41196174179749, "grad_norm": 0.36318093459298134, "learning_rate": 8.838276116586949e-06, "loss": 0.2348, "step": 4374 }, { "epoch": 1.4122845958271117, "grad_norm": 0.39139790342447267, "learning_rate": 8.835649269115076e-06, "loss": 0.3035, "step": 4375 }, { "epoch": 1.4126074498567336, "grad_norm": 0.3718776488269724, "learning_rate": 8.833022252418282e-06, "loss": 0.2885, "step": 4376 }, { "epoch": 1.4129303038863554, "grad_norm": 0.47449735683178507, "learning_rate": 8.830395066829403e-06, "loss": 0.2919, "step": 4377 }, { "epoch": 1.4132531579159773, "grad_norm": 0.36136224836060704, "learning_rate": 8.827767712681301e-06, "loss": 0.2758, "step": 4378 }, { "epoch": 1.4135760119455991, "grad_norm": 0.5343697226672777, "learning_rate": 8.825140190306857e-06, "loss": 0.3476, "step": 4379 }, { "epoch": 1.413898865975221, "grad_norm": 0.41320097258532146, "learning_rate": 8.822512500038977e-06, "loss": 0.2814, "step": 4380 }, { "epoch": 1.4142217200048428, "grad_norm": 0.4183042672389554, "learning_rate": 8.819884642210585e-06, "loss": 0.2993, "step": 4381 }, { "epoch": 1.4145445740344647, "grad_norm": 0.4900502551244944, "learning_rate": 8.817256617154625e-06, "loss": 0.3323, "step": 4382 }, { "epoch": 1.4148674280640865, "grad_norm": 0.4071550476066493, "learning_rate": 8.814628425204065e-06, "loss": 0.3337, "step": 4383 }, { "epoch": 1.4151902820937083, "grad_norm": 0.41595881014484554, "learning_rate": 8.812000066691896e-06, "loss": 0.3778, "step": 4384 }, { "epoch": 1.4155131361233302, "grad_norm": 0.456974547365914, "learning_rate": 8.809371541951124e-06, "loss": 0.3242, "step": 4385 }, { "epoch": 1.415835990152952, "grad_norm": 0.48046422981994114, "learning_rate": 8.806742851314782e-06, "loss": 0.3164, "step": 4386 }, { "epoch": 1.4161588441825739, "grad_norm": 0.3921624620280396, "learning_rate": 8.80411399511592e-06, "loss": 0.3059, "step": 4387 }, { "epoch": 1.4164816982121957, "grad_norm": 0.36319456272405315, "learning_rate": 8.801484973687609e-06, "loss": 0.3278, "step": 4388 }, { "epoch": 1.4168045522418176, "grad_norm": 0.4645723497831168, "learning_rate": 8.798855787362946e-06, "loss": 0.328, "step": 4389 }, { "epoch": 1.4171274062714394, "grad_norm": 0.4559200224396201, "learning_rate": 8.79622643647504e-06, "loss": 0.3224, "step": 4390 }, { "epoch": 1.4174502603010612, "grad_norm": 0.508106583835416, "learning_rate": 8.793596921357031e-06, "loss": 0.3115, "step": 4391 }, { "epoch": 1.4177731143306833, "grad_norm": 0.3893239264930858, "learning_rate": 8.790967242342075e-06, "loss": 0.311, "step": 4392 }, { "epoch": 1.4180959683603052, "grad_norm": 0.45769164649768845, "learning_rate": 8.788337399763347e-06, "loss": 0.2834, "step": 4393 }, { "epoch": 1.418418822389927, "grad_norm": 0.40382020475376845, "learning_rate": 8.785707393954045e-06, "loss": 0.2943, "step": 4394 }, { "epoch": 1.4187416764195488, "grad_norm": 0.4646273349813767, "learning_rate": 8.783077225247385e-06, "loss": 0.2906, "step": 4395 }, { "epoch": 1.4190645304491707, "grad_norm": 0.4572191305478703, "learning_rate": 8.78044689397661e-06, "loss": 0.2765, "step": 4396 }, { "epoch": 1.4193873844787925, "grad_norm": 0.4510260158739782, "learning_rate": 8.777816400474973e-06, "loss": 0.3832, "step": 4397 }, { "epoch": 1.4197102385084144, "grad_norm": 0.3790375780747563, "learning_rate": 8.77518574507576e-06, "loss": 0.3045, "step": 4398 }, { "epoch": 1.4200330925380362, "grad_norm": 0.416641119990556, "learning_rate": 8.77255492811227e-06, "loss": 0.3377, "step": 4399 }, { "epoch": 1.420355946567658, "grad_norm": 0.39615749268728784, "learning_rate": 8.769923949917821e-06, "loss": 0.309, "step": 4400 }, { "epoch": 1.42067880059728, "grad_norm": 0.4232412097612776, "learning_rate": 8.767292810825758e-06, "loss": 0.3605, "step": 4401 }, { "epoch": 1.4210016546269018, "grad_norm": 0.4347239080950649, "learning_rate": 8.764661511169441e-06, "loss": 0.3213, "step": 4402 }, { "epoch": 1.4213245086565236, "grad_norm": 0.736215337556729, "learning_rate": 8.762030051282252e-06, "loss": 0.2993, "step": 4403 }, { "epoch": 1.4216473626861454, "grad_norm": 0.4138951544933165, "learning_rate": 8.759398431497593e-06, "loss": 0.2743, "step": 4404 }, { "epoch": 1.4219702167157675, "grad_norm": 0.4049946406359305, "learning_rate": 8.756766652148887e-06, "loss": 0.3611, "step": 4405 }, { "epoch": 1.4222930707453894, "grad_norm": 0.4170514584654508, "learning_rate": 8.754134713569577e-06, "loss": 0.2591, "step": 4406 }, { "epoch": 1.4226159247750112, "grad_norm": 0.4017972911374444, "learning_rate": 8.751502616093124e-06, "loss": 0.2794, "step": 4407 }, { "epoch": 1.422938778804633, "grad_norm": 0.3433456019141109, "learning_rate": 8.748870360053014e-06, "loss": 0.2786, "step": 4408 }, { "epoch": 1.4232616328342549, "grad_norm": 0.48909471049208014, "learning_rate": 8.746237945782751e-06, "loss": 0.2376, "step": 4409 }, { "epoch": 1.4235844868638767, "grad_norm": 0.47153056282292516, "learning_rate": 8.743605373615854e-06, "loss": 0.2964, "step": 4410 }, { "epoch": 1.4239073408934986, "grad_norm": 0.392539798621469, "learning_rate": 8.740972643885871e-06, "loss": 0.2933, "step": 4411 }, { "epoch": 1.4242301949231204, "grad_norm": 0.3706949517252576, "learning_rate": 8.738339756926362e-06, "loss": 0.3195, "step": 4412 }, { "epoch": 1.4245530489527423, "grad_norm": 0.42060317424547083, "learning_rate": 8.735706713070909e-06, "loss": 0.2929, "step": 4413 }, { "epoch": 1.424875902982364, "grad_norm": 0.5629766913889623, "learning_rate": 8.733073512653119e-06, "loss": 0.2911, "step": 4414 }, { "epoch": 1.425198757011986, "grad_norm": 0.4131263934586968, "learning_rate": 8.730440156006614e-06, "loss": 0.315, "step": 4415 }, { "epoch": 1.4255216110416078, "grad_norm": 0.5043168817079169, "learning_rate": 8.727806643465036e-06, "loss": 0.2985, "step": 4416 }, { "epoch": 1.4258444650712296, "grad_norm": 0.7627628875275082, "learning_rate": 8.725172975362046e-06, "loss": 0.246, "step": 4417 }, { "epoch": 1.4261673191008515, "grad_norm": 0.386294844975604, "learning_rate": 8.722539152031326e-06, "loss": 0.2689, "step": 4418 }, { "epoch": 1.4264901731304733, "grad_norm": 0.34194592972902693, "learning_rate": 8.719905173806581e-06, "loss": 0.3117, "step": 4419 }, { "epoch": 1.4268130271600952, "grad_norm": 0.40567749400145076, "learning_rate": 8.717271041021527e-06, "loss": 0.3193, "step": 4420 }, { "epoch": 1.427135881189717, "grad_norm": 0.32678470115438485, "learning_rate": 8.714636754009913e-06, "loss": 0.2984, "step": 4421 }, { "epoch": 1.4274587352193389, "grad_norm": 0.34953872264077684, "learning_rate": 8.712002313105493e-06, "loss": 0.3056, "step": 4422 }, { "epoch": 1.4277815892489607, "grad_norm": 0.34595334061059135, "learning_rate": 8.709367718642048e-06, "loss": 0.3268, "step": 4423 }, { "epoch": 1.4281044432785825, "grad_norm": 0.4512559483479055, "learning_rate": 8.70673297095338e-06, "loss": 0.2735, "step": 4424 }, { "epoch": 1.4284272973082044, "grad_norm": 0.38175447895622117, "learning_rate": 8.704098070373305e-06, "loss": 0.3295, "step": 4425 }, { "epoch": 1.4287501513378265, "grad_norm": 0.32267952620176266, "learning_rate": 8.70146301723566e-06, "loss": 0.2975, "step": 4426 }, { "epoch": 1.4290730053674483, "grad_norm": 0.3980978496835364, "learning_rate": 8.69882781187431e-06, "loss": 0.3245, "step": 4427 }, { "epoch": 1.4293958593970701, "grad_norm": 0.4648430998589981, "learning_rate": 8.696192454623125e-06, "loss": 0.2839, "step": 4428 }, { "epoch": 1.429718713426692, "grad_norm": 0.4604340963012418, "learning_rate": 8.693556945816004e-06, "loss": 0.3379, "step": 4429 }, { "epoch": 1.4300415674563138, "grad_norm": 0.4623042577629449, "learning_rate": 8.690921285786859e-06, "loss": 0.2565, "step": 4430 }, { "epoch": 1.4303644214859357, "grad_norm": 0.4217451730849772, "learning_rate": 8.688285474869632e-06, "loss": 0.3089, "step": 4431 }, { "epoch": 1.4306872755155575, "grad_norm": 0.41189536822439193, "learning_rate": 8.685649513398268e-06, "loss": 0.3035, "step": 4432 }, { "epoch": 1.4310101295451794, "grad_norm": 0.5085778505964188, "learning_rate": 8.683013401706744e-06, "loss": 0.3067, "step": 4433 }, { "epoch": 1.4313329835748012, "grad_norm": 0.3824058590510703, "learning_rate": 8.680377140129051e-06, "loss": 0.2892, "step": 4434 }, { "epoch": 1.431655837604423, "grad_norm": 0.4431884118605375, "learning_rate": 8.677740728999202e-06, "loss": 0.3268, "step": 4435 }, { "epoch": 1.431978691634045, "grad_norm": 0.46047852965308805, "learning_rate": 8.675104168651221e-06, "loss": 0.2955, "step": 4436 }, { "epoch": 1.4323015456636667, "grad_norm": 0.5712657277400468, "learning_rate": 8.672467459419163e-06, "loss": 0.3189, "step": 4437 }, { "epoch": 1.4326243996932888, "grad_norm": 0.40101508895600685, "learning_rate": 8.66983060163709e-06, "loss": 0.2762, "step": 4438 }, { "epoch": 1.4329472537229107, "grad_norm": 0.43890975311589814, "learning_rate": 8.667193595639093e-06, "loss": 0.3236, "step": 4439 }, { "epoch": 1.4332701077525325, "grad_norm": 0.5062147378178575, "learning_rate": 8.664556441759274e-06, "loss": 0.3113, "step": 4440 }, { "epoch": 1.4335929617821543, "grad_norm": 0.3534210246846756, "learning_rate": 8.661919140331758e-06, "loss": 0.2291, "step": 4441 }, { "epoch": 1.4339158158117762, "grad_norm": 0.4445062716366326, "learning_rate": 8.659281691690686e-06, "loss": 0.3372, "step": 4442 }, { "epoch": 1.434238669841398, "grad_norm": 0.5262656454560501, "learning_rate": 8.65664409617022e-06, "loss": 0.3478, "step": 4443 }, { "epoch": 1.4345615238710199, "grad_norm": 0.41966079021718206, "learning_rate": 8.65400635410454e-06, "loss": 0.2794, "step": 4444 }, { "epoch": 1.4348843779006417, "grad_norm": 0.4346096462156049, "learning_rate": 8.651368465827842e-06, "loss": 0.3189, "step": 4445 }, { "epoch": 1.4352072319302636, "grad_norm": 0.4709968875643757, "learning_rate": 8.648730431674349e-06, "loss": 0.299, "step": 4446 }, { "epoch": 1.4355300859598854, "grad_norm": 0.369982859173352, "learning_rate": 8.64609225197829e-06, "loss": 0.2887, "step": 4447 }, { "epoch": 1.4358529399895072, "grad_norm": 0.7039732693156198, "learning_rate": 8.643453927073921e-06, "loss": 0.3383, "step": 4448 }, { "epoch": 1.436175794019129, "grad_norm": 0.3756080474072849, "learning_rate": 8.640815457295515e-06, "loss": 0.3111, "step": 4449 }, { "epoch": 1.436498648048751, "grad_norm": 0.3793067453688195, "learning_rate": 8.638176842977361e-06, "loss": 0.2824, "step": 4450 }, { "epoch": 1.4368215020783728, "grad_norm": 0.3860726774644299, "learning_rate": 8.635538084453768e-06, "loss": 0.285, "step": 4451 }, { "epoch": 1.4371443561079946, "grad_norm": 0.45604645947562444, "learning_rate": 8.632899182059063e-06, "loss": 0.3499, "step": 4452 }, { "epoch": 1.4374672101376165, "grad_norm": 0.4360238350124805, "learning_rate": 8.630260136127593e-06, "loss": 0.2909, "step": 4453 }, { "epoch": 1.4377900641672383, "grad_norm": 0.5025411653129758, "learning_rate": 8.62762094699372e-06, "loss": 0.2797, "step": 4454 }, { "epoch": 1.4381129181968602, "grad_norm": 0.4055570839908785, "learning_rate": 8.624981614991825e-06, "loss": 0.2935, "step": 4455 }, { "epoch": 1.438435772226482, "grad_norm": 0.4436288336315676, "learning_rate": 8.62234214045631e-06, "loss": 0.2286, "step": 4456 }, { "epoch": 1.4387586262561038, "grad_norm": 0.4497332641739054, "learning_rate": 8.619702523721591e-06, "loss": 0.3367, "step": 4457 }, { "epoch": 1.4390814802857257, "grad_norm": 0.38818539261316304, "learning_rate": 8.617062765122104e-06, "loss": 0.3407, "step": 4458 }, { "epoch": 1.4394043343153478, "grad_norm": 0.37270463988819996, "learning_rate": 8.6144228649923e-06, "loss": 0.2993, "step": 4459 }, { "epoch": 1.4397271883449696, "grad_norm": 0.4702430690353436, "learning_rate": 8.611782823666658e-06, "loss": 0.2869, "step": 4460 }, { "epoch": 1.4400500423745914, "grad_norm": 0.4588726390962058, "learning_rate": 8.60914264147966e-06, "loss": 0.2859, "step": 4461 }, { "epoch": 1.4403728964042133, "grad_norm": 0.48209584878451534, "learning_rate": 8.60650231876582e-06, "loss": 0.2361, "step": 4462 }, { "epoch": 1.4406957504338351, "grad_norm": 0.3644291378286889, "learning_rate": 8.603861855859656e-06, "loss": 0.3228, "step": 4463 }, { "epoch": 1.441018604463457, "grad_norm": 0.4088913282017207, "learning_rate": 8.601221253095719e-06, "loss": 0.2742, "step": 4464 }, { "epoch": 1.4413414584930788, "grad_norm": 0.4672949918353162, "learning_rate": 8.598580510808563e-06, "loss": 0.268, "step": 4465 }, { "epoch": 1.4416643125227007, "grad_norm": 0.42855090118455985, "learning_rate": 8.595939629332771e-06, "loss": 0.3251, "step": 4466 }, { "epoch": 1.4419871665523225, "grad_norm": 0.5196374448321877, "learning_rate": 8.593298609002936e-06, "loss": 0.2835, "step": 4467 }, { "epoch": 1.4423100205819444, "grad_norm": 0.3379850734511448, "learning_rate": 8.590657450153672e-06, "loss": 0.2926, "step": 4468 }, { "epoch": 1.4426328746115662, "grad_norm": 0.35764241033168354, "learning_rate": 8.588016153119611e-06, "loss": 0.304, "step": 4469 }, { "epoch": 1.442955728641188, "grad_norm": 0.522165722239899, "learning_rate": 8.585374718235402e-06, "loss": 0.274, "step": 4470 }, { "epoch": 1.4432785826708099, "grad_norm": 0.34608495187131144, "learning_rate": 8.582733145835713e-06, "loss": 0.3282, "step": 4471 }, { "epoch": 1.443601436700432, "grad_norm": 0.4267619009693293, "learning_rate": 8.580091436255224e-06, "loss": 0.3331, "step": 4472 }, { "epoch": 1.4439242907300538, "grad_norm": 0.46906900723719147, "learning_rate": 8.577449589828638e-06, "loss": 0.3364, "step": 4473 }, { "epoch": 1.4442471447596756, "grad_norm": 0.5130488656534508, "learning_rate": 8.57480760689067e-06, "loss": 0.2868, "step": 4474 }, { "epoch": 1.4445699987892975, "grad_norm": 0.38919406147827973, "learning_rate": 8.57216548777606e-06, "loss": 0.254, "step": 4475 }, { "epoch": 1.4448928528189193, "grad_norm": 0.4518085265522073, "learning_rate": 8.56952323281956e-06, "loss": 0.3042, "step": 4476 }, { "epoch": 1.4452157068485412, "grad_norm": 0.4958515948971227, "learning_rate": 8.56688084235594e-06, "loss": 0.3199, "step": 4477 }, { "epoch": 1.445538560878163, "grad_norm": 0.4360943818560699, "learning_rate": 8.564238316719984e-06, "loss": 0.2964, "step": 4478 }, { "epoch": 1.4458614149077849, "grad_norm": 0.4913804753591689, "learning_rate": 8.5615956562465e-06, "loss": 0.3171, "step": 4479 }, { "epoch": 1.4461842689374067, "grad_norm": 0.40408249056926904, "learning_rate": 8.558952861270308e-06, "loss": 0.2675, "step": 4480 }, { "epoch": 1.4465071229670285, "grad_norm": 0.4229365234618887, "learning_rate": 8.55630993212625e-06, "loss": 0.341, "step": 4481 }, { "epoch": 1.4468299769966504, "grad_norm": 0.3449015081859907, "learning_rate": 8.553666869149173e-06, "loss": 0.2873, "step": 4482 }, { "epoch": 1.4471528310262722, "grad_norm": 0.3928276254814827, "learning_rate": 8.551023672673957e-06, "loss": 0.3111, "step": 4483 }, { "epoch": 1.447475685055894, "grad_norm": 0.3480870883057694, "learning_rate": 8.548380343035489e-06, "loss": 0.2382, "step": 4484 }, { "epoch": 1.447798539085516, "grad_norm": 0.4448057483043129, "learning_rate": 8.545736880568674e-06, "loss": 0.3658, "step": 4485 }, { "epoch": 1.4481213931151378, "grad_norm": 0.5140134823619329, "learning_rate": 8.543093285608436e-06, "loss": 0.2961, "step": 4486 }, { "epoch": 1.4484442471447596, "grad_norm": 0.4644297827559153, "learning_rate": 8.540449558489718e-06, "loss": 0.334, "step": 4487 }, { "epoch": 1.4487671011743815, "grad_norm": 0.4519323426125486, "learning_rate": 8.537805699547472e-06, "loss": 0.3285, "step": 4488 }, { "epoch": 1.4490899552040033, "grad_norm": 0.4915762847006343, "learning_rate": 8.535161709116674e-06, "loss": 0.2374, "step": 4489 }, { "epoch": 1.4494128092336251, "grad_norm": 0.4433441015802962, "learning_rate": 8.532517587532311e-06, "loss": 0.2604, "step": 4490 }, { "epoch": 1.449735663263247, "grad_norm": 0.43706248887487376, "learning_rate": 8.529873335129393e-06, "loss": 0.3299, "step": 4491 }, { "epoch": 1.4500585172928688, "grad_norm": 0.472731633369192, "learning_rate": 8.527228952242942e-06, "loss": 0.3792, "step": 4492 }, { "epoch": 1.450381371322491, "grad_norm": 0.6033048836622752, "learning_rate": 8.524584439207997e-06, "loss": 0.34, "step": 4493 }, { "epoch": 1.4507042253521127, "grad_norm": 0.4008070883335862, "learning_rate": 8.521939796359615e-06, "loss": 0.2795, "step": 4494 }, { "epoch": 1.4510270793817346, "grad_norm": 0.4615424686018311, "learning_rate": 8.519295024032867e-06, "loss": 0.2638, "step": 4495 }, { "epoch": 1.4513499334113564, "grad_norm": 0.6013265790715674, "learning_rate": 8.516650122562843e-06, "loss": 0.3249, "step": 4496 }, { "epoch": 1.4516727874409783, "grad_norm": 0.5114268757936533, "learning_rate": 8.514005092284648e-06, "loss": 0.3282, "step": 4497 }, { "epoch": 1.4519956414706001, "grad_norm": 0.42220892703410606, "learning_rate": 8.511359933533406e-06, "loss": 0.2983, "step": 4498 }, { "epoch": 1.452318495500222, "grad_norm": 0.4587538231605152, "learning_rate": 8.508714646644251e-06, "loss": 0.2643, "step": 4499 }, { "epoch": 1.4526413495298438, "grad_norm": 0.4892521917490578, "learning_rate": 8.50606923195234e-06, "loss": 0.287, "step": 4500 }, { "epoch": 1.4529642035594657, "grad_norm": 0.4585522559892106, "learning_rate": 8.503423689792843e-06, "loss": 0.2744, "step": 4501 }, { "epoch": 1.4532870575890875, "grad_norm": 0.38015468416099246, "learning_rate": 8.500778020500948e-06, "loss": 0.2543, "step": 4502 }, { "epoch": 1.4536099116187093, "grad_norm": 0.37836771985295303, "learning_rate": 8.498132224411855e-06, "loss": 0.3168, "step": 4503 }, { "epoch": 1.4539327656483312, "grad_norm": 0.4329327473838034, "learning_rate": 8.495486301860782e-06, "loss": 0.2426, "step": 4504 }, { "epoch": 1.454255619677953, "grad_norm": 0.4177408837391922, "learning_rate": 8.492840253182968e-06, "loss": 0.3589, "step": 4505 }, { "epoch": 1.454578473707575, "grad_norm": 0.6515269909053015, "learning_rate": 8.49019407871366e-06, "loss": 0.2728, "step": 4506 }, { "epoch": 1.454901327737197, "grad_norm": 0.36358134326950337, "learning_rate": 8.487547778788124e-06, "loss": 0.268, "step": 4507 }, { "epoch": 1.4552241817668188, "grad_norm": 0.5790227359042767, "learning_rate": 8.484901353741647e-06, "loss": 0.2714, "step": 4508 }, { "epoch": 1.4555470357964406, "grad_norm": 0.4797393843777724, "learning_rate": 8.482254803909523e-06, "loss": 0.3049, "step": 4509 }, { "epoch": 1.4558698898260625, "grad_norm": 0.5788267262163574, "learning_rate": 8.479608129627069e-06, "loss": 0.2821, "step": 4510 }, { "epoch": 1.4561927438556843, "grad_norm": 0.5896795480762069, "learning_rate": 8.476961331229614e-06, "loss": 0.314, "step": 4511 }, { "epoch": 1.4565155978853062, "grad_norm": 0.4028909322098267, "learning_rate": 8.474314409052505e-06, "loss": 0.299, "step": 4512 }, { "epoch": 1.456838451914928, "grad_norm": 1.0259592556209707, "learning_rate": 8.4716673634311e-06, "loss": 0.3045, "step": 4513 }, { "epoch": 1.4571613059445498, "grad_norm": 0.541620126128615, "learning_rate": 8.469020194700777e-06, "loss": 0.2585, "step": 4514 }, { "epoch": 1.4574841599741717, "grad_norm": 0.3878482217935233, "learning_rate": 8.466372903196931e-06, "loss": 0.267, "step": 4515 }, { "epoch": 1.4578070140037935, "grad_norm": 0.42188502238368103, "learning_rate": 8.46372548925497e-06, "loss": 0.3008, "step": 4516 }, { "epoch": 1.4581298680334154, "grad_norm": 0.4528771643726014, "learning_rate": 8.461077953210316e-06, "loss": 0.2678, "step": 4517 }, { "epoch": 1.4584527220630372, "grad_norm": 0.43025527585905315, "learning_rate": 8.458430295398408e-06, "loss": 0.2963, "step": 4518 }, { "epoch": 1.458775576092659, "grad_norm": 0.33893528521142047, "learning_rate": 8.455782516154702e-06, "loss": 0.2884, "step": 4519 }, { "epoch": 1.459098430122281, "grad_norm": 0.4578617995748289, "learning_rate": 8.453134615814669e-06, "loss": 0.3555, "step": 4520 }, { "epoch": 1.4594212841519028, "grad_norm": 0.46733688668299517, "learning_rate": 8.45048659471379e-06, "loss": 0.31, "step": 4521 }, { "epoch": 1.4597441381815246, "grad_norm": 0.44714550745451925, "learning_rate": 8.44783845318757e-06, "loss": 0.2823, "step": 4522 }, { "epoch": 1.4600669922111464, "grad_norm": 0.40971260573846247, "learning_rate": 8.445190191571522e-06, "loss": 0.3145, "step": 4523 }, { "epoch": 1.4603898462407683, "grad_norm": 0.43400748472825035, "learning_rate": 8.44254181020118e-06, "loss": 0.3254, "step": 4524 }, { "epoch": 1.4607127002703901, "grad_norm": 0.43044121886477976, "learning_rate": 8.439893309412087e-06, "loss": 0.3408, "step": 4525 }, { "epoch": 1.461035554300012, "grad_norm": 0.37747749969809674, "learning_rate": 8.437244689539807e-06, "loss": 0.3145, "step": 4526 }, { "epoch": 1.461358408329634, "grad_norm": 0.39379619348062306, "learning_rate": 8.434595950919917e-06, "loss": 0.2631, "step": 4527 }, { "epoch": 1.4616812623592559, "grad_norm": 0.39208992785162894, "learning_rate": 8.431947093888005e-06, "loss": 0.2915, "step": 4528 }, { "epoch": 1.4620041163888777, "grad_norm": 0.42279577936769486, "learning_rate": 8.42929811877968e-06, "loss": 0.2701, "step": 4529 }, { "epoch": 1.4623269704184996, "grad_norm": 0.44194190806494493, "learning_rate": 8.426649025930562e-06, "loss": 0.2912, "step": 4530 }, { "epoch": 1.4626498244481214, "grad_norm": 0.377131891458176, "learning_rate": 8.423999815676289e-06, "loss": 0.2987, "step": 4531 }, { "epoch": 1.4629726784777433, "grad_norm": 0.42714205805557326, "learning_rate": 8.421350488352513e-06, "loss": 0.3164, "step": 4532 }, { "epoch": 1.463295532507365, "grad_norm": 0.5841216640190633, "learning_rate": 8.418701044294899e-06, "loss": 0.2744, "step": 4533 }, { "epoch": 1.463618386536987, "grad_norm": 0.424726189091372, "learning_rate": 8.416051483839126e-06, "loss": 0.3923, "step": 4534 }, { "epoch": 1.4639412405666088, "grad_norm": 0.3521236782846259, "learning_rate": 8.413401807320894e-06, "loss": 0.2866, "step": 4535 }, { "epoch": 1.4642640945962306, "grad_norm": 0.3522127691343815, "learning_rate": 8.410752015075909e-06, "loss": 0.2797, "step": 4536 }, { "epoch": 1.4645869486258525, "grad_norm": 0.3635421678749695, "learning_rate": 8.408102107439897e-06, "loss": 0.2646, "step": 4537 }, { "epoch": 1.4649098026554743, "grad_norm": 0.45113270324201143, "learning_rate": 8.4054520847486e-06, "loss": 0.3035, "step": 4538 }, { "epoch": 1.4652326566850962, "grad_norm": 0.524630269286866, "learning_rate": 8.402801947337769e-06, "loss": 0.3632, "step": 4539 }, { "epoch": 1.4655555107147182, "grad_norm": 0.4354510604454367, "learning_rate": 8.400151695543175e-06, "loss": 0.3454, "step": 4540 }, { "epoch": 1.46587836474434, "grad_norm": 0.411397870997111, "learning_rate": 8.397501329700602e-06, "loss": 0.3186, "step": 4541 }, { "epoch": 1.466201218773962, "grad_norm": 0.4056708548442054, "learning_rate": 8.394850850145847e-06, "loss": 0.2706, "step": 4542 }, { "epoch": 1.4665240728035838, "grad_norm": 0.3535912019048364, "learning_rate": 8.392200257214721e-06, "loss": 0.3008, "step": 4543 }, { "epoch": 1.4668469268332056, "grad_norm": 0.3921611330827551, "learning_rate": 8.38954955124305e-06, "loss": 0.2891, "step": 4544 }, { "epoch": 1.4671697808628275, "grad_norm": 0.36398633646182743, "learning_rate": 8.386898732566679e-06, "loss": 0.3464, "step": 4545 }, { "epoch": 1.4674926348924493, "grad_norm": 0.4040695058813285, "learning_rate": 8.384247801521456e-06, "loss": 0.3051, "step": 4546 }, { "epoch": 1.4678154889220711, "grad_norm": 1.0009242413408914, "learning_rate": 8.381596758443256e-06, "loss": 0.2868, "step": 4547 }, { "epoch": 1.468138342951693, "grad_norm": 0.36770798930184606, "learning_rate": 8.378945603667959e-06, "loss": 0.2831, "step": 4548 }, { "epoch": 1.4684611969813148, "grad_norm": 0.43013508533438277, "learning_rate": 8.376294337531467e-06, "loss": 0.3061, "step": 4549 }, { "epoch": 1.4687840510109367, "grad_norm": 0.43618035194440596, "learning_rate": 8.373642960369688e-06, "loss": 0.3471, "step": 4550 }, { "epoch": 1.4691069050405585, "grad_norm": 0.42322652669589794, "learning_rate": 8.37099147251855e-06, "loss": 0.3134, "step": 4551 }, { "epoch": 1.4694297590701804, "grad_norm": 0.5352644801741274, "learning_rate": 8.36833987431399e-06, "loss": 0.3011, "step": 4552 }, { "epoch": 1.4697526130998022, "grad_norm": 0.43837997011556806, "learning_rate": 8.365688166091964e-06, "loss": 0.2812, "step": 4553 }, { "epoch": 1.470075467129424, "grad_norm": 0.4382878806339186, "learning_rate": 8.36303634818844e-06, "loss": 0.3417, "step": 4554 }, { "epoch": 1.470398321159046, "grad_norm": 0.3909363195412178, "learning_rate": 8.360384420939398e-06, "loss": 0.3085, "step": 4555 }, { "epoch": 1.4707211751886677, "grad_norm": 0.3958806404788338, "learning_rate": 8.357732384680837e-06, "loss": 0.2793, "step": 4556 }, { "epoch": 1.4710440292182896, "grad_norm": 0.4288010197797688, "learning_rate": 8.355080239748763e-06, "loss": 0.3122, "step": 4557 }, { "epoch": 1.4713668832479114, "grad_norm": 0.37911492969957117, "learning_rate": 8.352427986479199e-06, "loss": 0.2972, "step": 4558 }, { "epoch": 1.4716897372775333, "grad_norm": 0.44970859973358596, "learning_rate": 8.349775625208186e-06, "loss": 0.2803, "step": 4559 }, { "epoch": 1.4720125913071551, "grad_norm": 0.4564800717074677, "learning_rate": 8.34712315627177e-06, "loss": 0.3038, "step": 4560 }, { "epoch": 1.4723354453367772, "grad_norm": 0.374430340301022, "learning_rate": 8.344470580006016e-06, "loss": 0.2562, "step": 4561 }, { "epoch": 1.472658299366399, "grad_norm": 0.46310888169996983, "learning_rate": 8.341817896747e-06, "loss": 0.3286, "step": 4562 }, { "epoch": 1.4729811533960209, "grad_norm": 0.38615228475120006, "learning_rate": 8.33916510683082e-06, "loss": 0.3137, "step": 4563 }, { "epoch": 1.4733040074256427, "grad_norm": 0.35814007332853137, "learning_rate": 8.336512210593575e-06, "loss": 0.2635, "step": 4564 }, { "epoch": 1.4736268614552646, "grad_norm": 0.5158458288910012, "learning_rate": 8.333859208371385e-06, "loss": 0.2785, "step": 4565 }, { "epoch": 1.4739497154848864, "grad_norm": 0.3945592276150231, "learning_rate": 8.331206100500382e-06, "loss": 0.2725, "step": 4566 }, { "epoch": 1.4742725695145082, "grad_norm": 0.400001224773691, "learning_rate": 8.32855288731671e-06, "loss": 0.2881, "step": 4567 }, { "epoch": 1.47459542354413, "grad_norm": 0.38930085853484003, "learning_rate": 8.32589956915653e-06, "loss": 0.3587, "step": 4568 }, { "epoch": 1.474918277573752, "grad_norm": 0.5176812669715487, "learning_rate": 8.323246146356007e-06, "loss": 0.3295, "step": 4569 }, { "epoch": 1.4752411316033738, "grad_norm": 0.37416090028487936, "learning_rate": 8.320592619251335e-06, "loss": 0.3139, "step": 4570 }, { "epoch": 1.4755639856329956, "grad_norm": 0.41474999340508145, "learning_rate": 8.317938988178707e-06, "loss": 0.3139, "step": 4571 }, { "epoch": 1.4758868396626175, "grad_norm": 0.4349573043128109, "learning_rate": 8.315285253474337e-06, "loss": 0.254, "step": 4572 }, { "epoch": 1.4762096936922395, "grad_norm": 0.865757055365188, "learning_rate": 8.312631415474447e-06, "loss": 0.2699, "step": 4573 }, { "epoch": 1.4765325477218614, "grad_norm": 0.47435946288874326, "learning_rate": 8.309977474515274e-06, "loss": 0.3027, "step": 4574 }, { "epoch": 1.4768554017514832, "grad_norm": 0.4879512661922134, "learning_rate": 8.307323430933073e-06, "loss": 0.3393, "step": 4575 }, { "epoch": 1.477178255781105, "grad_norm": 0.5048633873705307, "learning_rate": 8.304669285064102e-06, "loss": 0.3647, "step": 4576 }, { "epoch": 1.477501109810727, "grad_norm": 0.3617317580463476, "learning_rate": 8.302015037244644e-06, "loss": 0.3429, "step": 4577 }, { "epoch": 1.4778239638403488, "grad_norm": 0.4672846630743793, "learning_rate": 8.299360687810985e-06, "loss": 0.3064, "step": 4578 }, { "epoch": 1.4781468178699706, "grad_norm": 0.37850017484079806, "learning_rate": 8.296706237099426e-06, "loss": 0.3337, "step": 4579 }, { "epoch": 1.4784696718995924, "grad_norm": 0.38947335642139047, "learning_rate": 8.294051685446285e-06, "loss": 0.3472, "step": 4580 }, { "epoch": 1.4787925259292143, "grad_norm": 0.42627403455251883, "learning_rate": 8.291397033187888e-06, "loss": 0.2683, "step": 4581 }, { "epoch": 1.4791153799588361, "grad_norm": 0.3410028708220137, "learning_rate": 8.288742280660576e-06, "loss": 0.2894, "step": 4582 }, { "epoch": 1.479438233988458, "grad_norm": 0.4023260639979658, "learning_rate": 8.286087428200706e-06, "loss": 0.3032, "step": 4583 }, { "epoch": 1.4797610880180798, "grad_norm": 0.3979778018001487, "learning_rate": 8.28343247614464e-06, "loss": 0.2751, "step": 4584 }, { "epoch": 1.4800839420477017, "grad_norm": 0.39313384077347735, "learning_rate": 8.280777424828755e-06, "loss": 0.2988, "step": 4585 }, { "epoch": 1.4804067960773235, "grad_norm": 0.5258888064029402, "learning_rate": 8.27812227458945e-06, "loss": 0.2741, "step": 4586 }, { "epoch": 1.4807296501069453, "grad_norm": 0.40957787415554014, "learning_rate": 8.275467025763122e-06, "loss": 0.3504, "step": 4587 }, { "epoch": 1.4810525041365672, "grad_norm": 0.44753922505129135, "learning_rate": 8.272811678686192e-06, "loss": 0.2919, "step": 4588 }, { "epoch": 1.481375358166189, "grad_norm": 0.6089603404430264, "learning_rate": 8.270156233695085e-06, "loss": 0.288, "step": 4589 }, { "epoch": 1.4816982121958109, "grad_norm": 0.45390407441096287, "learning_rate": 8.267500691126244e-06, "loss": 0.3048, "step": 4590 }, { "epoch": 1.4820210662254327, "grad_norm": 0.4301410858462345, "learning_rate": 8.264845051316123e-06, "loss": 0.3301, "step": 4591 }, { "epoch": 1.4823439202550546, "grad_norm": 0.39102311875485984, "learning_rate": 8.262189314601183e-06, "loss": 0.2775, "step": 4592 }, { "epoch": 1.4826667742846764, "grad_norm": 0.44317670691479555, "learning_rate": 8.259533481317912e-06, "loss": 0.2726, "step": 4593 }, { "epoch": 1.4829896283142985, "grad_norm": 0.41535901692142857, "learning_rate": 8.25687755180279e-06, "loss": 0.2955, "step": 4594 }, { "epoch": 1.4833124823439203, "grad_norm": 0.3409560552723054, "learning_rate": 8.254221526392327e-06, "loss": 0.3101, "step": 4595 }, { "epoch": 1.4836353363735422, "grad_norm": 0.3782030112471005, "learning_rate": 8.251565405423035e-06, "loss": 0.3765, "step": 4596 }, { "epoch": 1.483958190403164, "grad_norm": 0.36209824366671334, "learning_rate": 8.248909189231443e-06, "loss": 0.3096, "step": 4597 }, { "epoch": 1.4842810444327859, "grad_norm": 0.43749943156790533, "learning_rate": 8.246252878154086e-06, "loss": 0.3423, "step": 4598 }, { "epoch": 1.4846038984624077, "grad_norm": 0.5070525723564548, "learning_rate": 8.243596472527519e-06, "loss": 0.2827, "step": 4599 }, { "epoch": 1.4849267524920295, "grad_norm": 0.37858582419527564, "learning_rate": 8.2409399726883e-06, "loss": 0.2762, "step": 4600 }, { "epoch": 1.4852496065216514, "grad_norm": 0.5241015616941349, "learning_rate": 8.238283378973005e-06, "loss": 0.287, "step": 4601 }, { "epoch": 1.4855724605512732, "grad_norm": 0.4149723006356424, "learning_rate": 8.235626691718227e-06, "loss": 0.3289, "step": 4602 }, { "epoch": 1.485895314580895, "grad_norm": 0.4459270501346115, "learning_rate": 8.23296991126056e-06, "loss": 0.3321, "step": 4603 }, { "epoch": 1.486218168610517, "grad_norm": 0.3916877162114998, "learning_rate": 8.230313037936614e-06, "loss": 0.3629, "step": 4604 }, { "epoch": 1.4865410226401388, "grad_norm": 0.41186346050362993, "learning_rate": 8.22765607208301e-06, "loss": 0.3073, "step": 4605 }, { "epoch": 1.4868638766697606, "grad_norm": 0.3415119144360223, "learning_rate": 8.224999014036387e-06, "loss": 0.2629, "step": 4606 }, { "epoch": 1.4871867306993827, "grad_norm": 0.44895394243629977, "learning_rate": 8.222341864133385e-06, "loss": 0.3082, "step": 4607 }, { "epoch": 1.4875095847290045, "grad_norm": 0.4261955503435299, "learning_rate": 8.219684622710662e-06, "loss": 0.4004, "step": 4608 }, { "epoch": 1.4878324387586264, "grad_norm": 0.41060447709279135, "learning_rate": 8.21702729010489e-06, "loss": 0.2988, "step": 4609 }, { "epoch": 1.4881552927882482, "grad_norm": 0.4534016433483701, "learning_rate": 8.214369866652748e-06, "loss": 0.3061, "step": 4610 }, { "epoch": 1.48847814681787, "grad_norm": 0.4193050002169447, "learning_rate": 8.211712352690927e-06, "loss": 0.3613, "step": 4611 }, { "epoch": 1.488801000847492, "grad_norm": 0.35538514395211784, "learning_rate": 8.209054748556133e-06, "loss": 0.2398, "step": 4612 }, { "epoch": 1.4891238548771137, "grad_norm": 0.8184460683866095, "learning_rate": 8.206397054585077e-06, "loss": 0.2753, "step": 4613 }, { "epoch": 1.4894467089067356, "grad_norm": 0.5277960672649978, "learning_rate": 8.203739271114488e-06, "loss": 0.2783, "step": 4614 }, { "epoch": 1.4897695629363574, "grad_norm": 0.37973965000349025, "learning_rate": 8.201081398481102e-06, "loss": 0.2286, "step": 4615 }, { "epoch": 1.4900924169659793, "grad_norm": 0.45484833029770166, "learning_rate": 8.198423437021667e-06, "loss": 0.3208, "step": 4616 }, { "epoch": 1.4904152709956011, "grad_norm": 0.665900294208401, "learning_rate": 8.195765387072945e-06, "loss": 0.2661, "step": 4617 }, { "epoch": 1.490738125025223, "grad_norm": 0.38300189218464, "learning_rate": 8.193107248971708e-06, "loss": 0.3014, "step": 4618 }, { "epoch": 1.4910609790548448, "grad_norm": 0.38444868193663984, "learning_rate": 8.190449023054734e-06, "loss": 0.3268, "step": 4619 }, { "epoch": 1.4913838330844666, "grad_norm": 0.4622548324113647, "learning_rate": 8.187790709658824e-06, "loss": 0.2963, "step": 4620 }, { "epoch": 1.4917066871140885, "grad_norm": 0.3746292514220944, "learning_rate": 8.185132309120774e-06, "loss": 0.3677, "step": 4621 }, { "epoch": 1.4920295411437103, "grad_norm": 0.3525319946861922, "learning_rate": 8.182473821777407e-06, "loss": 0.3164, "step": 4622 }, { "epoch": 1.4923523951733322, "grad_norm": 0.4371414718153511, "learning_rate": 8.179815247965546e-06, "loss": 0.3199, "step": 4623 }, { "epoch": 1.492675249202954, "grad_norm": 0.4368370525910021, "learning_rate": 8.177156588022027e-06, "loss": 0.3161, "step": 4624 }, { "epoch": 1.4929981032325759, "grad_norm": 0.5418133631863354, "learning_rate": 8.174497842283703e-06, "loss": 0.3319, "step": 4625 }, { "epoch": 1.4933209572621977, "grad_norm": 0.4293001677276482, "learning_rate": 8.17183901108743e-06, "loss": 0.2881, "step": 4626 }, { "epoch": 1.4936438112918196, "grad_norm": 0.3471500393693948, "learning_rate": 8.16918009477008e-06, "loss": 0.2658, "step": 4627 }, { "epoch": 1.4939666653214416, "grad_norm": 0.4220996580274946, "learning_rate": 8.166521093668533e-06, "loss": 0.2953, "step": 4628 }, { "epoch": 1.4942895193510635, "grad_norm": 0.3262290488713423, "learning_rate": 8.16386200811968e-06, "loss": 0.2438, "step": 4629 }, { "epoch": 1.4946123733806853, "grad_norm": 0.42191240641556893, "learning_rate": 8.161202838460428e-06, "loss": 0.2648, "step": 4630 }, { "epoch": 1.4949352274103072, "grad_norm": 0.3534348460634519, "learning_rate": 8.158543585027687e-06, "loss": 0.2889, "step": 4631 }, { "epoch": 1.495258081439929, "grad_norm": 0.503702972606603, "learning_rate": 8.155884248158377e-06, "loss": 0.3656, "step": 4632 }, { "epoch": 1.4955809354695508, "grad_norm": 0.5414008849065514, "learning_rate": 8.153224828189438e-06, "loss": 0.2943, "step": 4633 }, { "epoch": 1.4959037894991727, "grad_norm": 0.42684737768335734, "learning_rate": 8.150565325457813e-06, "loss": 0.3361, "step": 4634 }, { "epoch": 1.4962266435287945, "grad_norm": 0.3817793204062549, "learning_rate": 8.147905740300456e-06, "loss": 0.2925, "step": 4635 }, { "epoch": 1.4965494975584164, "grad_norm": 0.42164832249387885, "learning_rate": 8.145246073054337e-06, "loss": 0.3201, "step": 4636 }, { "epoch": 1.4968723515880382, "grad_norm": 0.41956677949511995, "learning_rate": 8.142586324056426e-06, "loss": 0.3145, "step": 4637 }, { "epoch": 1.49719520561766, "grad_norm": 0.4041003555329844, "learning_rate": 8.139926493643715e-06, "loss": 0.2919, "step": 4638 }, { "epoch": 1.497518059647282, "grad_norm": 0.4250045744335496, "learning_rate": 8.137266582153198e-06, "loss": 0.3209, "step": 4639 }, { "epoch": 1.4978409136769038, "grad_norm": 0.46421176389378227, "learning_rate": 8.13460658992188e-06, "loss": 0.3028, "step": 4640 }, { "epoch": 1.4981637677065258, "grad_norm": 0.384338368291091, "learning_rate": 8.131946517286782e-06, "loss": 0.3074, "step": 4641 }, { "epoch": 1.4984866217361477, "grad_norm": 0.4487839636454667, "learning_rate": 8.129286364584932e-06, "loss": 0.259, "step": 4642 }, { "epoch": 1.4988094757657695, "grad_norm": 0.44329252258084034, "learning_rate": 8.126626132153366e-06, "loss": 0.3268, "step": 4643 }, { "epoch": 1.4991323297953913, "grad_norm": 0.3718060031924859, "learning_rate": 8.12396582032913e-06, "loss": 0.3125, "step": 4644 }, { "epoch": 1.4994551838250132, "grad_norm": 0.4619204516549742, "learning_rate": 8.121305429449287e-06, "loss": 0.286, "step": 4645 }, { "epoch": 1.499778037854635, "grad_norm": 0.42787299896593134, "learning_rate": 8.1186449598509e-06, "loss": 0.2966, "step": 4646 }, { "epoch": 1.5001008918842569, "grad_norm": 0.36961787187309847, "learning_rate": 8.115984411871046e-06, "loss": 0.2911, "step": 4647 }, { "epoch": 1.5004237459138787, "grad_norm": 0.3396414774589171, "learning_rate": 8.113323785846816e-06, "loss": 0.2781, "step": 4648 }, { "epoch": 1.5007465999435006, "grad_norm": 0.3572992029014934, "learning_rate": 8.11066308211531e-06, "loss": 0.3103, "step": 4649 }, { "epoch": 1.5010694539731224, "grad_norm": 0.508597687445984, "learning_rate": 8.10800230101363e-06, "loss": 0.2923, "step": 4650 }, { "epoch": 1.5013923080027443, "grad_norm": 0.5760159703451724, "learning_rate": 8.105341442878899e-06, "loss": 0.3505, "step": 4651 }, { "epoch": 1.501715162032366, "grad_norm": 0.4635625267333443, "learning_rate": 8.102680508048239e-06, "loss": 0.3203, "step": 4652 }, { "epoch": 1.502038016061988, "grad_norm": 0.37241887660071754, "learning_rate": 8.100019496858789e-06, "loss": 0.3123, "step": 4653 }, { "epoch": 1.5023608700916098, "grad_norm": 0.4067572712942634, "learning_rate": 8.097358409647693e-06, "loss": 0.3258, "step": 4654 }, { "epoch": 1.5026837241212316, "grad_norm": 0.35533255537765374, "learning_rate": 8.094697246752111e-06, "loss": 0.2835, "step": 4655 }, { "epoch": 1.5030065781508535, "grad_norm": 0.3979587671771664, "learning_rate": 8.092036008509205e-06, "loss": 0.3205, "step": 4656 }, { "epoch": 1.5033294321804753, "grad_norm": 0.38268940396194023, "learning_rate": 8.089374695256152e-06, "loss": 0.3212, "step": 4657 }, { "epoch": 1.5036522862100972, "grad_norm": 0.9113755756031472, "learning_rate": 8.086713307330139e-06, "loss": 0.2882, "step": 4658 }, { "epoch": 1.503975140239719, "grad_norm": 0.41116330569810317, "learning_rate": 8.084051845068358e-06, "loss": 0.3302, "step": 4659 }, { "epoch": 1.5042979942693409, "grad_norm": 0.5057767069516218, "learning_rate": 8.08139030880801e-06, "loss": 0.3017, "step": 4660 }, { "epoch": 1.5046208482989627, "grad_norm": 0.37231481237159214, "learning_rate": 8.078728698886312e-06, "loss": 0.3172, "step": 4661 }, { "epoch": 1.5049437023285845, "grad_norm": 0.39606236130777084, "learning_rate": 8.076067015640486e-06, "loss": 0.2969, "step": 4662 }, { "epoch": 1.5052665563582064, "grad_norm": 0.3788687786111445, "learning_rate": 8.073405259407758e-06, "loss": 0.3371, "step": 4663 }, { "epoch": 1.5055894103878285, "grad_norm": 0.396004719502515, "learning_rate": 8.070743430525378e-06, "loss": 0.2794, "step": 4664 }, { "epoch": 1.5059122644174503, "grad_norm": 0.36903863449796326, "learning_rate": 8.068081529330589e-06, "loss": 0.344, "step": 4665 }, { "epoch": 1.5062351184470721, "grad_norm": 0.5673342882834087, "learning_rate": 8.065419556160652e-06, "loss": 0.3143, "step": 4666 }, { "epoch": 1.506557972476694, "grad_norm": 0.40908407959415416, "learning_rate": 8.062757511352836e-06, "loss": 0.3325, "step": 4667 }, { "epoch": 1.5068808265063158, "grad_norm": 0.4279119102243752, "learning_rate": 8.06009539524442e-06, "loss": 0.3218, "step": 4668 }, { "epoch": 1.5072036805359377, "grad_norm": 0.4130522894747661, "learning_rate": 8.057433208172687e-06, "loss": 0.2827, "step": 4669 }, { "epoch": 1.5075265345655595, "grad_norm": 0.35618465860826565, "learning_rate": 8.054770950474938e-06, "loss": 0.2931, "step": 4670 }, { "epoch": 1.5078493885951814, "grad_norm": 0.41418874725574434, "learning_rate": 8.05210862248847e-06, "loss": 0.3542, "step": 4671 }, { "epoch": 1.5081722426248032, "grad_norm": 0.3855470518399411, "learning_rate": 8.049446224550598e-06, "loss": 0.2838, "step": 4672 }, { "epoch": 1.5084950966544253, "grad_norm": 0.44173683565894634, "learning_rate": 8.04678375699865e-06, "loss": 0.3301, "step": 4673 }, { "epoch": 1.5088179506840471, "grad_norm": 0.4141592055130073, "learning_rate": 8.04412122016995e-06, "loss": 0.3446, "step": 4674 }, { "epoch": 1.509140804713669, "grad_norm": 0.39132684424062225, "learning_rate": 8.041458614401844e-06, "loss": 0.2808, "step": 4675 }, { "epoch": 1.5094636587432908, "grad_norm": 0.4555647093603353, "learning_rate": 8.038795940031674e-06, "loss": 0.2673, "step": 4676 }, { "epoch": 1.5097865127729126, "grad_norm": 0.5148942787538106, "learning_rate": 8.036133197396803e-06, "loss": 0.3145, "step": 4677 }, { "epoch": 1.5101093668025345, "grad_norm": 0.40055442750420606, "learning_rate": 8.033470386834594e-06, "loss": 0.3103, "step": 4678 }, { "epoch": 1.5104322208321563, "grad_norm": 0.38650086350905594, "learning_rate": 8.030807508682418e-06, "loss": 0.3053, "step": 4679 }, { "epoch": 1.5107550748617782, "grad_norm": 0.37869194279078117, "learning_rate": 8.028144563277664e-06, "loss": 0.3046, "step": 4680 }, { "epoch": 1.5110779288914, "grad_norm": 0.5611645211943563, "learning_rate": 8.025481550957721e-06, "loss": 0.3248, "step": 4681 }, { "epoch": 1.5114007829210219, "grad_norm": 0.38773581690078185, "learning_rate": 8.02281847205999e-06, "loss": 0.3004, "step": 4682 }, { "epoch": 1.5117236369506437, "grad_norm": 0.3799989282169731, "learning_rate": 8.020155326921878e-06, "loss": 0.2737, "step": 4683 }, { "epoch": 1.5120464909802656, "grad_norm": 0.37610548388549664, "learning_rate": 8.017492115880803e-06, "loss": 0.3254, "step": 4684 }, { "epoch": 1.5123693450098874, "grad_norm": 0.45735193980862465, "learning_rate": 8.014828839274188e-06, "loss": 0.3732, "step": 4685 }, { "epoch": 1.5126921990395092, "grad_norm": 0.4305284191399195, "learning_rate": 8.012165497439468e-06, "loss": 0.3223, "step": 4686 }, { "epoch": 1.513015053069131, "grad_norm": 0.82661421686118, "learning_rate": 8.009502090714083e-06, "loss": 0.3035, "step": 4687 }, { "epoch": 1.513337907098753, "grad_norm": 0.36638099236759025, "learning_rate": 8.006838619435489e-06, "loss": 0.2275, "step": 4688 }, { "epoch": 1.5136607611283748, "grad_norm": 0.46367450131240223, "learning_rate": 8.004175083941138e-06, "loss": 0.2598, "step": 4689 }, { "epoch": 1.5139836151579966, "grad_norm": 0.41570733769779145, "learning_rate": 8.001511484568499e-06, "loss": 0.2449, "step": 4690 }, { "epoch": 1.5143064691876185, "grad_norm": 0.46654587299819106, "learning_rate": 7.998847821655045e-06, "loss": 0.3644, "step": 4691 }, { "epoch": 1.5146293232172403, "grad_norm": 0.4026040708105308, "learning_rate": 7.99618409553826e-06, "loss": 0.3199, "step": 4692 }, { "epoch": 1.5149521772468622, "grad_norm": 0.7189692941043119, "learning_rate": 7.993520306555632e-06, "loss": 0.2898, "step": 4693 }, { "epoch": 1.515275031276484, "grad_norm": 0.3567359887176412, "learning_rate": 7.99085645504466e-06, "loss": 0.3356, "step": 4694 }, { "epoch": 1.5155978853061058, "grad_norm": 0.35215999901459477, "learning_rate": 7.98819254134285e-06, "loss": 0.3029, "step": 4695 }, { "epoch": 1.5159207393357277, "grad_norm": 0.4217825435896577, "learning_rate": 7.98552856578772e-06, "loss": 0.3351, "step": 4696 }, { "epoch": 1.5162435933653495, "grad_norm": 0.3668970411888325, "learning_rate": 7.982864528716791e-06, "loss": 0.3212, "step": 4697 }, { "epoch": 1.5165664473949716, "grad_norm": 0.40215526567743626, "learning_rate": 7.980200430467589e-06, "loss": 0.3631, "step": 4698 }, { "epoch": 1.5168893014245934, "grad_norm": 0.3401364585140859, "learning_rate": 7.977536271377652e-06, "loss": 0.268, "step": 4699 }, { "epoch": 1.5172121554542153, "grad_norm": 0.35221780862554924, "learning_rate": 7.97487205178453e-06, "loss": 0.2675, "step": 4700 }, { "epoch": 1.5175350094838371, "grad_norm": 0.403884432425157, "learning_rate": 7.972207772025773e-06, "loss": 0.312, "step": 4701 }, { "epoch": 1.517857863513459, "grad_norm": 0.4371961665740783, "learning_rate": 7.969543432438937e-06, "loss": 0.2531, "step": 4702 }, { "epoch": 1.5181807175430808, "grad_norm": 0.40607032712296076, "learning_rate": 7.966879033361596e-06, "loss": 0.3774, "step": 4703 }, { "epoch": 1.5185035715727027, "grad_norm": 0.37428685840714276, "learning_rate": 7.964214575131326e-06, "loss": 0.2958, "step": 4704 }, { "epoch": 1.5188264256023245, "grad_norm": 0.4073837054090945, "learning_rate": 7.961550058085711e-06, "loss": 0.3126, "step": 4705 }, { "epoch": 1.5191492796319466, "grad_norm": 0.37499149668645954, "learning_rate": 7.958885482562337e-06, "loss": 0.3016, "step": 4706 }, { "epoch": 1.5194721336615684, "grad_norm": 0.38475702289054764, "learning_rate": 7.956220848898803e-06, "loss": 0.3163, "step": 4707 }, { "epoch": 1.5197949876911903, "grad_norm": 0.40099681463065173, "learning_rate": 7.953556157432719e-06, "loss": 0.3506, "step": 4708 }, { "epoch": 1.520117841720812, "grad_norm": 0.4264447091541766, "learning_rate": 7.950891408501692e-06, "loss": 0.3114, "step": 4709 }, { "epoch": 1.520440695750434, "grad_norm": 0.40339206055998883, "learning_rate": 7.948226602443343e-06, "loss": 0.3183, "step": 4710 }, { "epoch": 1.5207635497800558, "grad_norm": 0.34700533980608345, "learning_rate": 7.945561739595304e-06, "loss": 0.3522, "step": 4711 }, { "epoch": 1.5210864038096776, "grad_norm": 0.40173486105490097, "learning_rate": 7.942896820295204e-06, "loss": 0.3412, "step": 4712 }, { "epoch": 1.5214092578392995, "grad_norm": 0.4151603910310323, "learning_rate": 7.94023184488069e-06, "loss": 0.3368, "step": 4713 }, { "epoch": 1.5217321118689213, "grad_norm": 0.3629280930118052, "learning_rate": 7.937566813689404e-06, "loss": 0.273, "step": 4714 }, { "epoch": 1.5220549658985432, "grad_norm": 0.36240464282978646, "learning_rate": 7.934901727059008e-06, "loss": 0.3529, "step": 4715 }, { "epoch": 1.522377819928165, "grad_norm": 0.436109867766775, "learning_rate": 7.93223658532716e-06, "loss": 0.3184, "step": 4716 }, { "epoch": 1.5227006739577869, "grad_norm": 0.35301219657480326, "learning_rate": 7.929571388831532e-06, "loss": 0.3245, "step": 4717 }, { "epoch": 1.5230235279874087, "grad_norm": 0.9901876076236438, "learning_rate": 7.9269061379098e-06, "loss": 0.2474, "step": 4718 }, { "epoch": 1.5233463820170305, "grad_norm": 0.5004380119738863, "learning_rate": 7.924240832899649e-06, "loss": 0.3908, "step": 4719 }, { "epoch": 1.5236692360466524, "grad_norm": 0.4009880530794106, "learning_rate": 7.92157547413877e-06, "loss": 0.2753, "step": 4720 }, { "epoch": 1.5239920900762742, "grad_norm": 0.39513944176326843, "learning_rate": 7.918910061964857e-06, "loss": 0.335, "step": 4721 }, { "epoch": 1.524314944105896, "grad_norm": 0.37058008149408694, "learning_rate": 7.916244596715617e-06, "loss": 0.3303, "step": 4722 }, { "epoch": 1.524637798135518, "grad_norm": 0.40408774196414116, "learning_rate": 7.913579078728759e-06, "loss": 0.3201, "step": 4723 }, { "epoch": 1.5249606521651398, "grad_norm": 0.40176634250984145, "learning_rate": 7.910913508342002e-06, "loss": 0.2767, "step": 4724 }, { "epoch": 1.5252835061947616, "grad_norm": 0.3129637669495216, "learning_rate": 7.908247885893069e-06, "loss": 0.2526, "step": 4725 }, { "epoch": 1.5256063602243835, "grad_norm": 0.48797449864217407, "learning_rate": 7.905582211719688e-06, "loss": 0.2887, "step": 4726 }, { "epoch": 1.5259292142540053, "grad_norm": 0.3729632377919217, "learning_rate": 7.902916486159604e-06, "loss": 0.2696, "step": 4727 }, { "epoch": 1.5262520682836271, "grad_norm": 0.42188070930663385, "learning_rate": 7.900250709550556e-06, "loss": 0.2872, "step": 4728 }, { "epoch": 1.526574922313249, "grad_norm": 0.40440807780984445, "learning_rate": 7.897584882230293e-06, "loss": 0.3565, "step": 4729 }, { "epoch": 1.5268977763428708, "grad_norm": 0.42206482877368307, "learning_rate": 7.894919004536573e-06, "loss": 0.3035, "step": 4730 }, { "epoch": 1.5272206303724927, "grad_norm": 0.42972780183634907, "learning_rate": 7.892253076807157e-06, "loss": 0.2903, "step": 4731 }, { "epoch": 1.5275434844021147, "grad_norm": 0.4402081182589045, "learning_rate": 7.889587099379819e-06, "loss": 0.2984, "step": 4732 }, { "epoch": 1.5278663384317366, "grad_norm": 0.3982213711557079, "learning_rate": 7.886921072592333e-06, "loss": 0.3253, "step": 4733 }, { "epoch": 1.5281891924613584, "grad_norm": 0.41878804907156847, "learning_rate": 7.884254996782476e-06, "loss": 0.2761, "step": 4734 }, { "epoch": 1.5285120464909803, "grad_norm": 0.4426467405216565, "learning_rate": 7.881588872288042e-06, "loss": 0.3434, "step": 4735 }, { "epoch": 1.5288349005206021, "grad_norm": 0.41202243839262864, "learning_rate": 7.878922699446823e-06, "loss": 0.332, "step": 4736 }, { "epoch": 1.529157754550224, "grad_norm": 0.37278017150417503, "learning_rate": 7.876256478596621e-06, "loss": 0.3495, "step": 4737 }, { "epoch": 1.5294806085798458, "grad_norm": 0.337120454097079, "learning_rate": 7.873590210075241e-06, "loss": 0.3036, "step": 4738 }, { "epoch": 1.5298034626094676, "grad_norm": 0.362620933403799, "learning_rate": 7.870923894220496e-06, "loss": 0.2574, "step": 4739 }, { "epoch": 1.5301263166390897, "grad_norm": 0.3700937258206615, "learning_rate": 7.868257531370202e-06, "loss": 0.2812, "step": 4740 }, { "epoch": 1.5304491706687116, "grad_norm": 0.40098776861339497, "learning_rate": 7.865591121862188e-06, "loss": 0.3414, "step": 4741 }, { "epoch": 1.5307720246983334, "grad_norm": 0.32890681539114497, "learning_rate": 7.862924666034281e-06, "loss": 0.2579, "step": 4742 }, { "epoch": 1.5310948787279552, "grad_norm": 0.37770404305104854, "learning_rate": 7.860258164224318e-06, "loss": 0.2891, "step": 4743 }, { "epoch": 1.531417732757577, "grad_norm": 0.3523190778950852, "learning_rate": 7.857591616770145e-06, "loss": 0.3163, "step": 4744 }, { "epoch": 1.531740586787199, "grad_norm": 0.3721509113944539, "learning_rate": 7.854925024009605e-06, "loss": 0.2712, "step": 4745 }, { "epoch": 1.5320634408168208, "grad_norm": 0.37382555859849964, "learning_rate": 7.852258386280552e-06, "loss": 0.2911, "step": 4746 }, { "epoch": 1.5323862948464426, "grad_norm": 0.44587961171591556, "learning_rate": 7.849591703920848e-06, "loss": 0.276, "step": 4747 }, { "epoch": 1.5327091488760645, "grad_norm": 0.39077578028387755, "learning_rate": 7.846924977268357e-06, "loss": 0.2874, "step": 4748 }, { "epoch": 1.5330320029056863, "grad_norm": 0.3636162176695034, "learning_rate": 7.844258206660947e-06, "loss": 0.294, "step": 4749 }, { "epoch": 1.5333548569353082, "grad_norm": 0.45475700943084857, "learning_rate": 7.841591392436499e-06, "loss": 0.3372, "step": 4750 }, { "epoch": 1.53367771096493, "grad_norm": 0.4037387463542772, "learning_rate": 7.838924534932891e-06, "loss": 0.2897, "step": 4751 }, { "epoch": 1.5340005649945518, "grad_norm": 0.4444258961593315, "learning_rate": 7.83625763448801e-06, "loss": 0.248, "step": 4752 }, { "epoch": 1.5343234190241737, "grad_norm": 0.3972496523482467, "learning_rate": 7.83359069143975e-06, "loss": 0.303, "step": 4753 }, { "epoch": 1.5346462730537955, "grad_norm": 0.3972076712892697, "learning_rate": 7.83092370612601e-06, "loss": 0.2991, "step": 4754 }, { "epoch": 1.5349691270834174, "grad_norm": 0.42709308402385404, "learning_rate": 7.828256678884692e-06, "loss": 0.3414, "step": 4755 }, { "epoch": 1.5352919811130392, "grad_norm": 0.39740216010635226, "learning_rate": 7.825589610053705e-06, "loss": 0.302, "step": 4756 }, { "epoch": 1.535614835142661, "grad_norm": 0.3619519420143342, "learning_rate": 7.822922499970961e-06, "loss": 0.3008, "step": 4757 }, { "epoch": 1.535937689172283, "grad_norm": 0.43803585532190714, "learning_rate": 7.82025534897438e-06, "loss": 0.2606, "step": 4758 }, { "epoch": 1.5362605432019047, "grad_norm": 0.42976575937185435, "learning_rate": 7.817588157401888e-06, "loss": 0.307, "step": 4759 }, { "epoch": 1.5365833972315266, "grad_norm": 0.39839279580812925, "learning_rate": 7.814920925591416e-06, "loss": 0.3253, "step": 4760 }, { "epoch": 1.5369062512611484, "grad_norm": 0.36847213558427044, "learning_rate": 7.812253653880895e-06, "loss": 0.2841, "step": 4761 }, { "epoch": 1.5372291052907703, "grad_norm": 0.3973888607735394, "learning_rate": 7.809586342608268e-06, "loss": 0.3263, "step": 4762 }, { "epoch": 1.5375519593203921, "grad_norm": 0.38395897109656335, "learning_rate": 7.806918992111475e-06, "loss": 0.3059, "step": 4763 }, { "epoch": 1.537874813350014, "grad_norm": 0.6578243874768249, "learning_rate": 7.804251602728472e-06, "loss": 0.3385, "step": 4764 }, { "epoch": 1.538197667379636, "grad_norm": 0.5816686646063752, "learning_rate": 7.801584174797205e-06, "loss": 0.3633, "step": 4765 }, { "epoch": 1.5385205214092579, "grad_norm": 0.41584692593870176, "learning_rate": 7.79891670865564e-06, "loss": 0.3105, "step": 4766 }, { "epoch": 1.5388433754388797, "grad_norm": 0.4229225356124637, "learning_rate": 7.796249204641741e-06, "loss": 0.2579, "step": 4767 }, { "epoch": 1.5391662294685016, "grad_norm": 0.4143932637706731, "learning_rate": 7.793581663093475e-06, "loss": 0.2957, "step": 4768 }, { "epoch": 1.5394890834981234, "grad_norm": 0.41078442689444006, "learning_rate": 7.790914084348818e-06, "loss": 0.3519, "step": 4769 }, { "epoch": 1.5398119375277453, "grad_norm": 0.3576637695413339, "learning_rate": 7.788246468745746e-06, "loss": 0.349, "step": 4770 }, { "epoch": 1.540134791557367, "grad_norm": 0.40365006828686334, "learning_rate": 7.785578816622244e-06, "loss": 0.3281, "step": 4771 }, { "epoch": 1.540457645586989, "grad_norm": 0.3961060492896476, "learning_rate": 7.782911128316298e-06, "loss": 0.3344, "step": 4772 }, { "epoch": 1.5407804996166108, "grad_norm": 0.39635881551213803, "learning_rate": 7.7802434041659e-06, "loss": 0.2619, "step": 4773 }, { "epoch": 1.5411033536462329, "grad_norm": 0.40645517704544415, "learning_rate": 7.777575644509053e-06, "loss": 0.306, "step": 4774 }, { "epoch": 1.5414262076758547, "grad_norm": 0.4560590528595242, "learning_rate": 7.77490784968375e-06, "loss": 0.3205, "step": 4775 }, { "epoch": 1.5417490617054765, "grad_norm": 0.41742953897046575, "learning_rate": 7.772240020028005e-06, "loss": 0.2969, "step": 4776 }, { "epoch": 1.5420719157350984, "grad_norm": 0.3528727765091835, "learning_rate": 7.76957215587982e-06, "loss": 0.2893, "step": 4777 }, { "epoch": 1.5423947697647202, "grad_norm": 0.39588760046257254, "learning_rate": 7.766904257577216e-06, "loss": 0.2709, "step": 4778 }, { "epoch": 1.542717623794342, "grad_norm": 0.39868526872848503, "learning_rate": 7.76423632545821e-06, "loss": 0.3035, "step": 4779 }, { "epoch": 1.543040477823964, "grad_norm": 0.4161459989878952, "learning_rate": 7.761568359860827e-06, "loss": 0.289, "step": 4780 }, { "epoch": 1.5433633318535858, "grad_norm": 0.39117942621694796, "learning_rate": 7.758900361123091e-06, "loss": 0.3095, "step": 4781 }, { "epoch": 1.5436861858832076, "grad_norm": 0.3668835863113809, "learning_rate": 7.756232329583035e-06, "loss": 0.2972, "step": 4782 }, { "epoch": 1.5440090399128295, "grad_norm": 0.3909235560407688, "learning_rate": 7.7535642655787e-06, "loss": 0.2899, "step": 4783 }, { "epoch": 1.5443318939424513, "grad_norm": 0.3963978956596621, "learning_rate": 7.750896169448117e-06, "loss": 0.329, "step": 4784 }, { "epoch": 1.5446547479720731, "grad_norm": 0.5704965386383044, "learning_rate": 7.748228041529337e-06, "loss": 0.3671, "step": 4785 }, { "epoch": 1.544977602001695, "grad_norm": 0.38475511003336454, "learning_rate": 7.745559882160407e-06, "loss": 0.3037, "step": 4786 }, { "epoch": 1.5453004560313168, "grad_norm": 0.3922171538239974, "learning_rate": 7.742891691679377e-06, "loss": 0.3085, "step": 4787 }, { "epoch": 1.5456233100609387, "grad_norm": 0.3744496568851844, "learning_rate": 7.740223470424303e-06, "loss": 0.2673, "step": 4788 }, { "epoch": 1.5459461640905605, "grad_norm": 0.4057752105348047, "learning_rate": 7.737555218733247e-06, "loss": 0.3094, "step": 4789 }, { "epoch": 1.5462690181201824, "grad_norm": 0.38821623372391273, "learning_rate": 7.734886936944271e-06, "loss": 0.2882, "step": 4790 }, { "epoch": 1.5465918721498042, "grad_norm": 0.38784723651948955, "learning_rate": 7.732218625395446e-06, "loss": 0.3594, "step": 4791 }, { "epoch": 1.546914726179426, "grad_norm": 0.4299846601842098, "learning_rate": 7.72955028442484e-06, "loss": 0.2233, "step": 4792 }, { "epoch": 1.547237580209048, "grad_norm": 0.4784500229431601, "learning_rate": 7.726881914370529e-06, "loss": 0.3605, "step": 4793 }, { "epoch": 1.5475604342386697, "grad_norm": 0.4521739935756854, "learning_rate": 7.724213515570592e-06, "loss": 0.3564, "step": 4794 }, { "epoch": 1.5478832882682916, "grad_norm": 0.36303067135237105, "learning_rate": 7.72154508836311e-06, "loss": 0.2766, "step": 4795 }, { "epoch": 1.5482061422979134, "grad_norm": 0.42339032077086314, "learning_rate": 7.71887663308617e-06, "loss": 0.3126, "step": 4796 }, { "epoch": 1.5485289963275353, "grad_norm": 0.41363987839576444, "learning_rate": 7.716208150077863e-06, "loss": 0.3227, "step": 4797 }, { "epoch": 1.5488518503571571, "grad_norm": 0.31211258972264666, "learning_rate": 7.713539639676282e-06, "loss": 0.2911, "step": 4798 }, { "epoch": 1.5491747043867792, "grad_norm": 0.4280919260372956, "learning_rate": 7.710871102219521e-06, "loss": 0.287, "step": 4799 }, { "epoch": 1.549497558416401, "grad_norm": 0.4175012075607156, "learning_rate": 7.708202538045685e-06, "loss": 0.2559, "step": 4800 }, { "epoch": 1.5498204124460229, "grad_norm": 0.4765302992451974, "learning_rate": 7.70553394749287e-06, "loss": 0.3039, "step": 4801 }, { "epoch": 1.5501432664756447, "grad_norm": 0.3694101230437422, "learning_rate": 7.702865330899189e-06, "loss": 0.3584, "step": 4802 }, { "epoch": 1.5504661205052666, "grad_norm": 0.3732403030371425, "learning_rate": 7.70019668860275e-06, "loss": 0.2843, "step": 4803 }, { "epoch": 1.5507889745348884, "grad_norm": 0.35013379214280715, "learning_rate": 7.697528020941665e-06, "loss": 0.2777, "step": 4804 }, { "epoch": 1.5511118285645102, "grad_norm": 0.37508861343793615, "learning_rate": 7.694859328254054e-06, "loss": 0.3044, "step": 4805 }, { "epoch": 1.551434682594132, "grad_norm": 0.4197515851455872, "learning_rate": 7.692190610878032e-06, "loss": 0.2493, "step": 4806 }, { "epoch": 1.551757536623754, "grad_norm": 0.4045530705458125, "learning_rate": 7.689521869151726e-06, "loss": 0.3016, "step": 4807 }, { "epoch": 1.552080390653376, "grad_norm": 0.42455785345418745, "learning_rate": 7.68685310341326e-06, "loss": 0.28, "step": 4808 }, { "epoch": 1.5524032446829978, "grad_norm": 0.40615922967890083, "learning_rate": 7.684184314000763e-06, "loss": 0.2734, "step": 4809 }, { "epoch": 1.5527260987126197, "grad_norm": 0.49014747046994495, "learning_rate": 7.681515501252367e-06, "loss": 0.3155, "step": 4810 }, { "epoch": 1.5530489527422415, "grad_norm": 0.3599401425710048, "learning_rate": 7.678846665506205e-06, "loss": 0.3248, "step": 4811 }, { "epoch": 1.5533718067718634, "grad_norm": 0.41121872782404884, "learning_rate": 7.676177807100419e-06, "loss": 0.2782, "step": 4812 }, { "epoch": 1.5536946608014852, "grad_norm": 0.30005004092318166, "learning_rate": 7.673508926373147e-06, "loss": 0.2386, "step": 4813 }, { "epoch": 1.554017514831107, "grad_norm": 0.3609704157681345, "learning_rate": 7.670840023662532e-06, "loss": 0.2837, "step": 4814 }, { "epoch": 1.554340368860729, "grad_norm": 0.3472822433509479, "learning_rate": 7.668171099306723e-06, "loss": 0.2978, "step": 4815 }, { "epoch": 1.5546632228903507, "grad_norm": 0.45264623349512795, "learning_rate": 7.665502153643866e-06, "loss": 0.3073, "step": 4816 }, { "epoch": 1.5549860769199726, "grad_norm": 0.4070797888113552, "learning_rate": 7.662833187012114e-06, "loss": 0.2356, "step": 4817 }, { "epoch": 1.5553089309495944, "grad_norm": 0.4931704637061138, "learning_rate": 7.660164199749624e-06, "loss": 0.3678, "step": 4818 }, { "epoch": 1.5556317849792163, "grad_norm": 0.4758052441152273, "learning_rate": 7.657495192194548e-06, "loss": 0.349, "step": 4819 }, { "epoch": 1.5559546390088381, "grad_norm": 0.4598234748888078, "learning_rate": 7.654826164685048e-06, "loss": 0.2806, "step": 4820 }, { "epoch": 1.55627749303846, "grad_norm": 0.3983716118184659, "learning_rate": 7.652157117559286e-06, "loss": 0.2992, "step": 4821 }, { "epoch": 1.5566003470680818, "grad_norm": 0.5381903275013481, "learning_rate": 7.649488051155427e-06, "loss": 0.3147, "step": 4822 }, { "epoch": 1.5569232010977037, "grad_norm": 0.44430820215176975, "learning_rate": 7.646818965811639e-06, "loss": 0.3347, "step": 4823 }, { "epoch": 1.5572460551273255, "grad_norm": 0.36137967740733, "learning_rate": 7.64414986186609e-06, "loss": 0.2739, "step": 4824 }, { "epoch": 1.5575689091569473, "grad_norm": 0.3935145371600316, "learning_rate": 7.64148073965695e-06, "loss": 0.2861, "step": 4825 }, { "epoch": 1.5578917631865692, "grad_norm": 0.6411825361307549, "learning_rate": 7.638811599522397e-06, "loss": 0.3656, "step": 4826 }, { "epoch": 1.558214617216191, "grad_norm": 0.4217222972792999, "learning_rate": 7.636142441800603e-06, "loss": 0.2926, "step": 4827 }, { "epoch": 1.5585374712458129, "grad_norm": 0.397572461412392, "learning_rate": 7.633473266829753e-06, "loss": 0.2405, "step": 4828 }, { "epoch": 1.5588603252754347, "grad_norm": 0.5037431186301946, "learning_rate": 7.630804074948022e-06, "loss": 0.2732, "step": 4829 }, { "epoch": 1.5591831793050566, "grad_norm": 0.38484827833273455, "learning_rate": 7.628134866493594e-06, "loss": 0.3672, "step": 4830 }, { "epoch": 1.5595060333346784, "grad_norm": 0.3725493829195228, "learning_rate": 7.625465641804656e-06, "loss": 0.3, "step": 4831 }, { "epoch": 1.5598288873643003, "grad_norm": 0.5252853317608372, "learning_rate": 7.622796401219392e-06, "loss": 0.2736, "step": 4832 }, { "epoch": 1.5601517413939223, "grad_norm": 0.5187124213473859, "learning_rate": 7.6201271450759964e-06, "loss": 0.2671, "step": 4833 }, { "epoch": 1.5604745954235442, "grad_norm": 0.49456764635570466, "learning_rate": 7.617457873712656e-06, "loss": 0.2699, "step": 4834 }, { "epoch": 1.560797449453166, "grad_norm": 0.45361635502458164, "learning_rate": 7.614788587467562e-06, "loss": 0.304, "step": 4835 }, { "epoch": 1.5611203034827879, "grad_norm": 0.45425702772755056, "learning_rate": 7.612119286678914e-06, "loss": 0.3386, "step": 4836 }, { "epoch": 1.5614431575124097, "grad_norm": 0.38178628951291305, "learning_rate": 7.609449971684906e-06, "loss": 0.3277, "step": 4837 }, { "epoch": 1.5617660115420315, "grad_norm": 0.33710617546096877, "learning_rate": 7.606780642823738e-06, "loss": 0.3194, "step": 4838 }, { "epoch": 1.5620888655716534, "grad_norm": 0.34928211322688574, "learning_rate": 7.604111300433612e-06, "loss": 0.2639, "step": 4839 }, { "epoch": 1.5624117196012752, "grad_norm": 0.4328323279911219, "learning_rate": 7.6014419448527245e-06, "loss": 0.2808, "step": 4840 }, { "epoch": 1.5627345736308973, "grad_norm": 0.3681366772452361, "learning_rate": 7.5987725764192835e-06, "loss": 0.2664, "step": 4841 }, { "epoch": 1.5630574276605191, "grad_norm": 0.45934161177382204, "learning_rate": 7.596103195471492e-06, "loss": 0.3218, "step": 4842 }, { "epoch": 1.563380281690141, "grad_norm": 0.5574400065598167, "learning_rate": 7.593433802347558e-06, "loss": 0.3173, "step": 4843 }, { "epoch": 1.5637031357197628, "grad_norm": 0.4000723516040887, "learning_rate": 7.59076439738569e-06, "loss": 0.2916, "step": 4844 }, { "epoch": 1.5640259897493847, "grad_norm": 0.3587662967556467, "learning_rate": 7.5880949809241e-06, "loss": 0.2719, "step": 4845 }, { "epoch": 1.5643488437790065, "grad_norm": 0.4446044293912364, "learning_rate": 7.585425553300995e-06, "loss": 0.2977, "step": 4846 }, { "epoch": 1.5646716978086284, "grad_norm": 0.37608343152124907, "learning_rate": 7.5827561148545905e-06, "loss": 0.3174, "step": 4847 }, { "epoch": 1.5649945518382502, "grad_norm": 0.3182073593829123, "learning_rate": 7.5800866659231015e-06, "loss": 0.2557, "step": 4848 }, { "epoch": 1.565317405867872, "grad_norm": 0.37306324369758154, "learning_rate": 7.577417206844741e-06, "loss": 0.3215, "step": 4849 }, { "epoch": 1.565640259897494, "grad_norm": 0.5197733147547597, "learning_rate": 7.574747737957725e-06, "loss": 0.2705, "step": 4850 }, { "epoch": 1.5659631139271157, "grad_norm": 0.4230079793498488, "learning_rate": 7.572078259600276e-06, "loss": 0.3137, "step": 4851 }, { "epoch": 1.5662859679567376, "grad_norm": 0.4045926947588496, "learning_rate": 7.56940877211061e-06, "loss": 0.2721, "step": 4852 }, { "epoch": 1.5666088219863594, "grad_norm": 0.4797917448827481, "learning_rate": 7.566739275826948e-06, "loss": 0.3427, "step": 4853 }, { "epoch": 1.5669316760159813, "grad_norm": 0.4115376972482118, "learning_rate": 7.564069771087512e-06, "loss": 0.2935, "step": 4854 }, { "epoch": 1.5672545300456031, "grad_norm": 0.44217028613776266, "learning_rate": 7.561400258230525e-06, "loss": 0.2953, "step": 4855 }, { "epoch": 1.567577384075225, "grad_norm": 0.4338491135245206, "learning_rate": 7.55873073759421e-06, "loss": 0.3049, "step": 4856 }, { "epoch": 1.5679002381048468, "grad_norm": 0.389185706989841, "learning_rate": 7.556061209516791e-06, "loss": 0.26, "step": 4857 }, { "epoch": 1.5682230921344686, "grad_norm": 0.5184740391074663, "learning_rate": 7.553391674336494e-06, "loss": 0.3222, "step": 4858 }, { "epoch": 1.5685459461640905, "grad_norm": 0.35665029973223583, "learning_rate": 7.550722132391544e-06, "loss": 0.257, "step": 4859 }, { "epoch": 1.5688688001937123, "grad_norm": 0.45604128809031785, "learning_rate": 7.54805258402017e-06, "loss": 0.3267, "step": 4860 }, { "epoch": 1.5691916542233342, "grad_norm": 0.6071455788983454, "learning_rate": 7.545383029560601e-06, "loss": 0.3182, "step": 4861 }, { "epoch": 1.569514508252956, "grad_norm": 0.3977351773626613, "learning_rate": 7.542713469351065e-06, "loss": 0.27, "step": 4862 }, { "epoch": 1.5698373622825779, "grad_norm": 0.3380858346231054, "learning_rate": 7.54004390372979e-06, "loss": 0.317, "step": 4863 }, { "epoch": 1.5701602163121997, "grad_norm": 0.40643646335111094, "learning_rate": 7.537374333035009e-06, "loss": 0.2911, "step": 4864 }, { "epoch": 1.5704830703418216, "grad_norm": 0.3815076643663719, "learning_rate": 7.5347047576049495e-06, "loss": 0.2853, "step": 4865 }, { "epoch": 1.5708059243714434, "grad_norm": 0.38869456731300744, "learning_rate": 7.532035177777846e-06, "loss": 0.2957, "step": 4866 }, { "epoch": 1.5711287784010655, "grad_norm": 0.4101757784283985, "learning_rate": 7.5293655938919285e-06, "loss": 0.2838, "step": 4867 }, { "epoch": 1.5714516324306873, "grad_norm": 0.38855545240932504, "learning_rate": 7.526696006285433e-06, "loss": 0.3103, "step": 4868 }, { "epoch": 1.5717744864603092, "grad_norm": 0.3996825947037892, "learning_rate": 7.524026415296586e-06, "loss": 0.3163, "step": 4869 }, { "epoch": 1.572097340489931, "grad_norm": 0.35348308671131595, "learning_rate": 7.521356821263629e-06, "loss": 0.284, "step": 4870 }, { "epoch": 1.5724201945195528, "grad_norm": 0.32800338358788805, "learning_rate": 7.518687224524792e-06, "loss": 0.3085, "step": 4871 }, { "epoch": 1.5727430485491747, "grad_norm": 0.407273320766358, "learning_rate": 7.5160176254183085e-06, "loss": 0.2804, "step": 4872 }, { "epoch": 1.5730659025787965, "grad_norm": 0.3661510296128456, "learning_rate": 7.513348024282415e-06, "loss": 0.3462, "step": 4873 }, { "epoch": 1.5733887566084184, "grad_norm": 0.44779006425028334, "learning_rate": 7.510678421455344e-06, "loss": 0.3807, "step": 4874 }, { "epoch": 1.5737116106380404, "grad_norm": 0.3639743513981956, "learning_rate": 7.508008817275331e-06, "loss": 0.3216, "step": 4875 }, { "epoch": 1.5740344646676623, "grad_norm": 0.40192462673734286, "learning_rate": 7.5053392120806124e-06, "loss": 0.2807, "step": 4876 }, { "epoch": 1.5743573186972841, "grad_norm": 0.49867468901138967, "learning_rate": 7.502669606209425e-06, "loss": 0.2513, "step": 4877 }, { "epoch": 1.574680172726906, "grad_norm": 0.37735687928021205, "learning_rate": 7.5e-06, "loss": 0.2946, "step": 4878 }, { "epoch": 1.5750030267565278, "grad_norm": 0.4571696269586679, "learning_rate": 7.4973303937905755e-06, "loss": 0.2981, "step": 4879 }, { "epoch": 1.5753258807861497, "grad_norm": 0.40369627837819516, "learning_rate": 7.494660787919388e-06, "loss": 0.2863, "step": 4880 }, { "epoch": 1.5756487348157715, "grad_norm": 0.38483074261860417, "learning_rate": 7.491991182724669e-06, "loss": 0.3233, "step": 4881 }, { "epoch": 1.5759715888453933, "grad_norm": 0.4451069101051973, "learning_rate": 7.489321578544657e-06, "loss": 0.302, "step": 4882 }, { "epoch": 1.5762944428750152, "grad_norm": 0.36827522882549196, "learning_rate": 7.486651975717587e-06, "loss": 0.3244, "step": 4883 }, { "epoch": 1.576617296904637, "grad_norm": 0.35892775027130397, "learning_rate": 7.483982374581692e-06, "loss": 0.2965, "step": 4884 }, { "epoch": 1.5769401509342589, "grad_norm": 0.33410305759416403, "learning_rate": 7.481312775475208e-06, "loss": 0.2667, "step": 4885 }, { "epoch": 1.5772630049638807, "grad_norm": 0.44633550121639276, "learning_rate": 7.478643178736372e-06, "loss": 0.3288, "step": 4886 }, { "epoch": 1.5775858589935026, "grad_norm": 0.3430343565825981, "learning_rate": 7.4759735847034126e-06, "loss": 0.2957, "step": 4887 }, { "epoch": 1.5779087130231244, "grad_norm": 0.3620384855248349, "learning_rate": 7.473303993714571e-06, "loss": 0.311, "step": 4888 }, { "epoch": 1.5782315670527463, "grad_norm": 0.4106148470926667, "learning_rate": 7.470634406108073e-06, "loss": 0.3311, "step": 4889 }, { "epoch": 1.578554421082368, "grad_norm": 0.358686184904754, "learning_rate": 7.467964822222156e-06, "loss": 0.2821, "step": 4890 }, { "epoch": 1.57887727511199, "grad_norm": 0.4243607721400594, "learning_rate": 7.465295242395051e-06, "loss": 0.2884, "step": 4891 }, { "epoch": 1.5792001291416118, "grad_norm": 0.3347416345695264, "learning_rate": 7.462625666964995e-06, "loss": 0.2924, "step": 4892 }, { "epoch": 1.5795229831712336, "grad_norm": 0.3453476292623292, "learning_rate": 7.459956096270211e-06, "loss": 0.3465, "step": 4893 }, { "epoch": 1.5798458372008555, "grad_norm": 0.36726656474963043, "learning_rate": 7.457286530648938e-06, "loss": 0.3131, "step": 4894 }, { "epoch": 1.5801686912304773, "grad_norm": 0.38804655004021693, "learning_rate": 7.4546169704394e-06, "loss": 0.3266, "step": 4895 }, { "epoch": 1.5804915452600992, "grad_norm": 0.3473922987648008, "learning_rate": 7.45194741597983e-06, "loss": 0.2763, "step": 4896 }, { "epoch": 1.580814399289721, "grad_norm": 0.36093659346620244, "learning_rate": 7.449277867608457e-06, "loss": 0.3381, "step": 4897 }, { "epoch": 1.5811372533193429, "grad_norm": 0.6949843799671765, "learning_rate": 7.446608325663508e-06, "loss": 0.2787, "step": 4898 }, { "epoch": 1.5814601073489647, "grad_norm": 0.3484811199218812, "learning_rate": 7.443938790483211e-06, "loss": 0.3021, "step": 4899 }, { "epoch": 1.5817829613785868, "grad_norm": 0.37805627046667833, "learning_rate": 7.44126926240579e-06, "loss": 0.2884, "step": 4900 }, { "epoch": 1.5821058154082086, "grad_norm": 0.46484314325743364, "learning_rate": 7.438599741769475e-06, "loss": 0.3284, "step": 4901 }, { "epoch": 1.5824286694378304, "grad_norm": 0.37618552536328004, "learning_rate": 7.435930228912488e-06, "loss": 0.2998, "step": 4902 }, { "epoch": 1.5827515234674523, "grad_norm": 0.36612265929402205, "learning_rate": 7.433260724173053e-06, "loss": 0.2399, "step": 4903 }, { "epoch": 1.5830743774970741, "grad_norm": 0.38122041079279506, "learning_rate": 7.43059122788939e-06, "loss": 0.3473, "step": 4904 }, { "epoch": 1.583397231526696, "grad_norm": 0.424853686702827, "learning_rate": 7.427921740399724e-06, "loss": 0.2841, "step": 4905 }, { "epoch": 1.5837200855563178, "grad_norm": 0.3885128350673456, "learning_rate": 7.425252262042275e-06, "loss": 0.3173, "step": 4906 }, { "epoch": 1.5840429395859397, "grad_norm": 0.3683882056911287, "learning_rate": 7.422582793155262e-06, "loss": 0.3145, "step": 4907 }, { "epoch": 1.5843657936155615, "grad_norm": 0.42582297721421275, "learning_rate": 7.4199133340769e-06, "loss": 0.3654, "step": 4908 }, { "epoch": 1.5846886476451836, "grad_norm": 0.48769231443362826, "learning_rate": 7.417243885145412e-06, "loss": 0.3175, "step": 4909 }, { "epoch": 1.5850115016748054, "grad_norm": 0.3931674253100992, "learning_rate": 7.414574446699006e-06, "loss": 0.2841, "step": 4910 }, { "epoch": 1.5853343557044273, "grad_norm": 0.4191229913131564, "learning_rate": 7.411905019075903e-06, "loss": 0.293, "step": 4911 }, { "epoch": 1.5856572097340491, "grad_norm": 0.4361143242816659, "learning_rate": 7.409235602614311e-06, "loss": 0.3197, "step": 4912 }, { "epoch": 1.585980063763671, "grad_norm": 0.4246060297792956, "learning_rate": 7.406566197652444e-06, "loss": 0.2778, "step": 4913 }, { "epoch": 1.5863029177932928, "grad_norm": 0.39201390840375133, "learning_rate": 7.403896804528508e-06, "loss": 0.2967, "step": 4914 }, { "epoch": 1.5866257718229146, "grad_norm": 0.38125478150108816, "learning_rate": 7.401227423580716e-06, "loss": 0.3256, "step": 4915 }, { "epoch": 1.5869486258525365, "grad_norm": 0.34811902317628085, "learning_rate": 7.398558055147277e-06, "loss": 0.3033, "step": 4916 }, { "epoch": 1.5872714798821583, "grad_norm": 0.45644195984148495, "learning_rate": 7.3958886995663885e-06, "loss": 0.3011, "step": 4917 }, { "epoch": 1.5875943339117802, "grad_norm": 0.3707799309070755, "learning_rate": 7.393219357176263e-06, "loss": 0.3567, "step": 4918 }, { "epoch": 1.587917187941402, "grad_norm": 0.3393524705186151, "learning_rate": 7.3905500283150925e-06, "loss": 0.2841, "step": 4919 }, { "epoch": 1.5882400419710239, "grad_norm": 0.4303145442215851, "learning_rate": 7.387880713321088e-06, "loss": 0.2981, "step": 4920 }, { "epoch": 1.5885628960006457, "grad_norm": 0.3925996494407294, "learning_rate": 7.3852114125324385e-06, "loss": 0.2834, "step": 4921 }, { "epoch": 1.5888857500302676, "grad_norm": 0.43413115144701264, "learning_rate": 7.382542126287346e-06, "loss": 0.2556, "step": 4922 }, { "epoch": 1.5892086040598894, "grad_norm": 0.42170591498942767, "learning_rate": 7.379872854924004e-06, "loss": 0.2899, "step": 4923 }, { "epoch": 1.5895314580895112, "grad_norm": 0.32138712331370867, "learning_rate": 7.377203598780609e-06, "loss": 0.2761, "step": 4924 }, { "epoch": 1.589854312119133, "grad_norm": 0.42483660697167824, "learning_rate": 7.374534358195344e-06, "loss": 0.3336, "step": 4925 }, { "epoch": 1.590177166148755, "grad_norm": 0.4298744508807626, "learning_rate": 7.371865133506408e-06, "loss": 0.3078, "step": 4926 }, { "epoch": 1.5905000201783768, "grad_norm": 0.3434726790888897, "learning_rate": 7.369195925051979e-06, "loss": 0.2734, "step": 4927 }, { "epoch": 1.5908228742079986, "grad_norm": 0.43743524188619043, "learning_rate": 7.366526733170249e-06, "loss": 0.3756, "step": 4928 }, { "epoch": 1.5911457282376205, "grad_norm": 0.36512563709537177, "learning_rate": 7.363857558199397e-06, "loss": 0.2562, "step": 4929 }, { "epoch": 1.5914685822672423, "grad_norm": 0.5072584512318304, "learning_rate": 7.361188400477606e-06, "loss": 0.257, "step": 4930 }, { "epoch": 1.5917914362968641, "grad_norm": 0.43535098883251466, "learning_rate": 7.358519260343051e-06, "loss": 0.3629, "step": 4931 }, { "epoch": 1.592114290326486, "grad_norm": 0.4417358054955215, "learning_rate": 7.355850138133914e-06, "loss": 0.2884, "step": 4932 }, { "epoch": 1.5924371443561078, "grad_norm": 0.6700112909752141, "learning_rate": 7.353181034188362e-06, "loss": 0.2993, "step": 4933 }, { "epoch": 1.59275999838573, "grad_norm": 0.4261188469731869, "learning_rate": 7.350511948844572e-06, "loss": 0.2713, "step": 4934 }, { "epoch": 1.5930828524153517, "grad_norm": 0.36609949917028284, "learning_rate": 7.347842882440716e-06, "loss": 0.2656, "step": 4935 }, { "epoch": 1.5934057064449736, "grad_norm": 0.3594844475717161, "learning_rate": 7.345173835314953e-06, "loss": 0.2827, "step": 4936 }, { "epoch": 1.5937285604745954, "grad_norm": 0.4581003009824041, "learning_rate": 7.3425048078054536e-06, "loss": 0.2938, "step": 4937 }, { "epoch": 1.5940514145042173, "grad_norm": 0.4580849769630758, "learning_rate": 7.339835800250377e-06, "loss": 0.3128, "step": 4938 }, { "epoch": 1.5943742685338391, "grad_norm": 0.4527703224160194, "learning_rate": 7.337166812987887e-06, "loss": 0.3076, "step": 4939 }, { "epoch": 1.594697122563461, "grad_norm": 0.40805916182838403, "learning_rate": 7.334497846356133e-06, "loss": 0.2915, "step": 4940 }, { "epoch": 1.5950199765930828, "grad_norm": 0.4811146685123204, "learning_rate": 7.33182890069328e-06, "loss": 0.3077, "step": 4941 }, { "epoch": 1.5953428306227047, "grad_norm": 0.4009853332759902, "learning_rate": 7.329159976337468e-06, "loss": 0.2895, "step": 4942 }, { "epoch": 1.5956656846523267, "grad_norm": 0.5571877069477863, "learning_rate": 7.326491073626855e-06, "loss": 0.2553, "step": 4943 }, { "epoch": 1.5959885386819486, "grad_norm": 0.47441604758514927, "learning_rate": 7.323822192899582e-06, "loss": 0.2569, "step": 4944 }, { "epoch": 1.5963113927115704, "grad_norm": 0.3916547818626131, "learning_rate": 7.321153334493796e-06, "loss": 0.3052, "step": 4945 }, { "epoch": 1.5966342467411923, "grad_norm": 0.36619941341443796, "learning_rate": 7.318484498747634e-06, "loss": 0.2947, "step": 4946 }, { "epoch": 1.596957100770814, "grad_norm": 0.3803210165606184, "learning_rate": 7.31581568599924e-06, "loss": 0.3197, "step": 4947 }, { "epoch": 1.597279954800436, "grad_norm": 0.7929801086753465, "learning_rate": 7.313146896586742e-06, "loss": 0.2475, "step": 4948 }, { "epoch": 1.5976028088300578, "grad_norm": 0.3914467317027698, "learning_rate": 7.310478130848276e-06, "loss": 0.3089, "step": 4949 }, { "epoch": 1.5979256628596796, "grad_norm": 0.4017575858681726, "learning_rate": 7.307809389121969e-06, "loss": 0.258, "step": 4950 }, { "epoch": 1.5982485168893015, "grad_norm": 0.4819746605664891, "learning_rate": 7.3051406717459484e-06, "loss": 0.3147, "step": 4951 }, { "epoch": 1.5985713709189233, "grad_norm": 0.4359481312952665, "learning_rate": 7.302471979058337e-06, "loss": 0.33, "step": 4952 }, { "epoch": 1.5988942249485452, "grad_norm": 0.4391565759383544, "learning_rate": 7.299803311397251e-06, "loss": 0.3039, "step": 4953 }, { "epoch": 1.599217078978167, "grad_norm": 0.374309451674163, "learning_rate": 7.297134669100812e-06, "loss": 0.2885, "step": 4954 }, { "epoch": 1.5995399330077889, "grad_norm": 0.3618182386175755, "learning_rate": 7.294466052507129e-06, "loss": 0.3449, "step": 4955 }, { "epoch": 1.5998627870374107, "grad_norm": 0.4150551746109557, "learning_rate": 7.291797461954318e-06, "loss": 0.2831, "step": 4956 }, { "epoch": 1.6001856410670325, "grad_norm": 0.48413940214241746, "learning_rate": 7.2891288977804785e-06, "loss": 0.331, "step": 4957 }, { "epoch": 1.6005084950966544, "grad_norm": 0.4084713731920299, "learning_rate": 7.2864603603237206e-06, "loss": 0.3277, "step": 4958 }, { "epoch": 1.6008313491262762, "grad_norm": 0.38345848177493774, "learning_rate": 7.283791849922138e-06, "loss": 0.2913, "step": 4959 }, { "epoch": 1.601154203155898, "grad_norm": 0.40743606427480794, "learning_rate": 7.2811233669138316e-06, "loss": 0.336, "step": 4960 }, { "epoch": 1.60147705718552, "grad_norm": 0.4527268464459356, "learning_rate": 7.278454911636892e-06, "loss": 0.3091, "step": 4961 }, { "epoch": 1.6017999112151418, "grad_norm": 0.3664705678749686, "learning_rate": 7.275786484429411e-06, "loss": 0.3399, "step": 4962 }, { "epoch": 1.6021227652447636, "grad_norm": 0.3984514453149619, "learning_rate": 7.273118085629473e-06, "loss": 0.32, "step": 4963 }, { "epoch": 1.6024456192743854, "grad_norm": 0.4219903777819815, "learning_rate": 7.270449715575163e-06, "loss": 0.2553, "step": 4964 }, { "epoch": 1.6027684733040073, "grad_norm": 0.412207036955148, "learning_rate": 7.267781374604555e-06, "loss": 0.3227, "step": 4965 }, { "epoch": 1.6030913273336291, "grad_norm": 0.3812787667167343, "learning_rate": 7.26511306305573e-06, "loss": 0.3487, "step": 4966 }, { "epoch": 1.603414181363251, "grad_norm": 0.3933379430648404, "learning_rate": 7.262444781266755e-06, "loss": 0.2842, "step": 4967 }, { "epoch": 1.603737035392873, "grad_norm": 0.418206157878715, "learning_rate": 7.259776529575699e-06, "loss": 0.3062, "step": 4968 }, { "epoch": 1.6040598894224949, "grad_norm": 0.4113523163741898, "learning_rate": 7.2571083083206244e-06, "loss": 0.2706, "step": 4969 }, { "epoch": 1.6043827434521167, "grad_norm": 0.31957722443949443, "learning_rate": 7.254440117839596e-06, "loss": 0.2898, "step": 4970 }, { "epoch": 1.6047055974817386, "grad_norm": 0.42090768541673995, "learning_rate": 7.251771958470664e-06, "loss": 0.361, "step": 4971 }, { "epoch": 1.6050284515113604, "grad_norm": 0.38904629255664847, "learning_rate": 7.249103830551882e-06, "loss": 0.3237, "step": 4972 }, { "epoch": 1.6053513055409823, "grad_norm": 0.38138206520026957, "learning_rate": 7.246435734421303e-06, "loss": 0.283, "step": 4973 }, { "epoch": 1.605674159570604, "grad_norm": 0.3611468909771359, "learning_rate": 7.243767670416964e-06, "loss": 0.3289, "step": 4974 }, { "epoch": 1.605997013600226, "grad_norm": 0.47964443509982674, "learning_rate": 7.241099638876911e-06, "loss": 0.3399, "step": 4975 }, { "epoch": 1.606319867629848, "grad_norm": 0.41816858198449264, "learning_rate": 7.238431640139175e-06, "loss": 0.2869, "step": 4976 }, { "epoch": 1.6066427216594699, "grad_norm": 0.40174070653722105, "learning_rate": 7.235763674541791e-06, "loss": 0.287, "step": 4977 }, { "epoch": 1.6069655756890917, "grad_norm": 0.4331865063842694, "learning_rate": 7.233095742422784e-06, "loss": 0.2968, "step": 4978 }, { "epoch": 1.6072884297187136, "grad_norm": 0.4176243741044237, "learning_rate": 7.230427844120181e-06, "loss": 0.3446, "step": 4979 }, { "epoch": 1.6076112837483354, "grad_norm": 0.32201032558135473, "learning_rate": 7.227759979971998e-06, "loss": 0.3127, "step": 4980 }, { "epoch": 1.6079341377779572, "grad_norm": 0.3895724149912796, "learning_rate": 7.225092150316252e-06, "loss": 0.3326, "step": 4981 }, { "epoch": 1.608256991807579, "grad_norm": 0.428422694113936, "learning_rate": 7.222424355490949e-06, "loss": 0.2639, "step": 4982 }, { "epoch": 1.608579845837201, "grad_norm": 0.39305763826533896, "learning_rate": 7.2197565958341e-06, "loss": 0.3386, "step": 4983 }, { "epoch": 1.6089026998668228, "grad_norm": 0.3841873048126827, "learning_rate": 7.2170888716837045e-06, "loss": 0.3209, "step": 4984 }, { "epoch": 1.6092255538964446, "grad_norm": 0.38580880300534187, "learning_rate": 7.21442118337776e-06, "loss": 0.2987, "step": 4985 }, { "epoch": 1.6095484079260665, "grad_norm": 0.44147971047814705, "learning_rate": 7.211753531254255e-06, "loss": 0.4167, "step": 4986 }, { "epoch": 1.6098712619556883, "grad_norm": 0.4876215662629399, "learning_rate": 7.2090859156511845e-06, "loss": 0.3168, "step": 4987 }, { "epoch": 1.6101941159853101, "grad_norm": 0.46065565879110715, "learning_rate": 7.206418336906525e-06, "loss": 0.2854, "step": 4988 }, { "epoch": 1.610516970014932, "grad_norm": 0.8364030642162847, "learning_rate": 7.2037507953582575e-06, "loss": 0.2563, "step": 4989 }, { "epoch": 1.6108398240445538, "grad_norm": 0.5488451415402156, "learning_rate": 7.20108329134436e-06, "loss": 0.3074, "step": 4990 }, { "epoch": 1.6111626780741757, "grad_norm": 0.44284373926016124, "learning_rate": 7.198415825202795e-06, "loss": 0.3379, "step": 4991 }, { "epoch": 1.6114855321037975, "grad_norm": 0.4110980458780911, "learning_rate": 7.1957483972715305e-06, "loss": 0.3462, "step": 4992 }, { "epoch": 1.6118083861334194, "grad_norm": 0.34621878660159006, "learning_rate": 7.193081007888524e-06, "loss": 0.3186, "step": 4993 }, { "epoch": 1.6121312401630412, "grad_norm": 0.427852599251729, "learning_rate": 7.190413657391734e-06, "loss": 0.3016, "step": 4994 }, { "epoch": 1.612454094192663, "grad_norm": 0.4441230015107353, "learning_rate": 7.187746346119103e-06, "loss": 0.298, "step": 4995 }, { "epoch": 1.612776948222285, "grad_norm": 0.4137405251286969, "learning_rate": 7.185079074408585e-06, "loss": 0.2475, "step": 4996 }, { "epoch": 1.6130998022519067, "grad_norm": 0.40924220894574403, "learning_rate": 7.182411842598111e-06, "loss": 0.2995, "step": 4997 }, { "epoch": 1.6134226562815286, "grad_norm": 0.40175781342392836, "learning_rate": 7.179744651025622e-06, "loss": 0.2903, "step": 4998 }, { "epoch": 1.6137455103111504, "grad_norm": 0.46551257562002996, "learning_rate": 7.177077500029042e-06, "loss": 0.3454, "step": 4999 }, { "epoch": 1.6140683643407723, "grad_norm": 0.4610312983920445, "learning_rate": 7.174410389946299e-06, "loss": 0.3033, "step": 5000 }, { "epoch": 1.6143912183703941, "grad_norm": 0.40188309387113746, "learning_rate": 7.171743321115309e-06, "loss": 0.2737, "step": 5001 }, { "epoch": 1.6147140724000162, "grad_norm": 0.3774600462710801, "learning_rate": 7.169076293873992e-06, "loss": 0.2796, "step": 5002 }, { "epoch": 1.615036926429638, "grad_norm": 0.3653443090447563, "learning_rate": 7.1664093085602505e-06, "loss": 0.3096, "step": 5003 }, { "epoch": 1.6153597804592599, "grad_norm": 0.38045194617168937, "learning_rate": 7.163742365511993e-06, "loss": 0.3179, "step": 5004 }, { "epoch": 1.6156826344888817, "grad_norm": 0.3878945377399526, "learning_rate": 7.1610754650671115e-06, "loss": 0.3032, "step": 5005 }, { "epoch": 1.6160054885185036, "grad_norm": 0.44394686275956735, "learning_rate": 7.1584086075635026e-06, "loss": 0.3187, "step": 5006 }, { "epoch": 1.6163283425481254, "grad_norm": 0.4101711313890086, "learning_rate": 7.155741793339054e-06, "loss": 0.2878, "step": 5007 }, { "epoch": 1.6166511965777473, "grad_norm": 0.4309619354681287, "learning_rate": 7.153075022731644e-06, "loss": 0.3389, "step": 5008 }, { "epoch": 1.616974050607369, "grad_norm": 0.37547973960693365, "learning_rate": 7.150408296079152e-06, "loss": 0.2972, "step": 5009 }, { "epoch": 1.6172969046369912, "grad_norm": 0.40234806379198607, "learning_rate": 7.147741613719446e-06, "loss": 0.2986, "step": 5010 }, { "epoch": 1.617619758666613, "grad_norm": 0.43992673887732403, "learning_rate": 7.145074975990397e-06, "loss": 0.2759, "step": 5011 }, { "epoch": 1.6179426126962348, "grad_norm": 0.4929919360600916, "learning_rate": 7.142408383229854e-06, "loss": 0.3083, "step": 5012 }, { "epoch": 1.6182654667258567, "grad_norm": 0.46026952098487767, "learning_rate": 7.139741835775681e-06, "loss": 0.2639, "step": 5013 }, { "epoch": 1.6185883207554785, "grad_norm": 0.39887339247296005, "learning_rate": 7.137075333965719e-06, "loss": 0.3402, "step": 5014 }, { "epoch": 1.6189111747851004, "grad_norm": 0.3833827451817886, "learning_rate": 7.134408878137813e-06, "loss": 0.3335, "step": 5015 }, { "epoch": 1.6192340288147222, "grad_norm": 0.37833576311261147, "learning_rate": 7.131742468629797e-06, "loss": 0.2912, "step": 5016 }, { "epoch": 1.619556882844344, "grad_norm": 0.39815430117219874, "learning_rate": 7.129076105779507e-06, "loss": 0.2558, "step": 5017 }, { "epoch": 1.619879736873966, "grad_norm": 0.3321081929757849, "learning_rate": 7.12640978992476e-06, "loss": 0.2599, "step": 5018 }, { "epoch": 1.6202025909035878, "grad_norm": 0.4227889015107674, "learning_rate": 7.1237435214033815e-06, "loss": 0.2832, "step": 5019 }, { "epoch": 1.6205254449332096, "grad_norm": 0.3366328008906609, "learning_rate": 7.1210773005531774e-06, "loss": 0.2782, "step": 5020 }, { "epoch": 1.6208482989628314, "grad_norm": 0.4145331024328971, "learning_rate": 7.1184111277119605e-06, "loss": 0.3317, "step": 5021 }, { "epoch": 1.6211711529924533, "grad_norm": 0.33708411129456944, "learning_rate": 7.115745003217526e-06, "loss": 0.2557, "step": 5022 }, { "epoch": 1.6214940070220751, "grad_norm": 0.5250835150140829, "learning_rate": 7.113078927407671e-06, "loss": 0.3254, "step": 5023 }, { "epoch": 1.621816861051697, "grad_norm": 0.36659660818099493, "learning_rate": 7.110412900620181e-06, "loss": 0.289, "step": 5024 }, { "epoch": 1.6221397150813188, "grad_norm": 0.42645233827141815, "learning_rate": 7.107746923192845e-06, "loss": 0.2918, "step": 5025 }, { "epoch": 1.6224625691109407, "grad_norm": 0.4722317445529332, "learning_rate": 7.105080995463429e-06, "loss": 0.3323, "step": 5026 }, { "epoch": 1.6227854231405625, "grad_norm": 0.44736613849583684, "learning_rate": 7.102415117769707e-06, "loss": 0.251, "step": 5027 }, { "epoch": 1.6231082771701844, "grad_norm": 0.428349710054276, "learning_rate": 7.099749290449446e-06, "loss": 0.2801, "step": 5028 }, { "epoch": 1.6234311311998062, "grad_norm": 0.4049868720937726, "learning_rate": 7.097083513840396e-06, "loss": 0.2696, "step": 5029 }, { "epoch": 1.623753985229428, "grad_norm": 0.47491756374680627, "learning_rate": 7.094417788280311e-06, "loss": 0.2454, "step": 5030 }, { "epoch": 1.6240768392590499, "grad_norm": 0.4094233776045748, "learning_rate": 7.091752114106932e-06, "loss": 0.2748, "step": 5031 }, { "epoch": 1.6243996932886717, "grad_norm": 0.3886150489354032, "learning_rate": 7.0890864916580004e-06, "loss": 0.3206, "step": 5032 }, { "epoch": 1.6247225473182936, "grad_norm": 0.37180318388979455, "learning_rate": 7.0864209212712416e-06, "loss": 0.2729, "step": 5033 }, { "epoch": 1.6250454013479154, "grad_norm": 0.3828943565962083, "learning_rate": 7.083755403284386e-06, "loss": 0.3161, "step": 5034 }, { "epoch": 1.6253682553775375, "grad_norm": 0.41565957010152593, "learning_rate": 7.081089938035144e-06, "loss": 0.2894, "step": 5035 }, { "epoch": 1.6256911094071593, "grad_norm": 0.3749509381283761, "learning_rate": 7.078424525861234e-06, "loss": 0.3036, "step": 5036 }, { "epoch": 1.6260139634367812, "grad_norm": 0.4067665389989686, "learning_rate": 7.075759167100352e-06, "loss": 0.307, "step": 5037 }, { "epoch": 1.626336817466403, "grad_norm": 0.3989532536046958, "learning_rate": 7.0730938620902014e-06, "loss": 0.2904, "step": 5038 }, { "epoch": 1.6266596714960249, "grad_norm": 0.4159778763686852, "learning_rate": 7.07042861116847e-06, "loss": 0.339, "step": 5039 }, { "epoch": 1.6269825255256467, "grad_norm": 0.3715224558693852, "learning_rate": 7.067763414672844e-06, "loss": 0.3013, "step": 5040 }, { "epoch": 1.6273053795552685, "grad_norm": 0.4853450265764134, "learning_rate": 7.065098272940994e-06, "loss": 0.285, "step": 5041 }, { "epoch": 1.6276282335848904, "grad_norm": 0.4042071727263603, "learning_rate": 7.062433186310599e-06, "loss": 0.2499, "step": 5042 }, { "epoch": 1.6279510876145122, "grad_norm": 0.33946734520051164, "learning_rate": 7.0597681551193126e-06, "loss": 0.2824, "step": 5043 }, { "epoch": 1.6282739416441343, "grad_norm": 0.38428175906703993, "learning_rate": 7.057103179704799e-06, "loss": 0.2722, "step": 5044 }, { "epoch": 1.6285967956737561, "grad_norm": 0.4227339176496193, "learning_rate": 7.054438260404697e-06, "loss": 0.3264, "step": 5045 }, { "epoch": 1.628919649703378, "grad_norm": 0.3944988156985482, "learning_rate": 7.051773397556656e-06, "loss": 0.3193, "step": 5046 }, { "epoch": 1.6292425037329998, "grad_norm": 0.4049806906281197, "learning_rate": 7.049108591498309e-06, "loss": 0.3433, "step": 5047 }, { "epoch": 1.6295653577626217, "grad_norm": 0.3909111133156342, "learning_rate": 7.046443842567281e-06, "loss": 0.3391, "step": 5048 }, { "epoch": 1.6298882117922435, "grad_norm": 0.39433189733909313, "learning_rate": 7.043779151101196e-06, "loss": 0.3318, "step": 5049 }, { "epoch": 1.6302110658218654, "grad_norm": 0.3747433615193816, "learning_rate": 7.041114517437662e-06, "loss": 0.2909, "step": 5050 }, { "epoch": 1.6305339198514872, "grad_norm": 0.36410088852166567, "learning_rate": 7.038449941914291e-06, "loss": 0.2444, "step": 5051 }, { "epoch": 1.630856773881109, "grad_norm": 0.3941190163969094, "learning_rate": 7.035785424868672e-06, "loss": 0.3122, "step": 5052 }, { "epoch": 1.631179627910731, "grad_norm": 0.4565308554113994, "learning_rate": 7.033120966638404e-06, "loss": 0.326, "step": 5053 }, { "epoch": 1.6315024819403527, "grad_norm": 0.3636692610163138, "learning_rate": 7.030456567561064e-06, "loss": 0.3001, "step": 5054 }, { "epoch": 1.6318253359699746, "grad_norm": 0.3832567045287476, "learning_rate": 7.027792227974232e-06, "loss": 0.3536, "step": 5055 }, { "epoch": 1.6321481899995964, "grad_norm": 0.44685150568966786, "learning_rate": 7.025127948215472e-06, "loss": 0.2978, "step": 5056 }, { "epoch": 1.6324710440292183, "grad_norm": 0.41513321928851366, "learning_rate": 7.02246372862235e-06, "loss": 0.319, "step": 5057 }, { "epoch": 1.6327938980588401, "grad_norm": 0.406585071269466, "learning_rate": 7.019799569532413e-06, "loss": 0.3098, "step": 5058 }, { "epoch": 1.633116752088462, "grad_norm": 0.5095327616360348, "learning_rate": 7.017135471283213e-06, "loss": 0.2584, "step": 5059 }, { "epoch": 1.6334396061180838, "grad_norm": 0.34737613703863407, "learning_rate": 7.01447143421228e-06, "loss": 0.289, "step": 5060 }, { "epoch": 1.6337624601477057, "grad_norm": 0.35972213594456626, "learning_rate": 7.0118074586571494e-06, "loss": 0.3138, "step": 5061 }, { "epoch": 1.6340853141773275, "grad_norm": 0.3763279476225025, "learning_rate": 7.0091435449553404e-06, "loss": 0.2869, "step": 5062 }, { "epoch": 1.6344081682069493, "grad_norm": 0.404483645139349, "learning_rate": 7.00647969344437e-06, "loss": 0.3008, "step": 5063 }, { "epoch": 1.6347310222365712, "grad_norm": 0.37870089668151785, "learning_rate": 7.003815904461742e-06, "loss": 0.3021, "step": 5064 }, { "epoch": 1.635053876266193, "grad_norm": 0.3573282548054672, "learning_rate": 7.001152178344955e-06, "loss": 0.2735, "step": 5065 }, { "epoch": 1.6353767302958149, "grad_norm": 0.3726389424770818, "learning_rate": 6.998488515431502e-06, "loss": 0.2994, "step": 5066 }, { "epoch": 1.6356995843254367, "grad_norm": 0.414010428975128, "learning_rate": 6.995824916058861e-06, "loss": 0.3513, "step": 5067 }, { "epoch": 1.6360224383550586, "grad_norm": 0.3857092051010873, "learning_rate": 6.993161380564513e-06, "loss": 0.2842, "step": 5068 }, { "epoch": 1.6363452923846806, "grad_norm": 0.5193243875180976, "learning_rate": 6.990497909285915e-06, "loss": 0.2992, "step": 5069 }, { "epoch": 1.6366681464143025, "grad_norm": 0.42217522709661953, "learning_rate": 6.987834502560533e-06, "loss": 0.3426, "step": 5070 }, { "epoch": 1.6369910004439243, "grad_norm": 0.43911827780598317, "learning_rate": 6.985171160725813e-06, "loss": 0.2666, "step": 5071 }, { "epoch": 1.6373138544735462, "grad_norm": 0.4467205961981618, "learning_rate": 6.9825078841192e-06, "loss": 0.3195, "step": 5072 }, { "epoch": 1.637636708503168, "grad_norm": 0.3674985584176251, "learning_rate": 6.979844673078123e-06, "loss": 0.3317, "step": 5073 }, { "epoch": 1.6379595625327898, "grad_norm": 0.3876274467184043, "learning_rate": 6.977181527940013e-06, "loss": 0.2805, "step": 5074 }, { "epoch": 1.6382824165624117, "grad_norm": 0.3828893113152972, "learning_rate": 6.974518449042281e-06, "loss": 0.31, "step": 5075 }, { "epoch": 1.6386052705920335, "grad_norm": 0.38231110632833476, "learning_rate": 6.9718554367223385e-06, "loss": 0.3173, "step": 5076 }, { "epoch": 1.6389281246216554, "grad_norm": 0.45581853607608336, "learning_rate": 6.969192491317583e-06, "loss": 0.3538, "step": 5077 }, { "epoch": 1.6392509786512774, "grad_norm": 0.4193205314709306, "learning_rate": 6.96652961316541e-06, "loss": 0.2815, "step": 5078 }, { "epoch": 1.6395738326808993, "grad_norm": 0.35542692709532686, "learning_rate": 6.963866802603198e-06, "loss": 0.3627, "step": 5079 }, { "epoch": 1.6398966867105211, "grad_norm": 0.4625147094728348, "learning_rate": 6.961204059968328e-06, "loss": 0.3345, "step": 5080 }, { "epoch": 1.640219540740143, "grad_norm": 0.45509892240503436, "learning_rate": 6.958541385598159e-06, "loss": 0.3058, "step": 5081 }, { "epoch": 1.6405423947697648, "grad_norm": 0.4454065440624649, "learning_rate": 6.955878779830049e-06, "loss": 0.3083, "step": 5082 }, { "epoch": 1.6408652487993867, "grad_norm": 0.42838053846935914, "learning_rate": 6.953216243001352e-06, "loss": 0.2642, "step": 5083 }, { "epoch": 1.6411881028290085, "grad_norm": 0.36716220102729913, "learning_rate": 6.950553775449401e-06, "loss": 0.3625, "step": 5084 }, { "epoch": 1.6415109568586304, "grad_norm": 0.6227122985977538, "learning_rate": 6.947891377511532e-06, "loss": 0.3146, "step": 5085 }, { "epoch": 1.6418338108882522, "grad_norm": 0.3854555297434133, "learning_rate": 6.9452290495250645e-06, "loss": 0.2739, "step": 5086 }, { "epoch": 1.642156664917874, "grad_norm": 0.4773753202739949, "learning_rate": 6.942566791827313e-06, "loss": 0.272, "step": 5087 }, { "epoch": 1.6424795189474959, "grad_norm": 0.41419493219610143, "learning_rate": 6.93990460475558e-06, "loss": 0.3106, "step": 5088 }, { "epoch": 1.6428023729771177, "grad_norm": 0.4221679051935396, "learning_rate": 6.937242488647164e-06, "loss": 0.2923, "step": 5089 }, { "epoch": 1.6431252270067396, "grad_norm": 0.3922399600153777, "learning_rate": 6.934580443839348e-06, "loss": 0.2907, "step": 5090 }, { "epoch": 1.6434480810363614, "grad_norm": 0.42045399622950386, "learning_rate": 6.931918470669414e-06, "loss": 0.3333, "step": 5091 }, { "epoch": 1.6437709350659833, "grad_norm": 0.47610521201625516, "learning_rate": 6.929256569474624e-06, "loss": 0.325, "step": 5092 }, { "epoch": 1.644093789095605, "grad_norm": 0.3529554415084156, "learning_rate": 6.926594740592243e-06, "loss": 0.3214, "step": 5093 }, { "epoch": 1.644416643125227, "grad_norm": 0.4451247536934805, "learning_rate": 6.923932984359516e-06, "loss": 0.3297, "step": 5094 }, { "epoch": 1.6447394971548488, "grad_norm": 0.4642868491283015, "learning_rate": 6.9212713011136906e-06, "loss": 0.3125, "step": 5095 }, { "epoch": 1.6450623511844706, "grad_norm": 0.37614996734396805, "learning_rate": 6.918609691191991e-06, "loss": 0.3371, "step": 5096 }, { "epoch": 1.6453852052140925, "grad_norm": 0.3625122543781101, "learning_rate": 6.915948154931646e-06, "loss": 0.3284, "step": 5097 }, { "epoch": 1.6457080592437143, "grad_norm": 0.35578004823591747, "learning_rate": 6.913286692669862e-06, "loss": 0.2732, "step": 5098 }, { "epoch": 1.6460309132733362, "grad_norm": 0.3962182184253204, "learning_rate": 6.910625304743851e-06, "loss": 0.3337, "step": 5099 }, { "epoch": 1.646353767302958, "grad_norm": 0.36197586324932013, "learning_rate": 6.907963991490797e-06, "loss": 0.296, "step": 5100 }, { "epoch": 1.6466766213325799, "grad_norm": 0.4163328994373151, "learning_rate": 6.905302753247891e-06, "loss": 0.2959, "step": 5101 }, { "epoch": 1.6469994753622017, "grad_norm": 0.3741336066025503, "learning_rate": 6.902641590352308e-06, "loss": 0.3237, "step": 5102 }, { "epoch": 1.6473223293918238, "grad_norm": 0.436122551236562, "learning_rate": 6.899980503141212e-06, "loss": 0.2689, "step": 5103 }, { "epoch": 1.6476451834214456, "grad_norm": 0.3498874413772278, "learning_rate": 6.897319491951763e-06, "loss": 0.2857, "step": 5104 }, { "epoch": 1.6479680374510675, "grad_norm": 0.4577197774860905, "learning_rate": 6.894658557121101e-06, "loss": 0.3665, "step": 5105 }, { "epoch": 1.6482908914806893, "grad_norm": 0.4062627073992372, "learning_rate": 6.89199769898637e-06, "loss": 0.3871, "step": 5106 }, { "epoch": 1.6486137455103111, "grad_norm": 0.40187158769099995, "learning_rate": 6.88933691788469e-06, "loss": 0.2854, "step": 5107 }, { "epoch": 1.648936599539933, "grad_norm": 0.3980297723770371, "learning_rate": 6.886676214153183e-06, "loss": 0.3526, "step": 5108 }, { "epoch": 1.6492594535695548, "grad_norm": 0.35812037936449703, "learning_rate": 6.884015588128954e-06, "loss": 0.3256, "step": 5109 }, { "epoch": 1.6495823075991767, "grad_norm": 0.38241432537667125, "learning_rate": 6.881355040149103e-06, "loss": 0.3031, "step": 5110 }, { "epoch": 1.6499051616287987, "grad_norm": 0.358455433043967, "learning_rate": 6.878694570550715e-06, "loss": 0.288, "step": 5111 }, { "epoch": 1.6502280156584206, "grad_norm": 0.43557855991075145, "learning_rate": 6.876034179670872e-06, "loss": 0.3073, "step": 5112 }, { "epoch": 1.6505508696880424, "grad_norm": 0.3806620926978151, "learning_rate": 6.873373867846636e-06, "loss": 0.3056, "step": 5113 }, { "epoch": 1.6508737237176643, "grad_norm": 0.36474685818360536, "learning_rate": 6.870713635415071e-06, "loss": 0.2896, "step": 5114 }, { "epoch": 1.6511965777472861, "grad_norm": 0.3740632486839991, "learning_rate": 6.868053482713219e-06, "loss": 0.2741, "step": 5115 }, { "epoch": 1.651519431776908, "grad_norm": 0.3547228488736507, "learning_rate": 6.865393410078122e-06, "loss": 0.288, "step": 5116 }, { "epoch": 1.6518422858065298, "grad_norm": 0.40575007543431196, "learning_rate": 6.862733417846804e-06, "loss": 0.2991, "step": 5117 }, { "epoch": 1.6521651398361517, "grad_norm": 0.3298802304856365, "learning_rate": 6.860073506356288e-06, "loss": 0.2948, "step": 5118 }, { "epoch": 1.6524879938657735, "grad_norm": 0.37685279124392873, "learning_rate": 6.857413675943575e-06, "loss": 0.3011, "step": 5119 }, { "epoch": 1.6528108478953953, "grad_norm": 0.3814611744891847, "learning_rate": 6.854753926945663e-06, "loss": 0.2729, "step": 5120 }, { "epoch": 1.6531337019250172, "grad_norm": 0.35452881324106317, "learning_rate": 6.852094259699544e-06, "loss": 0.2882, "step": 5121 }, { "epoch": 1.653456555954639, "grad_norm": 0.4185709505780391, "learning_rate": 6.849434674542187e-06, "loss": 0.348, "step": 5122 }, { "epoch": 1.6537794099842609, "grad_norm": 0.40457994184280593, "learning_rate": 6.846775171810564e-06, "loss": 0.318, "step": 5123 }, { "epoch": 1.6541022640138827, "grad_norm": 0.339281384222836, "learning_rate": 6.844115751841623e-06, "loss": 0.2759, "step": 5124 }, { "epoch": 1.6544251180435046, "grad_norm": 0.4418593314798805, "learning_rate": 6.841456414972316e-06, "loss": 0.3715, "step": 5125 }, { "epoch": 1.6547479720731264, "grad_norm": 0.3646028810548989, "learning_rate": 6.838797161539572e-06, "loss": 0.2892, "step": 5126 }, { "epoch": 1.6550708261027482, "grad_norm": 0.38758936262956173, "learning_rate": 6.8361379918803196e-06, "loss": 0.2749, "step": 5127 }, { "epoch": 1.65539368013237, "grad_norm": 0.34844109612293944, "learning_rate": 6.833478906331468e-06, "loss": 0.3205, "step": 5128 }, { "epoch": 1.655716534161992, "grad_norm": 0.3296006959147034, "learning_rate": 6.8308199052299224e-06, "loss": 0.2337, "step": 5129 }, { "epoch": 1.6560393881916138, "grad_norm": 0.5268300807737312, "learning_rate": 6.828160988912572e-06, "loss": 0.2915, "step": 5130 }, { "epoch": 1.6563622422212356, "grad_norm": 0.40900193014869474, "learning_rate": 6.825502157716301e-06, "loss": 0.2562, "step": 5131 }, { "epoch": 1.6566850962508575, "grad_norm": 0.35498437903668945, "learning_rate": 6.822843411977975e-06, "loss": 0.2679, "step": 5132 }, { "epoch": 1.6570079502804793, "grad_norm": 0.38651038113592906, "learning_rate": 6.820184752034458e-06, "loss": 0.2812, "step": 5133 }, { "epoch": 1.6573308043101012, "grad_norm": 0.3461725638153822, "learning_rate": 6.8175261782225954e-06, "loss": 0.2551, "step": 5134 }, { "epoch": 1.657653658339723, "grad_norm": 0.4350563101537184, "learning_rate": 6.814867690879228e-06, "loss": 0.3227, "step": 5135 }, { "epoch": 1.6579765123693448, "grad_norm": 0.4648753494826876, "learning_rate": 6.812209290341178e-06, "loss": 0.275, "step": 5136 }, { "epoch": 1.658299366398967, "grad_norm": 0.3836703336636174, "learning_rate": 6.809550976945264e-06, "loss": 0.2727, "step": 5137 }, { "epoch": 1.6586222204285888, "grad_norm": 0.43796437963845947, "learning_rate": 6.806892751028295e-06, "loss": 0.2922, "step": 5138 }, { "epoch": 1.6589450744582106, "grad_norm": 0.3456569944372795, "learning_rate": 6.804234612927055e-06, "loss": 0.2813, "step": 5139 }, { "epoch": 1.6592679284878324, "grad_norm": 0.44780946271165556, "learning_rate": 6.801576562978334e-06, "loss": 0.307, "step": 5140 }, { "epoch": 1.6595907825174543, "grad_norm": 0.4615476314144694, "learning_rate": 6.7989186015189e-06, "loss": 0.329, "step": 5141 }, { "epoch": 1.6599136365470761, "grad_norm": 0.342655670837574, "learning_rate": 6.796260728885515e-06, "loss": 0.3074, "step": 5142 }, { "epoch": 1.660236490576698, "grad_norm": 0.4082805477161892, "learning_rate": 6.793602945414923e-06, "loss": 0.326, "step": 5143 }, { "epoch": 1.6605593446063198, "grad_norm": 0.41806932864753293, "learning_rate": 6.790945251443869e-06, "loss": 0.3912, "step": 5144 }, { "epoch": 1.6608821986359419, "grad_norm": 0.3757999723890151, "learning_rate": 6.788287647309072e-06, "loss": 0.3066, "step": 5145 }, { "epoch": 1.6612050526655637, "grad_norm": 0.3937091527789471, "learning_rate": 6.785630133347254e-06, "loss": 0.3012, "step": 5146 }, { "epoch": 1.6615279066951856, "grad_norm": 0.42431739094982446, "learning_rate": 6.78297270989511e-06, "loss": 0.3014, "step": 5147 }, { "epoch": 1.6618507607248074, "grad_norm": 0.410923779230694, "learning_rate": 6.780315377289339e-06, "loss": 0.3062, "step": 5148 }, { "epoch": 1.6621736147544293, "grad_norm": 0.39489090246602254, "learning_rate": 6.777658135866616e-06, "loss": 0.3036, "step": 5149 }, { "epoch": 1.662496468784051, "grad_norm": 0.4622229766880008, "learning_rate": 6.775000985963617e-06, "loss": 0.3106, "step": 5150 }, { "epoch": 1.662819322813673, "grad_norm": 0.39792425259772074, "learning_rate": 6.7723439279169895e-06, "loss": 0.284, "step": 5151 }, { "epoch": 1.6631421768432948, "grad_norm": 0.359687282752114, "learning_rate": 6.769686962063389e-06, "loss": 0.3006, "step": 5152 }, { "epoch": 1.6634650308729166, "grad_norm": 0.37719368212320153, "learning_rate": 6.767030088739441e-06, "loss": 0.3228, "step": 5153 }, { "epoch": 1.6637878849025385, "grad_norm": 0.3756436332156313, "learning_rate": 6.764373308281775e-06, "loss": 0.3386, "step": 5154 }, { "epoch": 1.6641107389321603, "grad_norm": 0.37836038702897, "learning_rate": 6.761716621026994e-06, "loss": 0.3494, "step": 5155 }, { "epoch": 1.6644335929617822, "grad_norm": 0.3680678392050936, "learning_rate": 6.7590600273117015e-06, "loss": 0.3058, "step": 5156 }, { "epoch": 1.664756446991404, "grad_norm": 0.4746807255488818, "learning_rate": 6.756403527472484e-06, "loss": 0.332, "step": 5157 }, { "epoch": 1.6650793010210259, "grad_norm": 0.36044248439652105, "learning_rate": 6.753747121845914e-06, "loss": 0.2869, "step": 5158 }, { "epoch": 1.6654021550506477, "grad_norm": 0.39150940301553433, "learning_rate": 6.7510908107685595e-06, "loss": 0.2812, "step": 5159 }, { "epoch": 1.6657250090802695, "grad_norm": 0.39588274061922774, "learning_rate": 6.748434594576964e-06, "loss": 0.3144, "step": 5160 }, { "epoch": 1.6660478631098914, "grad_norm": 0.43421336573828423, "learning_rate": 6.745778473607674e-06, "loss": 0.2954, "step": 5161 }, { "epoch": 1.6663707171395132, "grad_norm": 0.3969036383516019, "learning_rate": 6.74312244819721e-06, "loss": 0.3077, "step": 5162 }, { "epoch": 1.666693571169135, "grad_norm": 0.3733332913274998, "learning_rate": 6.740466518682091e-06, "loss": 0.2704, "step": 5163 }, { "epoch": 1.667016425198757, "grad_norm": 0.4154330687447364, "learning_rate": 6.737810685398818e-06, "loss": 0.277, "step": 5164 }, { "epoch": 1.6673392792283788, "grad_norm": 0.4011005721537231, "learning_rate": 6.735154948683882e-06, "loss": 0.2791, "step": 5165 }, { "epoch": 1.6676621332580006, "grad_norm": 0.3575875362628165, "learning_rate": 6.732499308873758e-06, "loss": 0.3016, "step": 5166 }, { "epoch": 1.6679849872876225, "grad_norm": 0.3998593839264106, "learning_rate": 6.729843766304918e-06, "loss": 0.3048, "step": 5167 }, { "epoch": 1.6683078413172443, "grad_norm": 0.40387356498470645, "learning_rate": 6.7271883213138105e-06, "loss": 0.3638, "step": 5168 }, { "epoch": 1.6686306953468661, "grad_norm": 0.391188922659017, "learning_rate": 6.72453297423688e-06, "loss": 0.3377, "step": 5169 }, { "epoch": 1.6689535493764882, "grad_norm": 0.39784983063208246, "learning_rate": 6.721877725410551e-06, "loss": 0.3058, "step": 5170 }, { "epoch": 1.66927640340611, "grad_norm": 0.3559460806458013, "learning_rate": 6.719222575171245e-06, "loss": 0.2637, "step": 5171 }, { "epoch": 1.669599257435732, "grad_norm": 0.4350999400520472, "learning_rate": 6.716567523855362e-06, "loss": 0.2853, "step": 5172 }, { "epoch": 1.6699221114653537, "grad_norm": 0.35528300719428846, "learning_rate": 6.713912571799298e-06, "loss": 0.2876, "step": 5173 }, { "epoch": 1.6702449654949756, "grad_norm": 0.37248028728232424, "learning_rate": 6.711257719339424e-06, "loss": 0.2917, "step": 5174 }, { "epoch": 1.6705678195245974, "grad_norm": 0.42197660137601595, "learning_rate": 6.708602966812113e-06, "loss": 0.2532, "step": 5175 }, { "epoch": 1.6708906735542193, "grad_norm": 0.35874576868437835, "learning_rate": 6.7059483145537175e-06, "loss": 0.3092, "step": 5176 }, { "epoch": 1.6712135275838411, "grad_norm": 0.3913694155756494, "learning_rate": 6.703293762900575e-06, "loss": 0.3132, "step": 5177 }, { "epoch": 1.671536381613463, "grad_norm": 0.3810880130218826, "learning_rate": 6.700639312189018e-06, "loss": 0.2947, "step": 5178 }, { "epoch": 1.671859235643085, "grad_norm": 0.4301975993369276, "learning_rate": 6.697984962755356e-06, "loss": 0.2753, "step": 5179 }, { "epoch": 1.6721820896727069, "grad_norm": 0.31511952109275515, "learning_rate": 6.695330714935898e-06, "loss": 0.2852, "step": 5180 }, { "epoch": 1.6725049437023287, "grad_norm": 0.43260814154682287, "learning_rate": 6.692676569066928e-06, "loss": 0.36, "step": 5181 }, { "epoch": 1.6728277977319506, "grad_norm": 0.43410119835686967, "learning_rate": 6.690022525484728e-06, "loss": 0.293, "step": 5182 }, { "epoch": 1.6731506517615724, "grad_norm": 0.3767964527921394, "learning_rate": 6.687368584525554e-06, "loss": 0.3378, "step": 5183 }, { "epoch": 1.6734735057911942, "grad_norm": 0.3697239314242563, "learning_rate": 6.684714746525667e-06, "loss": 0.3375, "step": 5184 }, { "epoch": 1.673796359820816, "grad_norm": 0.36727081177486137, "learning_rate": 6.682061011821294e-06, "loss": 0.2774, "step": 5185 }, { "epoch": 1.674119213850438, "grad_norm": 0.39487655484783635, "learning_rate": 6.679407380748666e-06, "loss": 0.2948, "step": 5186 }, { "epoch": 1.6744420678800598, "grad_norm": 0.4018745787123365, "learning_rate": 6.676753853643994e-06, "loss": 0.3246, "step": 5187 }, { "epoch": 1.6747649219096816, "grad_norm": 0.38350682695489685, "learning_rate": 6.674100430843475e-06, "loss": 0.315, "step": 5188 }, { "epoch": 1.6750877759393035, "grad_norm": 0.35054883902260986, "learning_rate": 6.671447112683291e-06, "loss": 0.2466, "step": 5189 }, { "epoch": 1.6754106299689253, "grad_norm": 0.3781985393004534, "learning_rate": 6.668793899499621e-06, "loss": 0.2626, "step": 5190 }, { "epoch": 1.6757334839985472, "grad_norm": 0.36112660125495055, "learning_rate": 6.666140791628616e-06, "loss": 0.3092, "step": 5191 }, { "epoch": 1.676056338028169, "grad_norm": 0.3222585858663578, "learning_rate": 6.6634877894064245e-06, "loss": 0.2857, "step": 5192 }, { "epoch": 1.6763791920577908, "grad_norm": 0.37497490513698084, "learning_rate": 6.660834893169182e-06, "loss": 0.2659, "step": 5193 }, { "epoch": 1.6767020460874127, "grad_norm": 0.37266060622005076, "learning_rate": 6.658182103252999e-06, "loss": 0.3211, "step": 5194 }, { "epoch": 1.6770249001170345, "grad_norm": 0.37498843865167225, "learning_rate": 6.655529419993986e-06, "loss": 0.2366, "step": 5195 }, { "epoch": 1.6773477541466564, "grad_norm": 0.3771021225757389, "learning_rate": 6.652876843728232e-06, "loss": 0.3162, "step": 5196 }, { "epoch": 1.6776706081762782, "grad_norm": 0.4270608402382048, "learning_rate": 6.650224374791816e-06, "loss": 0.2883, "step": 5197 }, { "epoch": 1.6779934622059, "grad_norm": 0.41702053859949983, "learning_rate": 6.6475720135208e-06, "loss": 0.2552, "step": 5198 }, { "epoch": 1.678316316235522, "grad_norm": 0.4334369910391575, "learning_rate": 6.644919760251238e-06, "loss": 0.271, "step": 5199 }, { "epoch": 1.6786391702651438, "grad_norm": 0.3334401680398576, "learning_rate": 6.642267615319163e-06, "loss": 0.2436, "step": 5200 }, { "epoch": 1.6789620242947656, "grad_norm": 0.4339624719398536, "learning_rate": 6.639615579060603e-06, "loss": 0.3308, "step": 5201 }, { "epoch": 1.6792848783243874, "grad_norm": 0.42076354434724644, "learning_rate": 6.6369636518115614e-06, "loss": 0.2656, "step": 5202 }, { "epoch": 1.6796077323540093, "grad_norm": 0.42845876973660413, "learning_rate": 6.634311833908037e-06, "loss": 0.3261, "step": 5203 }, { "epoch": 1.6799305863836314, "grad_norm": 0.4163671267099993, "learning_rate": 6.63166012568601e-06, "loss": 0.2945, "step": 5204 }, { "epoch": 1.6802534404132532, "grad_norm": 0.4686299072015517, "learning_rate": 6.629008527481453e-06, "loss": 0.3135, "step": 5205 }, { "epoch": 1.680576294442875, "grad_norm": 0.39853238551254344, "learning_rate": 6.626357039630312e-06, "loss": 0.2939, "step": 5206 }, { "epoch": 1.6808991484724969, "grad_norm": 0.6276841830236778, "learning_rate": 6.623705662468536e-06, "loss": 0.2798, "step": 5207 }, { "epoch": 1.6812220025021187, "grad_norm": 0.4040537289794548, "learning_rate": 6.6210543963320405e-06, "loss": 0.2754, "step": 5208 }, { "epoch": 1.6815448565317406, "grad_norm": 0.41560288634526, "learning_rate": 6.6184032415567465e-06, "loss": 0.279, "step": 5209 }, { "epoch": 1.6818677105613624, "grad_norm": 0.41186837699725376, "learning_rate": 6.615752198478546e-06, "loss": 0.3057, "step": 5210 }, { "epoch": 1.6821905645909843, "grad_norm": 0.777338700448737, "learning_rate": 6.613101267433323e-06, "loss": 0.2747, "step": 5211 }, { "epoch": 1.682513418620606, "grad_norm": 0.3421447903799921, "learning_rate": 6.6104504487569505e-06, "loss": 0.2637, "step": 5212 }, { "epoch": 1.6828362726502282, "grad_norm": 0.3894859834476371, "learning_rate": 6.607799742785278e-06, "loss": 0.2584, "step": 5213 }, { "epoch": 1.68315912667985, "grad_norm": 0.3665059674518832, "learning_rate": 6.605149149854155e-06, "loss": 0.3091, "step": 5214 }, { "epoch": 1.6834819807094719, "grad_norm": 0.4297384382405036, "learning_rate": 6.602498670299397e-06, "loss": 0.3587, "step": 5215 }, { "epoch": 1.6838048347390937, "grad_norm": 0.36202759272228374, "learning_rate": 6.599848304456825e-06, "loss": 0.3084, "step": 5216 }, { "epoch": 1.6841276887687155, "grad_norm": 0.34731801306292287, "learning_rate": 6.597198052662231e-06, "loss": 0.2992, "step": 5217 }, { "epoch": 1.6844505427983374, "grad_norm": 0.42368454458771937, "learning_rate": 6.5945479152514015e-06, "loss": 0.2969, "step": 5218 }, { "epoch": 1.6847733968279592, "grad_norm": 0.4467738028982834, "learning_rate": 6.591897892560105e-06, "loss": 0.2803, "step": 5219 }, { "epoch": 1.685096250857581, "grad_norm": 0.40254338957456826, "learning_rate": 6.589247984924094e-06, "loss": 0.3141, "step": 5220 }, { "epoch": 1.685419104887203, "grad_norm": 0.37266114389416843, "learning_rate": 6.586598192679108e-06, "loss": 0.2737, "step": 5221 }, { "epoch": 1.6857419589168248, "grad_norm": 0.45683915709177014, "learning_rate": 6.583948516160876e-06, "loss": 0.3067, "step": 5222 }, { "epoch": 1.6860648129464466, "grad_norm": 0.3385847684766518, "learning_rate": 6.581298955705103e-06, "loss": 0.2749, "step": 5223 }, { "epoch": 1.6863876669760685, "grad_norm": 0.4782184883826213, "learning_rate": 6.57864951164749e-06, "loss": 0.2843, "step": 5224 }, { "epoch": 1.6867105210056903, "grad_norm": 0.3986750321475926, "learning_rate": 6.5760001843237115e-06, "loss": 0.3206, "step": 5225 }, { "epoch": 1.6870333750353121, "grad_norm": 0.5197298752238678, "learning_rate": 6.57335097406944e-06, "loss": 0.2858, "step": 5226 }, { "epoch": 1.687356229064934, "grad_norm": 0.4063531798064161, "learning_rate": 6.570701881220321e-06, "loss": 0.2906, "step": 5227 }, { "epoch": 1.6876790830945558, "grad_norm": 0.3912322795689359, "learning_rate": 6.568052906111998e-06, "loss": 0.3445, "step": 5228 }, { "epoch": 1.6880019371241777, "grad_norm": 0.37945304645817896, "learning_rate": 6.565404049080086e-06, "loss": 0.3151, "step": 5229 }, { "epoch": 1.6883247911537995, "grad_norm": 0.5181637984873968, "learning_rate": 6.562755310460192e-06, "loss": 0.2734, "step": 5230 }, { "epoch": 1.6886476451834214, "grad_norm": 0.39639973872893064, "learning_rate": 6.560106690587914e-06, "loss": 0.294, "step": 5231 }, { "epoch": 1.6889704992130432, "grad_norm": 0.4354950325222672, "learning_rate": 6.55745818979882e-06, "loss": 0.3181, "step": 5232 }, { "epoch": 1.689293353242665, "grad_norm": 0.33453968551288393, "learning_rate": 6.554809808428479e-06, "loss": 0.3193, "step": 5233 }, { "epoch": 1.689616207272287, "grad_norm": 0.39142876745761945, "learning_rate": 6.5521615468124306e-06, "loss": 0.2694, "step": 5234 }, { "epoch": 1.6899390613019087, "grad_norm": 0.4057213643229583, "learning_rate": 6.549513405286211e-06, "loss": 0.276, "step": 5235 }, { "epoch": 1.6902619153315306, "grad_norm": 0.45856664452374657, "learning_rate": 6.546865384185331e-06, "loss": 0.3285, "step": 5236 }, { "epoch": 1.6905847693611524, "grad_norm": 0.38093157257230564, "learning_rate": 6.544217483845299e-06, "loss": 0.2869, "step": 5237 }, { "epoch": 1.6909076233907745, "grad_norm": 0.41895007275115254, "learning_rate": 6.541569704601591e-06, "loss": 0.3098, "step": 5238 }, { "epoch": 1.6912304774203963, "grad_norm": 0.4330085871517576, "learning_rate": 6.538922046789686e-06, "loss": 0.346, "step": 5239 }, { "epoch": 1.6915533314500182, "grad_norm": 0.3491123236210534, "learning_rate": 6.536274510745031e-06, "loss": 0.3255, "step": 5240 }, { "epoch": 1.69187618547964, "grad_norm": 0.3362521038889595, "learning_rate": 6.53362709680307e-06, "loss": 0.2721, "step": 5241 }, { "epoch": 1.6921990395092619, "grad_norm": 0.3793036265313247, "learning_rate": 6.530979805299225e-06, "loss": 0.2988, "step": 5242 }, { "epoch": 1.6925218935388837, "grad_norm": 0.41741381454154597, "learning_rate": 6.528332636568903e-06, "loss": 0.3295, "step": 5243 }, { "epoch": 1.6928447475685056, "grad_norm": 0.35464897352487007, "learning_rate": 6.5256855909474975e-06, "loss": 0.2623, "step": 5244 }, { "epoch": 1.6931676015981274, "grad_norm": 0.4231578877376744, "learning_rate": 6.5230386687703885e-06, "loss": 0.2996, "step": 5245 }, { "epoch": 1.6934904556277495, "grad_norm": 0.43335047821533884, "learning_rate": 6.5203918703729315e-06, "loss": 0.3113, "step": 5246 }, { "epoch": 1.6938133096573713, "grad_norm": 0.3707073721933738, "learning_rate": 6.5177451960904796e-06, "loss": 0.2654, "step": 5247 }, { "epoch": 1.6941361636869932, "grad_norm": 0.4291840070346685, "learning_rate": 6.5150986462583544e-06, "loss": 0.3496, "step": 5248 }, { "epoch": 1.694459017716615, "grad_norm": 0.4207158936344401, "learning_rate": 6.512452221211875e-06, "loss": 0.3642, "step": 5249 }, { "epoch": 1.6947818717462368, "grad_norm": 0.40784600625813916, "learning_rate": 6.509805921286342e-06, "loss": 0.3045, "step": 5250 }, { "epoch": 1.6951047257758587, "grad_norm": 0.35791962443406344, "learning_rate": 6.507159746817034e-06, "loss": 0.2942, "step": 5251 }, { "epoch": 1.6954275798054805, "grad_norm": 0.3833652405358677, "learning_rate": 6.504513698139219e-06, "loss": 0.278, "step": 5252 }, { "epoch": 1.6957504338351024, "grad_norm": 0.4385173275722838, "learning_rate": 6.501867775588146e-06, "loss": 0.339, "step": 5253 }, { "epoch": 1.6960732878647242, "grad_norm": 0.44090438310603625, "learning_rate": 6.499221979499055e-06, "loss": 0.2652, "step": 5254 }, { "epoch": 1.696396141894346, "grad_norm": 0.4662870984774905, "learning_rate": 6.496576310207157e-06, "loss": 0.3064, "step": 5255 }, { "epoch": 1.696718995923968, "grad_norm": 0.3959038861507378, "learning_rate": 6.493930768047662e-06, "loss": 0.3017, "step": 5256 }, { "epoch": 1.6970418499535898, "grad_norm": 0.37152363195109667, "learning_rate": 6.491285353355749e-06, "loss": 0.3111, "step": 5257 }, { "epoch": 1.6973647039832116, "grad_norm": 0.3493795026901185, "learning_rate": 6.488640066466596e-06, "loss": 0.3513, "step": 5258 }, { "epoch": 1.6976875580128334, "grad_norm": 0.4196512917600485, "learning_rate": 6.485994907715352e-06, "loss": 0.2884, "step": 5259 }, { "epoch": 1.6980104120424553, "grad_norm": 0.44632441803486894, "learning_rate": 6.4833498774371594e-06, "loss": 0.3338, "step": 5260 }, { "epoch": 1.6983332660720771, "grad_norm": 0.42448493458103465, "learning_rate": 6.480704975967135e-06, "loss": 0.3036, "step": 5261 }, { "epoch": 1.698656120101699, "grad_norm": 0.41269858515422453, "learning_rate": 6.478060203640389e-06, "loss": 0.3007, "step": 5262 }, { "epoch": 1.6989789741313208, "grad_norm": 0.4177023350213712, "learning_rate": 6.475415560792005e-06, "loss": 0.3444, "step": 5263 }, { "epoch": 1.6993018281609427, "grad_norm": 0.39828020834810896, "learning_rate": 6.47277104775706e-06, "loss": 0.3632, "step": 5264 }, { "epoch": 1.6996246821905645, "grad_norm": 0.38044609372133276, "learning_rate": 6.470126664870609e-06, "loss": 0.3199, "step": 5265 }, { "epoch": 1.6999475362201864, "grad_norm": 0.330578036136411, "learning_rate": 6.467482412467688e-06, "loss": 0.2487, "step": 5266 }, { "epoch": 1.7002703902498082, "grad_norm": 0.350474261428314, "learning_rate": 6.464838290883327e-06, "loss": 0.2941, "step": 5267 }, { "epoch": 1.70059324427943, "grad_norm": 0.39355997752237587, "learning_rate": 6.462194300452527e-06, "loss": 0.3018, "step": 5268 }, { "epoch": 1.7009160983090519, "grad_norm": 0.4232851284488262, "learning_rate": 6.459550441510283e-06, "loss": 0.2536, "step": 5269 }, { "epoch": 1.7012389523386737, "grad_norm": 0.48324947522936157, "learning_rate": 6.4569067143915625e-06, "loss": 0.2633, "step": 5270 }, { "epoch": 1.7015618063682956, "grad_norm": 0.4027079731038547, "learning_rate": 6.454263119431328e-06, "loss": 0.2639, "step": 5271 }, { "epoch": 1.7018846603979176, "grad_norm": 0.36147123194273234, "learning_rate": 6.4516196569645125e-06, "loss": 0.2249, "step": 5272 }, { "epoch": 1.7022075144275395, "grad_norm": 0.4167795689835278, "learning_rate": 6.448976327326045e-06, "loss": 0.3304, "step": 5273 }, { "epoch": 1.7025303684571613, "grad_norm": 0.3287059520503591, "learning_rate": 6.446333130850829e-06, "loss": 0.2554, "step": 5274 }, { "epoch": 1.7028532224867832, "grad_norm": 0.41385177381652966, "learning_rate": 6.443690067873754e-06, "loss": 0.359, "step": 5275 }, { "epoch": 1.703176076516405, "grad_norm": 0.4532022948656287, "learning_rate": 6.441047138729691e-06, "loss": 0.2619, "step": 5276 }, { "epoch": 1.7034989305460269, "grad_norm": 0.4545330291143829, "learning_rate": 6.438404343753502e-06, "loss": 0.2731, "step": 5277 }, { "epoch": 1.7038217845756487, "grad_norm": 0.3320546972065369, "learning_rate": 6.435761683280017e-06, "loss": 0.2558, "step": 5278 }, { "epoch": 1.7041446386052705, "grad_norm": 0.4055872752527504, "learning_rate": 6.433119157644064e-06, "loss": 0.3042, "step": 5279 }, { "epoch": 1.7044674926348926, "grad_norm": 0.39517321543945777, "learning_rate": 6.430476767180441e-06, "loss": 0.2783, "step": 5280 }, { "epoch": 1.7047903466645145, "grad_norm": 0.4234557860008341, "learning_rate": 6.427834512223941e-06, "loss": 0.2706, "step": 5281 }, { "epoch": 1.7051132006941363, "grad_norm": 0.40023455538067165, "learning_rate": 6.42519239310933e-06, "loss": 0.2762, "step": 5282 }, { "epoch": 1.7054360547237581, "grad_norm": 0.35061749976474293, "learning_rate": 6.422550410171365e-06, "loss": 0.3595, "step": 5283 }, { "epoch": 1.70575890875338, "grad_norm": 0.4225113029287837, "learning_rate": 6.4199085637447766e-06, "loss": 0.2667, "step": 5284 }, { "epoch": 1.7060817627830018, "grad_norm": 0.35922217243200416, "learning_rate": 6.417266854164287e-06, "loss": 0.2821, "step": 5285 }, { "epoch": 1.7064046168126237, "grad_norm": 0.42146154050940704, "learning_rate": 6.414625281764598e-06, "loss": 0.3302, "step": 5286 }, { "epoch": 1.7067274708422455, "grad_norm": 0.41958197608402137, "learning_rate": 6.411983846880388e-06, "loss": 0.2791, "step": 5287 }, { "epoch": 1.7070503248718674, "grad_norm": 0.40678217774101005, "learning_rate": 6.409342549846331e-06, "loss": 0.3249, "step": 5288 }, { "epoch": 1.7073731789014892, "grad_norm": 0.4686834788340464, "learning_rate": 6.4067013909970655e-06, "loss": 0.3175, "step": 5289 }, { "epoch": 1.707696032931111, "grad_norm": 0.3174629069592371, "learning_rate": 6.404060370667231e-06, "loss": 0.3115, "step": 5290 }, { "epoch": 1.708018886960733, "grad_norm": 0.3465388827671656, "learning_rate": 6.401419489191436e-06, "loss": 0.2867, "step": 5291 }, { "epoch": 1.7083417409903547, "grad_norm": 0.35203869494671713, "learning_rate": 6.398778746904283e-06, "loss": 0.3449, "step": 5292 }, { "epoch": 1.7086645950199766, "grad_norm": 0.40671166220658017, "learning_rate": 6.396138144140343e-06, "loss": 0.3129, "step": 5293 }, { "epoch": 1.7089874490495984, "grad_norm": 0.3803525028265709, "learning_rate": 6.393497681234183e-06, "loss": 0.2566, "step": 5294 }, { "epoch": 1.7093103030792203, "grad_norm": 0.34649690743933853, "learning_rate": 6.39085735852034e-06, "loss": 0.3056, "step": 5295 }, { "epoch": 1.7096331571088421, "grad_norm": 0.42180328109033977, "learning_rate": 6.388217176333344e-06, "loss": 0.3172, "step": 5296 }, { "epoch": 1.709956011138464, "grad_norm": 0.4231962202290393, "learning_rate": 6.3855771350077e-06, "loss": 0.2847, "step": 5297 }, { "epoch": 1.7102788651680858, "grad_norm": 0.3618250614170623, "learning_rate": 6.3829372348778985e-06, "loss": 0.3294, "step": 5298 }, { "epoch": 1.7106017191977076, "grad_norm": 0.4420768836405238, "learning_rate": 6.380297476278411e-06, "loss": 0.2638, "step": 5299 }, { "epoch": 1.7109245732273295, "grad_norm": 0.406686014651812, "learning_rate": 6.377657859543692e-06, "loss": 0.3048, "step": 5300 }, { "epoch": 1.7112474272569513, "grad_norm": 0.7336791290523165, "learning_rate": 6.375018385008175e-06, "loss": 0.2942, "step": 5301 }, { "epoch": 1.7115702812865732, "grad_norm": 0.44924807757764096, "learning_rate": 6.372379053006283e-06, "loss": 0.3387, "step": 5302 }, { "epoch": 1.711893135316195, "grad_norm": 0.3462651525496735, "learning_rate": 6.369739863872408e-06, "loss": 0.2483, "step": 5303 }, { "epoch": 1.7122159893458169, "grad_norm": 0.42161067006414715, "learning_rate": 6.367100817940936e-06, "loss": 0.2522, "step": 5304 }, { "epoch": 1.712538843375439, "grad_norm": 0.3968413722085201, "learning_rate": 6.364461915546232e-06, "loss": 0.3391, "step": 5305 }, { "epoch": 1.7128616974050608, "grad_norm": 0.41023280139067236, "learning_rate": 6.3618231570226385e-06, "loss": 0.2676, "step": 5306 }, { "epoch": 1.7131845514346826, "grad_norm": 0.3616605573609256, "learning_rate": 6.359184542704486e-06, "loss": 0.307, "step": 5307 }, { "epoch": 1.7135074054643045, "grad_norm": 0.3368486278383235, "learning_rate": 6.356546072926078e-06, "loss": 0.3138, "step": 5308 }, { "epoch": 1.7138302594939263, "grad_norm": 0.4083659365408918, "learning_rate": 6.353907748021712e-06, "loss": 0.3149, "step": 5309 }, { "epoch": 1.7141531135235482, "grad_norm": 0.3569931835934152, "learning_rate": 6.351269568325651e-06, "loss": 0.2836, "step": 5310 }, { "epoch": 1.71447596755317, "grad_norm": 0.40003241839189546, "learning_rate": 6.3486315341721585e-06, "loss": 0.3619, "step": 5311 }, { "epoch": 1.7147988215827918, "grad_norm": 0.4546597198359572, "learning_rate": 6.345993645895462e-06, "loss": 0.2943, "step": 5312 }, { "epoch": 1.7151216756124137, "grad_norm": 0.3854764200685924, "learning_rate": 6.343355903829781e-06, "loss": 0.2672, "step": 5313 }, { "epoch": 1.7154445296420358, "grad_norm": 0.41362931021514127, "learning_rate": 6.3407183083093145e-06, "loss": 0.3125, "step": 5314 }, { "epoch": 1.7157673836716576, "grad_norm": 0.4581496092375105, "learning_rate": 6.338080859668245e-06, "loss": 0.2515, "step": 5315 }, { "epoch": 1.7160902377012794, "grad_norm": 0.413416929383488, "learning_rate": 6.335443558240727e-06, "loss": 0.284, "step": 5316 }, { "epoch": 1.7164130917309013, "grad_norm": 0.39265379487671437, "learning_rate": 6.33280640436091e-06, "loss": 0.2887, "step": 5317 }, { "epoch": 1.7167359457605231, "grad_norm": 0.3766590996793365, "learning_rate": 6.3301693983629106e-06, "loss": 0.2808, "step": 5318 }, { "epoch": 1.717058799790145, "grad_norm": 0.36307470625755484, "learning_rate": 6.327532540580839e-06, "loss": 0.3467, "step": 5319 }, { "epoch": 1.7173816538197668, "grad_norm": 0.36944595210170583, "learning_rate": 6.324895831348779e-06, "loss": 0.2771, "step": 5320 }, { "epoch": 1.7177045078493887, "grad_norm": 0.35307716457202715, "learning_rate": 6.322259271000801e-06, "loss": 0.3089, "step": 5321 }, { "epoch": 1.7180273618790105, "grad_norm": 0.365037122479509, "learning_rate": 6.3196228598709485e-06, "loss": 0.3047, "step": 5322 }, { "epoch": 1.7183502159086324, "grad_norm": 0.3549812113632063, "learning_rate": 6.316986598293255e-06, "loss": 0.3032, "step": 5323 }, { "epoch": 1.7186730699382542, "grad_norm": 0.4025584559227368, "learning_rate": 6.314350486601734e-06, "loss": 0.2688, "step": 5324 }, { "epoch": 1.718995923967876, "grad_norm": 0.3303928512305026, "learning_rate": 6.3117145251303695e-06, "loss": 0.2368, "step": 5325 }, { "epoch": 1.7193187779974979, "grad_norm": 0.40882768847311324, "learning_rate": 6.309078714213141e-06, "loss": 0.3031, "step": 5326 }, { "epoch": 1.7196416320271197, "grad_norm": 0.3721234371856971, "learning_rate": 6.3064430541839976e-06, "loss": 0.3172, "step": 5327 }, { "epoch": 1.7199644860567416, "grad_norm": 0.38801338062703805, "learning_rate": 6.303807545376876e-06, "loss": 0.3163, "step": 5328 }, { "epoch": 1.7202873400863634, "grad_norm": 0.4102386261222869, "learning_rate": 6.30117218812569e-06, "loss": 0.3434, "step": 5329 }, { "epoch": 1.7206101941159853, "grad_norm": 0.3982190258738184, "learning_rate": 6.29853698276434e-06, "loss": 0.2831, "step": 5330 }, { "epoch": 1.720933048145607, "grad_norm": 0.45313341855019945, "learning_rate": 6.2959019296266975e-06, "loss": 0.3056, "step": 5331 }, { "epoch": 1.721255902175229, "grad_norm": 0.49120445672942065, "learning_rate": 6.293267029046624e-06, "loss": 0.3401, "step": 5332 }, { "epoch": 1.7215787562048508, "grad_norm": 0.349635816746031, "learning_rate": 6.2906322813579545e-06, "loss": 0.3077, "step": 5333 }, { "epoch": 1.7219016102344726, "grad_norm": 0.3871131903325668, "learning_rate": 6.287997686894512e-06, "loss": 0.3348, "step": 5334 }, { "epoch": 1.7222244642640945, "grad_norm": 0.4328953694267899, "learning_rate": 6.28536324599009e-06, "loss": 0.2849, "step": 5335 }, { "epoch": 1.7225473182937163, "grad_norm": 0.33326680206971215, "learning_rate": 6.282728958978473e-06, "loss": 0.2837, "step": 5336 }, { "epoch": 1.7228701723233382, "grad_norm": 0.33412026636189346, "learning_rate": 6.280094826193421e-06, "loss": 0.2896, "step": 5337 }, { "epoch": 1.72319302635296, "grad_norm": 0.4163256675493144, "learning_rate": 6.277460847968677e-06, "loss": 0.2584, "step": 5338 }, { "epoch": 1.723515880382582, "grad_norm": 0.39421329300057734, "learning_rate": 6.274827024637957e-06, "loss": 0.2171, "step": 5339 }, { "epoch": 1.723838734412204, "grad_norm": 0.4183461636416767, "learning_rate": 6.272193356534965e-06, "loss": 0.3088, "step": 5340 }, { "epoch": 1.7241615884418258, "grad_norm": 0.36126578448794894, "learning_rate": 6.269559843993387e-06, "loss": 0.3309, "step": 5341 }, { "epoch": 1.7244844424714476, "grad_norm": 0.4492433127632595, "learning_rate": 6.26692648734688e-06, "loss": 0.349, "step": 5342 }, { "epoch": 1.7248072965010695, "grad_norm": 0.35058996348286536, "learning_rate": 6.264293286929092e-06, "loss": 0.2568, "step": 5343 }, { "epoch": 1.7251301505306913, "grad_norm": 0.48996531302420837, "learning_rate": 6.26166024307364e-06, "loss": 0.2984, "step": 5344 }, { "epoch": 1.7254530045603131, "grad_norm": 0.5219623690186753, "learning_rate": 6.25902735611413e-06, "loss": 0.316, "step": 5345 }, { "epoch": 1.725775858589935, "grad_norm": 0.38443358686503926, "learning_rate": 6.256394626384145e-06, "loss": 0.2958, "step": 5346 }, { "epoch": 1.7260987126195568, "grad_norm": 0.3716137035974319, "learning_rate": 6.253762054217251e-06, "loss": 0.3355, "step": 5347 }, { "epoch": 1.726421566649179, "grad_norm": 0.37047959314589496, "learning_rate": 6.251129639946985e-06, "loss": 0.3105, "step": 5348 }, { "epoch": 1.7267444206788007, "grad_norm": 0.37897979925643643, "learning_rate": 6.2484973839068774e-06, "loss": 0.296, "step": 5349 }, { "epoch": 1.7270672747084226, "grad_norm": 0.38801156005347126, "learning_rate": 6.245865286430425e-06, "loss": 0.3401, "step": 5350 }, { "epoch": 1.7273901287380444, "grad_norm": 0.43764495140431725, "learning_rate": 6.243233347851115e-06, "loss": 0.2577, "step": 5351 }, { "epoch": 1.7277129827676663, "grad_norm": 0.5332109443783325, "learning_rate": 6.240601568502407e-06, "loss": 0.2984, "step": 5352 }, { "epoch": 1.7280358367972881, "grad_norm": 0.41653134893183025, "learning_rate": 6.23796994871775e-06, "loss": 0.2986, "step": 5353 }, { "epoch": 1.72835869082691, "grad_norm": 0.4031230339871719, "learning_rate": 6.23533848883056e-06, "loss": 0.286, "step": 5354 }, { "epoch": 1.7286815448565318, "grad_norm": 0.39613514449859055, "learning_rate": 6.232707189174243e-06, "loss": 0.3078, "step": 5355 }, { "epoch": 1.7290043988861536, "grad_norm": 0.41126489365636937, "learning_rate": 6.230076050082179e-06, "loss": 0.2522, "step": 5356 }, { "epoch": 1.7293272529157755, "grad_norm": 0.3839943212950905, "learning_rate": 6.227445071887733e-06, "loss": 0.3021, "step": 5357 }, { "epoch": 1.7296501069453973, "grad_norm": 0.41738610399156956, "learning_rate": 6.22481425492424e-06, "loss": 0.2637, "step": 5358 }, { "epoch": 1.7299729609750192, "grad_norm": 4.371378800175724, "learning_rate": 6.2221835995250256e-06, "loss": 0.29, "step": 5359 }, { "epoch": 1.730295815004641, "grad_norm": 0.4610292409657048, "learning_rate": 6.219553106023392e-06, "loss": 0.2907, "step": 5360 }, { "epoch": 1.7306186690342629, "grad_norm": 0.39333085754950065, "learning_rate": 6.216922774752614e-06, "loss": 0.2798, "step": 5361 }, { "epoch": 1.7309415230638847, "grad_norm": 0.3896669041978629, "learning_rate": 6.214292606045956e-06, "loss": 0.351, "step": 5362 }, { "epoch": 1.7312643770935066, "grad_norm": 0.4084496202156179, "learning_rate": 6.2116626002366526e-06, "loss": 0.2951, "step": 5363 }, { "epoch": 1.7315872311231284, "grad_norm": 0.3859876048426515, "learning_rate": 6.2090327576579255e-06, "loss": 0.2691, "step": 5364 }, { "epoch": 1.7319100851527502, "grad_norm": 0.34963069738437835, "learning_rate": 6.206403078642967e-06, "loss": 0.2987, "step": 5365 }, { "epoch": 1.732232939182372, "grad_norm": 0.4926830840903704, "learning_rate": 6.203773563524962e-06, "loss": 0.2775, "step": 5366 }, { "epoch": 1.732555793211994, "grad_norm": 0.3767175736190219, "learning_rate": 6.201144212637057e-06, "loss": 0.3178, "step": 5367 }, { "epoch": 1.7328786472416158, "grad_norm": 0.33965395884199623, "learning_rate": 6.198515026312393e-06, "loss": 0.2406, "step": 5368 }, { "epoch": 1.7332015012712376, "grad_norm": 0.38756279912810093, "learning_rate": 6.195886004884082e-06, "loss": 0.2861, "step": 5369 }, { "epoch": 1.7335243553008595, "grad_norm": 0.43718996289384243, "learning_rate": 6.193257148685221e-06, "loss": 0.3085, "step": 5370 }, { "epoch": 1.7338472093304813, "grad_norm": 0.40919013693715944, "learning_rate": 6.1906284580488765e-06, "loss": 0.3867, "step": 5371 }, { "epoch": 1.7341700633601032, "grad_norm": 0.4505889550443009, "learning_rate": 6.187999933308107e-06, "loss": 0.245, "step": 5372 }, { "epoch": 1.7344929173897252, "grad_norm": 0.3436920255980016, "learning_rate": 6.185371574795935e-06, "loss": 0.3015, "step": 5373 }, { "epoch": 1.734815771419347, "grad_norm": 0.4136569654826856, "learning_rate": 6.182743382845377e-06, "loss": 0.3286, "step": 5374 }, { "epoch": 1.735138625448969, "grad_norm": 0.43464442748427107, "learning_rate": 6.180115357789416e-06, "loss": 0.2652, "step": 5375 }, { "epoch": 1.7354614794785908, "grad_norm": 0.4523889234316265, "learning_rate": 6.177487499961025e-06, "loss": 0.2742, "step": 5376 }, { "epoch": 1.7357843335082126, "grad_norm": 0.3921105933132553, "learning_rate": 6.174859809693144e-06, "loss": 0.3301, "step": 5377 }, { "epoch": 1.7361071875378344, "grad_norm": 0.4224717855450254, "learning_rate": 6.172232287318699e-06, "loss": 0.2714, "step": 5378 }, { "epoch": 1.7364300415674563, "grad_norm": 0.3512997974563588, "learning_rate": 6.169604933170599e-06, "loss": 0.3335, "step": 5379 }, { "epoch": 1.7367528955970781, "grad_norm": 0.39249896628980707, "learning_rate": 6.166977747581719e-06, "loss": 0.2427, "step": 5380 }, { "epoch": 1.7370757496267002, "grad_norm": 0.461777431012935, "learning_rate": 6.1643507308849245e-06, "loss": 0.3258, "step": 5381 }, { "epoch": 1.737398603656322, "grad_norm": 0.38972171945916445, "learning_rate": 6.1617238834130515e-06, "loss": 0.3029, "step": 5382 }, { "epoch": 1.7377214576859439, "grad_norm": 0.36632764416007574, "learning_rate": 6.15909720549892e-06, "loss": 0.2664, "step": 5383 }, { "epoch": 1.7380443117155657, "grad_norm": 0.3565752825604332, "learning_rate": 6.1564706974753246e-06, "loss": 0.2777, "step": 5384 }, { "epoch": 1.7383671657451876, "grad_norm": 0.3531524975037431, "learning_rate": 6.1538443596750445e-06, "loss": 0.2673, "step": 5385 }, { "epoch": 1.7386900197748094, "grad_norm": 0.3654942347580973, "learning_rate": 6.151218192430828e-06, "loss": 0.3145, "step": 5386 }, { "epoch": 1.7390128738044313, "grad_norm": 0.3561319123838858, "learning_rate": 6.148592196075411e-06, "loss": 0.3203, "step": 5387 }, { "epoch": 1.739335727834053, "grad_norm": 0.384390382416985, "learning_rate": 6.1459663709415e-06, "loss": 0.308, "step": 5388 }, { "epoch": 1.739658581863675, "grad_norm": 0.3223169287797435, "learning_rate": 6.1433407173617884e-06, "loss": 0.2576, "step": 5389 }, { "epoch": 1.7399814358932968, "grad_norm": 0.6271728326315602, "learning_rate": 6.1407152356689365e-06, "loss": 0.2602, "step": 5390 }, { "epoch": 1.7403042899229186, "grad_norm": 0.46097110614819503, "learning_rate": 6.1380899261955924e-06, "loss": 0.291, "step": 5391 }, { "epoch": 1.7406271439525405, "grad_norm": 0.41145741415262127, "learning_rate": 6.1354647892743786e-06, "loss": 0.369, "step": 5392 }, { "epoch": 1.7409499979821623, "grad_norm": 0.35024044272304183, "learning_rate": 6.1328398252379e-06, "loss": 0.2609, "step": 5393 }, { "epoch": 1.7412728520117842, "grad_norm": 0.37882928485463097, "learning_rate": 6.13021503441873e-06, "loss": 0.4123, "step": 5394 }, { "epoch": 1.741595706041406, "grad_norm": 0.5598025878629865, "learning_rate": 6.1275904171494326e-06, "loss": 0.318, "step": 5395 }, { "epoch": 1.7419185600710279, "grad_norm": 0.34945377543124806, "learning_rate": 6.1249659737625346e-06, "loss": 0.2683, "step": 5396 }, { "epoch": 1.7422414141006497, "grad_norm": 0.38606578741785214, "learning_rate": 6.122341704590555e-06, "loss": 0.2514, "step": 5397 }, { "epoch": 1.7425642681302715, "grad_norm": 0.3850170396491347, "learning_rate": 6.1197176099659876e-06, "loss": 0.2851, "step": 5398 }, { "epoch": 1.7428871221598934, "grad_norm": 0.43671540339714154, "learning_rate": 6.117093690221295e-06, "loss": 0.3012, "step": 5399 }, { "epoch": 1.7432099761895152, "grad_norm": 0.45446560935044356, "learning_rate": 6.1144699456889285e-06, "loss": 0.3088, "step": 5400 }, { "epoch": 1.743532830219137, "grad_norm": 0.527631846786702, "learning_rate": 6.111846376701311e-06, "loss": 0.2684, "step": 5401 }, { "epoch": 1.743855684248759, "grad_norm": 0.42956805801621395, "learning_rate": 6.1092229835908495e-06, "loss": 0.3341, "step": 5402 }, { "epoch": 1.7441785382783808, "grad_norm": 0.39087444442980285, "learning_rate": 6.106599766689916e-06, "loss": 0.3108, "step": 5403 }, { "epoch": 1.7445013923080026, "grad_norm": 0.37508538193329016, "learning_rate": 6.103976726330879e-06, "loss": 0.3179, "step": 5404 }, { "epoch": 1.7448242463376245, "grad_norm": 0.406338225298442, "learning_rate": 6.101353862846063e-06, "loss": 0.3033, "step": 5405 }, { "epoch": 1.7451471003672463, "grad_norm": 0.45617639053747433, "learning_rate": 6.098731176567788e-06, "loss": 0.2882, "step": 5406 }, { "epoch": 1.7454699543968684, "grad_norm": 0.40193096468779876, "learning_rate": 6.096108667828343e-06, "loss": 0.2836, "step": 5407 }, { "epoch": 1.7457928084264902, "grad_norm": 0.36625408230533996, "learning_rate": 6.093486336959999e-06, "loss": 0.2948, "step": 5408 }, { "epoch": 1.746115662456112, "grad_norm": 0.38876411232260394, "learning_rate": 6.090864184294996e-06, "loss": 0.26, "step": 5409 }, { "epoch": 1.746438516485734, "grad_norm": 0.40814752354979134, "learning_rate": 6.0882422101655645e-06, "loss": 0.3791, "step": 5410 }, { "epoch": 1.7467613705153557, "grad_norm": 0.41174312589480494, "learning_rate": 6.085620414903896e-06, "loss": 0.3524, "step": 5411 }, { "epoch": 1.7470842245449776, "grad_norm": 0.3346988484538372, "learning_rate": 6.082998798842178e-06, "loss": 0.3127, "step": 5412 }, { "epoch": 1.7474070785745994, "grad_norm": 0.41349698371342725, "learning_rate": 6.080377362312556e-06, "loss": 0.3189, "step": 5413 }, { "epoch": 1.7477299326042213, "grad_norm": 0.4059847941623417, "learning_rate": 6.077756105647169e-06, "loss": 0.3301, "step": 5414 }, { "epoch": 1.7480527866338433, "grad_norm": 0.34703130319033937, "learning_rate": 6.075135029178126e-06, "loss": 0.3637, "step": 5415 }, { "epoch": 1.7483756406634652, "grad_norm": 0.32405374012496707, "learning_rate": 6.07251413323751e-06, "loss": 0.242, "step": 5416 }, { "epoch": 1.748698494693087, "grad_norm": 0.39192358491055307, "learning_rate": 6.069893418157392e-06, "loss": 0.3113, "step": 5417 }, { "epoch": 1.7490213487227089, "grad_norm": 0.43216935631640635, "learning_rate": 6.067272884269804e-06, "loss": 0.3356, "step": 5418 }, { "epoch": 1.7493442027523307, "grad_norm": 0.5417115391266822, "learning_rate": 6.064652531906774e-06, "loss": 0.2824, "step": 5419 }, { "epoch": 1.7496670567819526, "grad_norm": 0.4004530027265118, "learning_rate": 6.062032361400286e-06, "loss": 0.3087, "step": 5420 }, { "epoch": 1.7499899108115744, "grad_norm": 0.4143354894851902, "learning_rate": 6.059412373082322e-06, "loss": 0.2573, "step": 5421 }, { "epoch": 1.7503127648411962, "grad_norm": 0.431467845409809, "learning_rate": 6.056792567284823e-06, "loss": 0.2496, "step": 5422 }, { "epoch": 1.750635618870818, "grad_norm": 0.37492693347883044, "learning_rate": 6.05417294433972e-06, "loss": 0.3084, "step": 5423 }, { "epoch": 1.75095847290044, "grad_norm": 0.3229158662528019, "learning_rate": 6.051553504578913e-06, "loss": 0.267, "step": 5424 }, { "epoch": 1.7512813269300618, "grad_norm": 0.4198701036234845, "learning_rate": 6.0489342483342845e-06, "loss": 0.2684, "step": 5425 }, { "epoch": 1.7516041809596836, "grad_norm": 0.7643904644135052, "learning_rate": 6.0463151759376855e-06, "loss": 0.327, "step": 5426 }, { "epoch": 1.7519270349893055, "grad_norm": 0.38757071560830003, "learning_rate": 6.043696287720956e-06, "loss": 0.2436, "step": 5427 }, { "epoch": 1.7522498890189273, "grad_norm": 0.41332844357712567, "learning_rate": 6.041077584015898e-06, "loss": 0.3086, "step": 5428 }, { "epoch": 1.7525727430485492, "grad_norm": 0.35988996869654516, "learning_rate": 6.038459065154301e-06, "loss": 0.2727, "step": 5429 }, { "epoch": 1.752895597078171, "grad_norm": 0.41428200724984376, "learning_rate": 6.035840731467928e-06, "loss": 0.297, "step": 5430 }, { "epoch": 1.7532184511077928, "grad_norm": 0.4095577288689146, "learning_rate": 6.033222583288521e-06, "loss": 0.2811, "step": 5431 }, { "epoch": 1.7535413051374147, "grad_norm": 0.34434761966096966, "learning_rate": 6.030604620947789e-06, "loss": 0.3154, "step": 5432 }, { "epoch": 1.7538641591670365, "grad_norm": 0.3179196489392863, "learning_rate": 6.027986844777428e-06, "loss": 0.2562, "step": 5433 }, { "epoch": 1.7541870131966584, "grad_norm": 0.37059712405217116, "learning_rate": 6.0253692551091105e-06, "loss": 0.2758, "step": 5434 }, { "epoch": 1.7545098672262802, "grad_norm": 0.9274685595652147, "learning_rate": 6.022751852274474e-06, "loss": 0.2664, "step": 5435 }, { "epoch": 1.754832721255902, "grad_norm": 0.3492013016307682, "learning_rate": 6.020134636605147e-06, "loss": 0.3031, "step": 5436 }, { "epoch": 1.755155575285524, "grad_norm": 0.3920612844359965, "learning_rate": 6.017517608432719e-06, "loss": 0.3141, "step": 5437 }, { "epoch": 1.7554784293151457, "grad_norm": 0.3883600670567535, "learning_rate": 6.0149007680887725e-06, "loss": 0.2927, "step": 5438 }, { "epoch": 1.7558012833447676, "grad_norm": 0.43054394513021266, "learning_rate": 6.012284115904851e-06, "loss": 0.2499, "step": 5439 }, { "epoch": 1.7561241373743894, "grad_norm": 0.3658499850091626, "learning_rate": 6.009667652212486e-06, "loss": 0.3383, "step": 5440 }, { "epoch": 1.7564469914040115, "grad_norm": 0.35938730447386885, "learning_rate": 6.007051377343174e-06, "loss": 0.3474, "step": 5441 }, { "epoch": 1.7567698454336333, "grad_norm": 0.41075851842604955, "learning_rate": 6.004435291628401e-06, "loss": 0.2943, "step": 5442 }, { "epoch": 1.7570926994632552, "grad_norm": 0.43006828065429675, "learning_rate": 6.001819395399613e-06, "loss": 0.3189, "step": 5443 }, { "epoch": 1.757415553492877, "grad_norm": 0.40749054102241816, "learning_rate": 5.999203688988247e-06, "loss": 0.3157, "step": 5444 }, { "epoch": 1.7577384075224989, "grad_norm": 0.3320896257714008, "learning_rate": 5.996588172725705e-06, "loss": 0.2818, "step": 5445 }, { "epoch": 1.7580612615521207, "grad_norm": 0.34381428002721753, "learning_rate": 5.993972846943372e-06, "loss": 0.304, "step": 5446 }, { "epoch": 1.7583841155817426, "grad_norm": 0.4730309769127142, "learning_rate": 5.991357711972603e-06, "loss": 0.2798, "step": 5447 }, { "epoch": 1.7587069696113644, "grad_norm": 0.5049188321277693, "learning_rate": 5.9887427681447385e-06, "loss": 0.3366, "step": 5448 }, { "epoch": 1.7590298236409865, "grad_norm": 0.4158012561780119, "learning_rate": 5.986128015791081e-06, "loss": 0.2808, "step": 5449 }, { "epoch": 1.7593526776706083, "grad_norm": 0.4011911164141758, "learning_rate": 5.983513455242923e-06, "loss": 0.3123, "step": 5450 }, { "epoch": 1.7596755317002302, "grad_norm": 0.4758423059425398, "learning_rate": 5.980899086831518e-06, "loss": 0.3059, "step": 5451 }, { "epoch": 1.759998385729852, "grad_norm": 0.42873968199232804, "learning_rate": 5.978284910888107e-06, "loss": 0.2902, "step": 5452 }, { "epoch": 1.7603212397594739, "grad_norm": 0.3577418383604496, "learning_rate": 5.975670927743904e-06, "loss": 0.2822, "step": 5453 }, { "epoch": 1.7606440937890957, "grad_norm": 0.3864581734836529, "learning_rate": 5.973057137730094e-06, "loss": 0.3393, "step": 5454 }, { "epoch": 1.7609669478187175, "grad_norm": 0.47482409349828114, "learning_rate": 5.970443541177843e-06, "loss": 0.3788, "step": 5455 }, { "epoch": 1.7612898018483394, "grad_norm": 0.43649888228824746, "learning_rate": 5.9678301384182874e-06, "loss": 0.3226, "step": 5456 }, { "epoch": 1.7616126558779612, "grad_norm": 0.3679099347260598, "learning_rate": 5.965216929782547e-06, "loss": 0.3493, "step": 5457 }, { "epoch": 1.761935509907583, "grad_norm": 0.41561556632405733, "learning_rate": 5.962603915601703e-06, "loss": 0.2759, "step": 5458 }, { "epoch": 1.762258363937205, "grad_norm": 0.34485756956046365, "learning_rate": 5.959991096206831e-06, "loss": 0.3028, "step": 5459 }, { "epoch": 1.7625812179668268, "grad_norm": 0.3136455300079347, "learning_rate": 5.957378471928962e-06, "loss": 0.2628, "step": 5460 }, { "epoch": 1.7629040719964486, "grad_norm": 0.36350376290466363, "learning_rate": 5.954766043099119e-06, "loss": 0.3018, "step": 5461 }, { "epoch": 1.7632269260260705, "grad_norm": 0.41102918706600156, "learning_rate": 5.952153810048288e-06, "loss": 0.2759, "step": 5462 }, { "epoch": 1.7635497800556923, "grad_norm": 0.5325357904902228, "learning_rate": 5.949541773107441e-06, "loss": 0.3529, "step": 5463 }, { "epoch": 1.7638726340853141, "grad_norm": 0.4026734263054087, "learning_rate": 5.9469299326075135e-06, "loss": 0.3311, "step": 5464 }, { "epoch": 1.764195488114936, "grad_norm": 0.388174311263594, "learning_rate": 5.944318288879428e-06, "loss": 0.305, "step": 5465 }, { "epoch": 1.7645183421445578, "grad_norm": 0.4118274746464863, "learning_rate": 5.9417068422540695e-06, "loss": 0.289, "step": 5466 }, { "epoch": 1.7648411961741797, "grad_norm": 0.42572679931507323, "learning_rate": 5.939095593062311e-06, "loss": 0.3454, "step": 5467 }, { "epoch": 1.7651640502038015, "grad_norm": 0.4196979794183185, "learning_rate": 5.936484541634989e-06, "loss": 0.3074, "step": 5468 }, { "epoch": 1.7654869042334234, "grad_norm": 0.40733692192328874, "learning_rate": 5.933873688302921e-06, "loss": 0.2978, "step": 5469 }, { "epoch": 1.7658097582630452, "grad_norm": 0.49375137769989075, "learning_rate": 5.931263033396903e-06, "loss": 0.3027, "step": 5470 }, { "epoch": 1.766132612292667, "grad_norm": 0.38521463410733237, "learning_rate": 5.928652577247696e-06, "loss": 0.2719, "step": 5471 }, { "epoch": 1.766455466322289, "grad_norm": 0.3675685064618591, "learning_rate": 5.926042320186047e-06, "loss": 0.2798, "step": 5472 }, { "epoch": 1.7667783203519107, "grad_norm": 0.41935752811246396, "learning_rate": 5.923432262542666e-06, "loss": 0.256, "step": 5473 }, { "epoch": 1.7671011743815328, "grad_norm": 0.42095621531342714, "learning_rate": 5.920822404648249e-06, "loss": 0.2495, "step": 5474 }, { "epoch": 1.7674240284111546, "grad_norm": 0.384212078925661, "learning_rate": 5.9182127468334565e-06, "loss": 0.2327, "step": 5475 }, { "epoch": 1.7677468824407765, "grad_norm": 0.3826240151107739, "learning_rate": 5.915603289428932e-06, "loss": 0.2785, "step": 5476 }, { "epoch": 1.7680697364703983, "grad_norm": 0.39285125339019333, "learning_rate": 5.912994032765287e-06, "loss": 0.3052, "step": 5477 }, { "epoch": 1.7683925905000202, "grad_norm": 0.332710004386641, "learning_rate": 5.910384977173114e-06, "loss": 0.2542, "step": 5478 }, { "epoch": 1.768715444529642, "grad_norm": 0.41960951143586256, "learning_rate": 5.907776122982974e-06, "loss": 0.316, "step": 5479 }, { "epoch": 1.7690382985592639, "grad_norm": 0.3449959274728714, "learning_rate": 5.905167470525409e-06, "loss": 0.3359, "step": 5480 }, { "epoch": 1.7693611525888857, "grad_norm": 0.43511202811965116, "learning_rate": 5.902559020130927e-06, "loss": 0.3118, "step": 5481 }, { "epoch": 1.7696840066185076, "grad_norm": 0.4160922744426565, "learning_rate": 5.8999507721300195e-06, "loss": 0.2774, "step": 5482 }, { "epoch": 1.7700068606481296, "grad_norm": 0.40050141322262667, "learning_rate": 5.897342726853144e-06, "loss": 0.3363, "step": 5483 }, { "epoch": 1.7703297146777515, "grad_norm": 0.5232988414668338, "learning_rate": 5.8947348846307376e-06, "loss": 0.3193, "step": 5484 }, { "epoch": 1.7706525687073733, "grad_norm": 0.35976622636195543, "learning_rate": 5.89212724579321e-06, "loss": 0.3194, "step": 5485 }, { "epoch": 1.7709754227369952, "grad_norm": 0.47036790039429316, "learning_rate": 5.889519810670951e-06, "loss": 0.2878, "step": 5486 }, { "epoch": 1.771298276766617, "grad_norm": 0.4541698263866414, "learning_rate": 5.886912579594308e-06, "loss": 0.2799, "step": 5487 }, { "epoch": 1.7716211307962388, "grad_norm": 0.34913927571869446, "learning_rate": 5.884305552893623e-06, "loss": 0.3044, "step": 5488 }, { "epoch": 1.7719439848258607, "grad_norm": 0.30329253733156875, "learning_rate": 5.8816987308992e-06, "loss": 0.3049, "step": 5489 }, { "epoch": 1.7722668388554825, "grad_norm": 0.35953947960817967, "learning_rate": 5.879092113941318e-06, "loss": 0.2729, "step": 5490 }, { "epoch": 1.7725896928851044, "grad_norm": 0.43829129848037957, "learning_rate": 5.876485702350235e-06, "loss": 0.2969, "step": 5491 }, { "epoch": 1.7729125469147262, "grad_norm": 0.39782067187787284, "learning_rate": 5.873879496456176e-06, "loss": 0.312, "step": 5492 }, { "epoch": 1.773235400944348, "grad_norm": 0.3379463304351298, "learning_rate": 5.871273496589346e-06, "loss": 0.29, "step": 5493 }, { "epoch": 1.77355825497397, "grad_norm": 0.42122907116735037, "learning_rate": 5.86866770307992e-06, "loss": 0.2874, "step": 5494 }, { "epoch": 1.7738811090035917, "grad_norm": 0.6170816682375692, "learning_rate": 5.866062116258052e-06, "loss": 0.2672, "step": 5495 }, { "epoch": 1.7742039630332136, "grad_norm": 0.5405183861411257, "learning_rate": 5.863456736453861e-06, "loss": 0.2587, "step": 5496 }, { "epoch": 1.7745268170628354, "grad_norm": 0.35485809783675876, "learning_rate": 5.860851563997451e-06, "loss": 0.3003, "step": 5497 }, { "epoch": 1.7748496710924573, "grad_norm": 0.397566384222925, "learning_rate": 5.858246599218888e-06, "loss": 0.3365, "step": 5498 }, { "epoch": 1.7751725251220791, "grad_norm": 0.38646568153822064, "learning_rate": 5.85564184244822e-06, "loss": 0.2756, "step": 5499 }, { "epoch": 1.775495379151701, "grad_norm": 0.385401666365679, "learning_rate": 5.853037294015467e-06, "loss": 0.2531, "step": 5500 }, { "epoch": 1.7758182331813228, "grad_norm": 0.37754047311300093, "learning_rate": 5.850432954250622e-06, "loss": 0.2877, "step": 5501 }, { "epoch": 1.7761410872109447, "grad_norm": 0.41807443763608976, "learning_rate": 5.8478288234836475e-06, "loss": 0.329, "step": 5502 }, { "epoch": 1.7764639412405665, "grad_norm": 0.3838269578234552, "learning_rate": 5.845224902044489e-06, "loss": 0.3151, "step": 5503 }, { "epoch": 1.7767867952701883, "grad_norm": 0.4053507147317904, "learning_rate": 5.8426211902630535e-06, "loss": 0.2937, "step": 5504 }, { "epoch": 1.7771096492998102, "grad_norm": 0.337091459562413, "learning_rate": 5.840017688469234e-06, "loss": 0.2436, "step": 5505 }, { "epoch": 1.777432503329432, "grad_norm": 0.3713270438558607, "learning_rate": 5.837414396992885e-06, "loss": 0.3222, "step": 5506 }, { "epoch": 1.7777553573590539, "grad_norm": 0.6217123411816575, "learning_rate": 5.834811316163842e-06, "loss": 0.3142, "step": 5507 }, { "epoch": 1.778078211388676, "grad_norm": 0.3828447396469399, "learning_rate": 5.832208446311913e-06, "loss": 0.259, "step": 5508 }, { "epoch": 1.7784010654182978, "grad_norm": 0.4066105600697875, "learning_rate": 5.829605787766878e-06, "loss": 0.3132, "step": 5509 }, { "epoch": 1.7787239194479196, "grad_norm": 0.47989817099050386, "learning_rate": 5.827003340858489e-06, "loss": 0.2742, "step": 5510 }, { "epoch": 1.7790467734775415, "grad_norm": 0.45426026705671946, "learning_rate": 5.824401105916472e-06, "loss": 0.317, "step": 5511 }, { "epoch": 1.7793696275071633, "grad_norm": 0.40054039787289775, "learning_rate": 5.82179908327053e-06, "loss": 0.2916, "step": 5512 }, { "epoch": 1.7796924815367852, "grad_norm": 0.3503887506574728, "learning_rate": 5.8191972732503294e-06, "loss": 0.3361, "step": 5513 }, { "epoch": 1.780015335566407, "grad_norm": 0.46065645063481975, "learning_rate": 5.816595676185523e-06, "loss": 0.3, "step": 5514 }, { "epoch": 1.7803381895960289, "grad_norm": 0.4504508973455537, "learning_rate": 5.8139942924057235e-06, "loss": 0.2834, "step": 5515 }, { "epoch": 1.780661043625651, "grad_norm": 0.3828489656024216, "learning_rate": 5.811393122240528e-06, "loss": 0.2922, "step": 5516 }, { "epoch": 1.7809838976552728, "grad_norm": 0.41794436328208334, "learning_rate": 5.808792166019496e-06, "loss": 0.3359, "step": 5517 }, { "epoch": 1.7813067516848946, "grad_norm": 0.467818679820564, "learning_rate": 5.80619142407217e-06, "loss": 0.2658, "step": 5518 }, { "epoch": 1.7816296057145165, "grad_norm": 0.5180475242584746, "learning_rate": 5.803590896728056e-06, "loss": 0.3229, "step": 5519 }, { "epoch": 1.7819524597441383, "grad_norm": 0.4450942594545749, "learning_rate": 5.800990584316642e-06, "loss": 0.297, "step": 5520 }, { "epoch": 1.7822753137737601, "grad_norm": 0.39655725729960906, "learning_rate": 5.798390487167379e-06, "loss": 0.3606, "step": 5521 }, { "epoch": 1.782598167803382, "grad_norm": 0.4152193205430396, "learning_rate": 5.7957906056096996e-06, "loss": 0.3119, "step": 5522 }, { "epoch": 1.7829210218330038, "grad_norm": 0.3595455953484686, "learning_rate": 5.793190939973001e-06, "loss": 0.2603, "step": 5523 }, { "epoch": 1.7832438758626257, "grad_norm": 0.38803347089626977, "learning_rate": 5.790591490586662e-06, "loss": 0.2949, "step": 5524 }, { "epoch": 1.7835667298922475, "grad_norm": 0.4123387841923781, "learning_rate": 5.7879922577800255e-06, "loss": 0.3406, "step": 5525 }, { "epoch": 1.7838895839218694, "grad_norm": 0.42963306031933046, "learning_rate": 5.785393241882412e-06, "loss": 0.3274, "step": 5526 }, { "epoch": 1.7842124379514912, "grad_norm": 0.4110828522068207, "learning_rate": 5.782794443223117e-06, "loss": 0.3126, "step": 5527 }, { "epoch": 1.784535291981113, "grad_norm": 0.4397594695372374, "learning_rate": 5.780195862131398e-06, "loss": 0.2941, "step": 5528 }, { "epoch": 1.784858146010735, "grad_norm": 0.41310932259767313, "learning_rate": 5.777597498936499e-06, "loss": 0.3136, "step": 5529 }, { "epoch": 1.7851810000403567, "grad_norm": 0.4794179038636059, "learning_rate": 5.77499935396762e-06, "loss": 0.2697, "step": 5530 }, { "epoch": 1.7855038540699786, "grad_norm": 0.4244119394521111, "learning_rate": 5.77240142755395e-06, "loss": 0.2479, "step": 5531 }, { "epoch": 1.7858267080996004, "grad_norm": 0.33878941932806667, "learning_rate": 5.76980372002464e-06, "loss": 0.248, "step": 5532 }, { "epoch": 1.7861495621292223, "grad_norm": 0.48994957882413503, "learning_rate": 5.7672062317088165e-06, "loss": 0.2673, "step": 5533 }, { "epoch": 1.7864724161588441, "grad_norm": 0.4819520162194827, "learning_rate": 5.764608962935576e-06, "loss": 0.3812, "step": 5534 }, { "epoch": 1.786795270188466, "grad_norm": 0.3738530892483669, "learning_rate": 5.762011914033992e-06, "loss": 0.3071, "step": 5535 }, { "epoch": 1.7871181242180878, "grad_norm": 0.3888820150195311, "learning_rate": 5.759415085333102e-06, "loss": 0.2492, "step": 5536 }, { "epoch": 1.7874409782477096, "grad_norm": 0.37632650416819297, "learning_rate": 5.756818477161928e-06, "loss": 0.3097, "step": 5537 }, { "epoch": 1.7877638322773315, "grad_norm": 0.3990019198702324, "learning_rate": 5.7542220898494486e-06, "loss": 0.2933, "step": 5538 }, { "epoch": 1.7880866863069533, "grad_norm": 0.34471317886344705, "learning_rate": 5.7516259237246285e-06, "loss": 0.3657, "step": 5539 }, { "epoch": 1.7884095403365752, "grad_norm": 0.523181174975862, "learning_rate": 5.749029979116394e-06, "loss": 0.2755, "step": 5540 }, { "epoch": 1.788732394366197, "grad_norm": 0.3384901310681172, "learning_rate": 5.7464342563536535e-06, "loss": 0.3306, "step": 5541 }, { "epoch": 1.789055248395819, "grad_norm": 0.4269076454872017, "learning_rate": 5.743838755765274e-06, "loss": 0.2955, "step": 5542 }, { "epoch": 1.789378102425441, "grad_norm": 0.41062044572081907, "learning_rate": 5.741243477680107e-06, "loss": 0.2722, "step": 5543 }, { "epoch": 1.7897009564550628, "grad_norm": 0.44835777555299305, "learning_rate": 5.738648422426972e-06, "loss": 0.2823, "step": 5544 }, { "epoch": 1.7900238104846846, "grad_norm": 0.45973398671887183, "learning_rate": 5.736053590334653e-06, "loss": 0.2795, "step": 5545 }, { "epoch": 1.7903466645143065, "grad_norm": 0.41250878480100733, "learning_rate": 5.733458981731917e-06, "loss": 0.3132, "step": 5546 }, { "epoch": 1.7906695185439283, "grad_norm": 0.3619478407875433, "learning_rate": 5.730864596947495e-06, "loss": 0.2901, "step": 5547 }, { "epoch": 1.7909923725735502, "grad_norm": 0.4198749073651092, "learning_rate": 5.728270436310091e-06, "loss": 0.3467, "step": 5548 }, { "epoch": 1.791315226603172, "grad_norm": 0.37787616919892403, "learning_rate": 5.725676500148383e-06, "loss": 0.2923, "step": 5549 }, { "epoch": 1.791638080632794, "grad_norm": 0.4209530003314208, "learning_rate": 5.723082788791022e-06, "loss": 0.2838, "step": 5550 }, { "epoch": 1.791960934662416, "grad_norm": 0.3877085960712498, "learning_rate": 5.720489302566622e-06, "loss": 0.3057, "step": 5551 }, { "epoch": 1.7922837886920377, "grad_norm": 0.455861560269861, "learning_rate": 5.7178960418037795e-06, "loss": 0.3243, "step": 5552 }, { "epoch": 1.7926066427216596, "grad_norm": 0.3630386074613222, "learning_rate": 5.7153030068310505e-06, "loss": 0.2904, "step": 5553 }, { "epoch": 1.7929294967512814, "grad_norm": 0.3745299550204321, "learning_rate": 5.712710197976975e-06, "loss": 0.2719, "step": 5554 }, { "epoch": 1.7932523507809033, "grad_norm": 0.4150552908292056, "learning_rate": 5.710117615570056e-06, "loss": 0.3279, "step": 5555 }, { "epoch": 1.7935752048105251, "grad_norm": 0.34119758763587554, "learning_rate": 5.707525259938771e-06, "loss": 0.2539, "step": 5556 }, { "epoch": 1.793898058840147, "grad_norm": 0.44667818895275446, "learning_rate": 5.704933131411565e-06, "loss": 0.321, "step": 5557 }, { "epoch": 1.7942209128697688, "grad_norm": 0.3811180216405595, "learning_rate": 5.702341230316863e-06, "loss": 0.264, "step": 5558 }, { "epoch": 1.7945437668993907, "grad_norm": 0.45735532441929494, "learning_rate": 5.699749556983048e-06, "loss": 0.314, "step": 5559 }, { "epoch": 1.7948666209290125, "grad_norm": 0.4607025040275053, "learning_rate": 5.697158111738489e-06, "loss": 0.3251, "step": 5560 }, { "epoch": 1.7951894749586343, "grad_norm": 0.38283173081723987, "learning_rate": 5.69456689491151e-06, "loss": 0.3303, "step": 5561 }, { "epoch": 1.7955123289882562, "grad_norm": 0.310099536424096, "learning_rate": 5.69197590683042e-06, "loss": 0.3104, "step": 5562 }, { "epoch": 1.795835183017878, "grad_norm": 0.3501304672059674, "learning_rate": 5.689385147823494e-06, "loss": 0.3255, "step": 5563 }, { "epoch": 1.7961580370474999, "grad_norm": 0.35805093866482696, "learning_rate": 5.686794618218976e-06, "loss": 0.3461, "step": 5564 }, { "epoch": 1.7964808910771217, "grad_norm": 0.3956495961268003, "learning_rate": 5.684204318345084e-06, "loss": 0.2892, "step": 5565 }, { "epoch": 1.7968037451067436, "grad_norm": 0.37452305209069026, "learning_rate": 5.681614248530002e-06, "loss": 0.3059, "step": 5566 }, { "epoch": 1.7971265991363654, "grad_norm": 0.365568451824504, "learning_rate": 5.679024409101895e-06, "loss": 0.3554, "step": 5567 }, { "epoch": 1.7974494531659873, "grad_norm": 0.435506295567691, "learning_rate": 5.6764348003888834e-06, "loss": 0.3131, "step": 5568 }, { "epoch": 1.797772307195609, "grad_norm": 0.3455837414694504, "learning_rate": 5.673845422719075e-06, "loss": 0.2914, "step": 5569 }, { "epoch": 1.798095161225231, "grad_norm": 0.345842340676848, "learning_rate": 5.671256276420534e-06, "loss": 0.2968, "step": 5570 }, { "epoch": 1.7984180152548528, "grad_norm": 0.31145127714455223, "learning_rate": 5.668667361821305e-06, "loss": 0.2502, "step": 5571 }, { "epoch": 1.7987408692844746, "grad_norm": 0.3803227653970733, "learning_rate": 5.666078679249398e-06, "loss": 0.2928, "step": 5572 }, { "epoch": 1.7990637233140965, "grad_norm": 0.38112892335004067, "learning_rate": 5.663490229032801e-06, "loss": 0.379, "step": 5573 }, { "epoch": 1.7993865773437183, "grad_norm": 0.3351565999166047, "learning_rate": 5.660902011499459e-06, "loss": 0.2711, "step": 5574 }, { "epoch": 1.7997094313733402, "grad_norm": 0.5542972957428504, "learning_rate": 5.658314026977302e-06, "loss": 0.2804, "step": 5575 }, { "epoch": 1.8000322854029622, "grad_norm": 0.38443207980679395, "learning_rate": 5.65572627579422e-06, "loss": 0.284, "step": 5576 }, { "epoch": 1.800355139432584, "grad_norm": 0.48428056048706153, "learning_rate": 5.6531387582780796e-06, "loss": 0.3409, "step": 5577 }, { "epoch": 1.800677993462206, "grad_norm": 0.38878212785844024, "learning_rate": 5.650551474756716e-06, "loss": 0.3008, "step": 5578 }, { "epoch": 1.8010008474918278, "grad_norm": 0.3720505074616607, "learning_rate": 5.647964425557934e-06, "loss": 0.3006, "step": 5579 }, { "epoch": 1.8013237015214496, "grad_norm": 0.4292511895350967, "learning_rate": 5.645377611009506e-06, "loss": 0.2859, "step": 5580 }, { "epoch": 1.8016465555510714, "grad_norm": 0.3662064520278397, "learning_rate": 5.642791031439181e-06, "loss": 0.2701, "step": 5581 }, { "epoch": 1.8019694095806933, "grad_norm": 0.35264018698894445, "learning_rate": 5.640204687174679e-06, "loss": 0.2744, "step": 5582 }, { "epoch": 1.8022922636103151, "grad_norm": 0.40409726155346726, "learning_rate": 5.637618578543678e-06, "loss": 0.3187, "step": 5583 }, { "epoch": 1.8026151176399372, "grad_norm": 0.38221767707947707, "learning_rate": 5.635032705873842e-06, "loss": 0.3518, "step": 5584 }, { "epoch": 1.802937971669559, "grad_norm": 0.42919316894437803, "learning_rate": 5.63244706949279e-06, "loss": 0.3244, "step": 5585 }, { "epoch": 1.803260825699181, "grad_norm": 0.35588335742564225, "learning_rate": 5.6298616697281255e-06, "loss": 0.322, "step": 5586 }, { "epoch": 1.8035836797288027, "grad_norm": 0.43894011895401763, "learning_rate": 5.627276506907412e-06, "loss": 0.3289, "step": 5587 }, { "epoch": 1.8039065337584246, "grad_norm": 0.3967022716876395, "learning_rate": 5.624691581358184e-06, "loss": 0.3189, "step": 5588 }, { "epoch": 1.8042293877880464, "grad_norm": 0.4934643743963268, "learning_rate": 5.622106893407951e-06, "loss": 0.2975, "step": 5589 }, { "epoch": 1.8045522418176683, "grad_norm": 0.385726211751364, "learning_rate": 5.619522443384191e-06, "loss": 0.3273, "step": 5590 }, { "epoch": 1.8048750958472901, "grad_norm": 0.35442040766498184, "learning_rate": 5.616938231614344e-06, "loss": 0.2856, "step": 5591 }, { "epoch": 1.805197949876912, "grad_norm": 0.3535312684848603, "learning_rate": 5.614354258425834e-06, "loss": 0.3162, "step": 5592 }, { "epoch": 1.8055208039065338, "grad_norm": 0.4731953110779823, "learning_rate": 5.611770524146039e-06, "loss": 0.2316, "step": 5593 }, { "epoch": 1.8058436579361556, "grad_norm": 0.38912965349890843, "learning_rate": 5.609187029102321e-06, "loss": 0.3604, "step": 5594 }, { "epoch": 1.8061665119657775, "grad_norm": 0.36872044717613317, "learning_rate": 5.606603773621999e-06, "loss": 0.292, "step": 5595 }, { "epoch": 1.8064893659953993, "grad_norm": 0.3608073298275265, "learning_rate": 5.604020758032377e-06, "loss": 0.3178, "step": 5596 }, { "epoch": 1.8068122200250212, "grad_norm": 0.38113023495262016, "learning_rate": 5.60143798266071e-06, "loss": 0.3141, "step": 5597 }, { "epoch": 1.807135074054643, "grad_norm": 0.326284400309786, "learning_rate": 5.598855447834239e-06, "loss": 0.2661, "step": 5598 }, { "epoch": 1.8074579280842649, "grad_norm": 0.3313054142339996, "learning_rate": 5.596273153880162e-06, "loss": 0.3151, "step": 5599 }, { "epoch": 1.8077807821138867, "grad_norm": 0.3341720520856866, "learning_rate": 5.593691101125654e-06, "loss": 0.2735, "step": 5600 }, { "epoch": 1.8081036361435086, "grad_norm": 0.47887140209716866, "learning_rate": 5.591109289897862e-06, "loss": 0.3248, "step": 5601 }, { "epoch": 1.8084264901731304, "grad_norm": 0.3679436456905182, "learning_rate": 5.58852772052389e-06, "loss": 0.3307, "step": 5602 }, { "epoch": 1.8087493442027522, "grad_norm": 0.3879447412015418, "learning_rate": 5.5859463933308276e-06, "loss": 0.3384, "step": 5603 }, { "epoch": 1.809072198232374, "grad_norm": 0.359036574072371, "learning_rate": 5.583365308645718e-06, "loss": 0.3412, "step": 5604 }, { "epoch": 1.809395052261996, "grad_norm": 0.4244751623253591, "learning_rate": 5.580784466795588e-06, "loss": 0.273, "step": 5605 }, { "epoch": 1.8097179062916178, "grad_norm": 0.3582971963490569, "learning_rate": 5.578203868107419e-06, "loss": 0.2721, "step": 5606 }, { "epoch": 1.8100407603212396, "grad_norm": 0.4125364192370815, "learning_rate": 5.575623512908179e-06, "loss": 0.3105, "step": 5607 }, { "epoch": 1.8103636143508615, "grad_norm": 0.4145913556002381, "learning_rate": 5.573043401524785e-06, "loss": 0.2445, "step": 5608 }, { "epoch": 1.8106864683804835, "grad_norm": 0.6792294661105339, "learning_rate": 5.57046353428414e-06, "loss": 0.3278, "step": 5609 }, { "epoch": 1.8110093224101054, "grad_norm": 0.39519971109366336, "learning_rate": 5.567883911513108e-06, "loss": 0.2929, "step": 5610 }, { "epoch": 1.8113321764397272, "grad_norm": 0.39109629059630924, "learning_rate": 5.565304533538525e-06, "loss": 0.2778, "step": 5611 }, { "epoch": 1.811655030469349, "grad_norm": 0.3781149713404633, "learning_rate": 5.56272540068719e-06, "loss": 0.2974, "step": 5612 }, { "epoch": 1.811977884498971, "grad_norm": 0.4828182407183595, "learning_rate": 5.560146513285884e-06, "loss": 0.3061, "step": 5613 }, { "epoch": 1.8123007385285927, "grad_norm": 0.36728537214809115, "learning_rate": 5.557567871661339e-06, "loss": 0.3641, "step": 5614 }, { "epoch": 1.8126235925582146, "grad_norm": 0.3406396726326474, "learning_rate": 5.5549894761402734e-06, "loss": 0.2793, "step": 5615 }, { "epoch": 1.8129464465878364, "grad_norm": 0.7590009496286744, "learning_rate": 5.55241132704936e-06, "loss": 0.3256, "step": 5616 }, { "epoch": 1.8132693006174583, "grad_norm": 0.35512866780850366, "learning_rate": 5.549833424715248e-06, "loss": 0.232, "step": 5617 }, { "epoch": 1.8135921546470803, "grad_norm": 0.48552285402660755, "learning_rate": 5.547255769464558e-06, "loss": 0.3253, "step": 5618 }, { "epoch": 1.8139150086767022, "grad_norm": 0.4321803973463674, "learning_rate": 5.544678361623871e-06, "loss": 0.2943, "step": 5619 }, { "epoch": 1.814237862706324, "grad_norm": 0.4159439208524907, "learning_rate": 5.542101201519747e-06, "loss": 0.3326, "step": 5620 }, { "epoch": 1.8145607167359459, "grad_norm": 0.37164633910130973, "learning_rate": 5.539524289478701e-06, "loss": 0.3269, "step": 5621 }, { "epoch": 1.8148835707655677, "grad_norm": 0.4350000972275223, "learning_rate": 5.53694762582723e-06, "loss": 0.2992, "step": 5622 }, { "epoch": 1.8152064247951896, "grad_norm": 0.32014425036563005, "learning_rate": 5.53437121089179e-06, "loss": 0.2534, "step": 5623 }, { "epoch": 1.8155292788248114, "grad_norm": 0.35842495970844157, "learning_rate": 5.531795044998813e-06, "loss": 0.2983, "step": 5624 }, { "epoch": 1.8158521328544333, "grad_norm": 0.39657147705207896, "learning_rate": 5.529219128474689e-06, "loss": 0.3172, "step": 5625 }, { "epoch": 1.816174986884055, "grad_norm": 0.3721841473861226, "learning_rate": 5.526643461645791e-06, "loss": 0.2837, "step": 5626 }, { "epoch": 1.816497840913677, "grad_norm": 0.4621088014134099, "learning_rate": 5.5240680448384465e-06, "loss": 0.3535, "step": 5627 }, { "epoch": 1.8168206949432988, "grad_norm": 0.3871808573349131, "learning_rate": 5.521492878378964e-06, "loss": 0.3442, "step": 5628 }, { "epoch": 1.8171435489729206, "grad_norm": 0.4171550773266917, "learning_rate": 5.518917962593605e-06, "loss": 0.314, "step": 5629 }, { "epoch": 1.8174664030025425, "grad_norm": 0.43511550132995086, "learning_rate": 5.516343297808616e-06, "loss": 0.2993, "step": 5630 }, { "epoch": 1.8177892570321643, "grad_norm": 0.3959898294148138, "learning_rate": 5.5137688843501954e-06, "loss": 0.3861, "step": 5631 }, { "epoch": 1.8181121110617862, "grad_norm": 0.6135440269298746, "learning_rate": 5.5111947225445236e-06, "loss": 0.3298, "step": 5632 }, { "epoch": 1.818434965091408, "grad_norm": 0.3244936852921675, "learning_rate": 5.5086208127177425e-06, "loss": 0.2933, "step": 5633 }, { "epoch": 1.8187578191210299, "grad_norm": 0.5083531914819572, "learning_rate": 5.5060471551959614e-06, "loss": 0.2398, "step": 5634 }, { "epoch": 1.8190806731506517, "grad_norm": 0.35502960526231603, "learning_rate": 5.5034737503052594e-06, "loss": 0.2602, "step": 5635 }, { "epoch": 1.8194035271802735, "grad_norm": 0.36355015155485304, "learning_rate": 5.500900598371682e-06, "loss": 0.2957, "step": 5636 }, { "epoch": 1.8197263812098954, "grad_norm": 0.4072408273717202, "learning_rate": 5.4983276997212496e-06, "loss": 0.315, "step": 5637 }, { "epoch": 1.8200492352395172, "grad_norm": 0.40066606630840923, "learning_rate": 5.495755054679939e-06, "loss": 0.3472, "step": 5638 }, { "epoch": 1.820372089269139, "grad_norm": 0.4454767194785982, "learning_rate": 5.493182663573705e-06, "loss": 0.3592, "step": 5639 }, { "epoch": 1.820694943298761, "grad_norm": 0.37168261233510885, "learning_rate": 5.490610526728461e-06, "loss": 0.2536, "step": 5640 }, { "epoch": 1.8210177973283828, "grad_norm": 0.3372431751456083, "learning_rate": 5.488038644470099e-06, "loss": 0.2768, "step": 5641 }, { "epoch": 1.8213406513580046, "grad_norm": 0.4540601381024098, "learning_rate": 5.4854670171244685e-06, "loss": 0.2553, "step": 5642 }, { "epoch": 1.8216635053876267, "grad_norm": 0.34187200454727207, "learning_rate": 5.482895645017394e-06, "loss": 0.2696, "step": 5643 }, { "epoch": 1.8219863594172485, "grad_norm": 0.3649793633598135, "learning_rate": 5.480324528474661e-06, "loss": 0.2912, "step": 5644 }, { "epoch": 1.8223092134468704, "grad_norm": 0.41196551156481304, "learning_rate": 5.477753667822034e-06, "loss": 0.3005, "step": 5645 }, { "epoch": 1.8226320674764922, "grad_norm": 0.4727405717316549, "learning_rate": 5.475183063385227e-06, "loss": 0.3391, "step": 5646 }, { "epoch": 1.822954921506114, "grad_norm": 0.3020251614941979, "learning_rate": 5.472612715489942e-06, "loss": 0.3059, "step": 5647 }, { "epoch": 1.823277775535736, "grad_norm": 0.4064959431875453, "learning_rate": 5.470042624461831e-06, "loss": 0.2972, "step": 5648 }, { "epoch": 1.8236006295653577, "grad_norm": 0.35172154399591343, "learning_rate": 5.467472790626525e-06, "loss": 0.2902, "step": 5649 }, { "epoch": 1.8239234835949796, "grad_norm": 0.45998354442665074, "learning_rate": 5.464903214309616e-06, "loss": 0.2811, "step": 5650 }, { "epoch": 1.8242463376246014, "grad_norm": 0.44485703962474654, "learning_rate": 5.46233389583667e-06, "loss": 0.2662, "step": 5651 }, { "epoch": 1.8245691916542235, "grad_norm": 0.37679183000246286, "learning_rate": 5.45976483553321e-06, "loss": 0.319, "step": 5652 }, { "epoch": 1.8248920456838453, "grad_norm": 0.4187565220616703, "learning_rate": 5.45719603372474e-06, "loss": 0.2843, "step": 5653 }, { "epoch": 1.8252148997134672, "grad_norm": 0.4946757352153717, "learning_rate": 5.454627490736717e-06, "loss": 0.2589, "step": 5654 }, { "epoch": 1.825537753743089, "grad_norm": 0.4066788211935638, "learning_rate": 5.452059206894573e-06, "loss": 0.3561, "step": 5655 }, { "epoch": 1.8258606077727109, "grad_norm": 0.5656299426464615, "learning_rate": 5.44949118252371e-06, "loss": 0.2678, "step": 5656 }, { "epoch": 1.8261834618023327, "grad_norm": 0.4296308605244071, "learning_rate": 5.4469234179494875e-06, "loss": 0.3252, "step": 5657 }, { "epoch": 1.8265063158319546, "grad_norm": 0.4303277184970379, "learning_rate": 5.444355913497243e-06, "loss": 0.3656, "step": 5658 }, { "epoch": 1.8268291698615764, "grad_norm": 0.47771746083666156, "learning_rate": 5.441788669492272e-06, "loss": 0.3008, "step": 5659 }, { "epoch": 1.8271520238911982, "grad_norm": 0.44654663849908466, "learning_rate": 5.439221686259845e-06, "loss": 0.2655, "step": 5660 }, { "epoch": 1.82747487792082, "grad_norm": 0.3474336863120708, "learning_rate": 5.4366549641251894e-06, "loss": 0.295, "step": 5661 }, { "epoch": 1.827797731950442, "grad_norm": 0.4156244338229464, "learning_rate": 5.434088503413511e-06, "loss": 0.3189, "step": 5662 }, { "epoch": 1.8281205859800638, "grad_norm": 0.7059189944456777, "learning_rate": 5.431522304449972e-06, "loss": 0.3146, "step": 5663 }, { "epoch": 1.8284434400096856, "grad_norm": 0.38197587511678727, "learning_rate": 5.428956367559709e-06, "loss": 0.2887, "step": 5664 }, { "epoch": 1.8287662940393075, "grad_norm": 0.42536953109808073, "learning_rate": 5.426390693067821e-06, "loss": 0.3398, "step": 5665 }, { "epoch": 1.8290891480689293, "grad_norm": 0.40868001178580565, "learning_rate": 5.4238252812993814e-06, "loss": 0.282, "step": 5666 }, { "epoch": 1.8294120020985511, "grad_norm": 0.4283831285552794, "learning_rate": 5.4212601325794145e-06, "loss": 0.3016, "step": 5667 }, { "epoch": 1.829734856128173, "grad_norm": 0.33099039350332804, "learning_rate": 5.418695247232931e-06, "loss": 0.3156, "step": 5668 }, { "epoch": 1.8300577101577948, "grad_norm": 0.3399596297076998, "learning_rate": 5.416130625584889e-06, "loss": 0.2727, "step": 5669 }, { "epoch": 1.8303805641874167, "grad_norm": 0.4001845884649666, "learning_rate": 5.413566267960231e-06, "loss": 0.3344, "step": 5670 }, { "epoch": 1.8307034182170385, "grad_norm": 0.4098714677242265, "learning_rate": 5.41100217468385e-06, "loss": 0.3212, "step": 5671 }, { "epoch": 1.8310262722466604, "grad_norm": 0.39325071003834916, "learning_rate": 5.408438346080618e-06, "loss": 0.2769, "step": 5672 }, { "epoch": 1.8313491262762822, "grad_norm": 0.3743204087622822, "learning_rate": 5.405874782475367e-06, "loss": 0.2601, "step": 5673 }, { "epoch": 1.831671980305904, "grad_norm": 0.46905074444509437, "learning_rate": 5.403311484192897e-06, "loss": 0.2895, "step": 5674 }, { "epoch": 1.831994834335526, "grad_norm": 0.4363912475483492, "learning_rate": 5.4007484515579765e-06, "loss": 0.3137, "step": 5675 }, { "epoch": 1.8323176883651477, "grad_norm": 0.40905996462976574, "learning_rate": 5.3981856848953346e-06, "loss": 0.2956, "step": 5676 }, { "epoch": 1.8326405423947698, "grad_norm": 0.43521229725728106, "learning_rate": 5.395623184529676e-06, "loss": 0.3644, "step": 5677 }, { "epoch": 1.8329633964243917, "grad_norm": 0.42946096663549643, "learning_rate": 5.393060950785657e-06, "loss": 0.3504, "step": 5678 }, { "epoch": 1.8332862504540135, "grad_norm": 0.40730149473930566, "learning_rate": 5.390498983987917e-06, "loss": 0.3062, "step": 5679 }, { "epoch": 1.8336091044836353, "grad_norm": 0.38861618834933837, "learning_rate": 5.387937284461049e-06, "loss": 0.3365, "step": 5680 }, { "epoch": 1.8339319585132572, "grad_norm": 1.026892573786342, "learning_rate": 5.385375852529619e-06, "loss": 0.2549, "step": 5681 }, { "epoch": 1.834254812542879, "grad_norm": 0.4454459398431162, "learning_rate": 5.382814688518156e-06, "loss": 0.3106, "step": 5682 }, { "epoch": 1.8345776665725009, "grad_norm": 0.3828685657422503, "learning_rate": 5.380253792751158e-06, "loss": 0.3206, "step": 5683 }, { "epoch": 1.8349005206021227, "grad_norm": 0.3199655859144824, "learning_rate": 5.3776931655530844e-06, "loss": 0.3121, "step": 5684 }, { "epoch": 1.8352233746317448, "grad_norm": 0.4434187070615745, "learning_rate": 5.375132807248367e-06, "loss": 0.3594, "step": 5685 }, { "epoch": 1.8355462286613666, "grad_norm": 0.4664939050771899, "learning_rate": 5.372572718161392e-06, "loss": 0.2664, "step": 5686 }, { "epoch": 1.8358690826909885, "grad_norm": 0.36847778383067004, "learning_rate": 5.370012898616527e-06, "loss": 0.2362, "step": 5687 }, { "epoch": 1.8361919367206103, "grad_norm": 0.3781491031203242, "learning_rate": 5.367453348938094e-06, "loss": 0.2758, "step": 5688 }, { "epoch": 1.8365147907502322, "grad_norm": 0.3027852829689635, "learning_rate": 5.364894069450385e-06, "loss": 0.2522, "step": 5689 }, { "epoch": 1.836837644779854, "grad_norm": 0.6500365847255462, "learning_rate": 5.3623350604776554e-06, "loss": 0.2875, "step": 5690 }, { "epoch": 1.8371604988094759, "grad_norm": 0.418499732370877, "learning_rate": 5.35977632234413e-06, "loss": 0.3133, "step": 5691 }, { "epoch": 1.8374833528390977, "grad_norm": 0.318436913066264, "learning_rate": 5.357217855374e-06, "loss": 0.2571, "step": 5692 }, { "epoch": 1.8378062068687195, "grad_norm": 0.33733295577057243, "learning_rate": 5.354659659891414e-06, "loss": 0.2968, "step": 5693 }, { "epoch": 1.8381290608983414, "grad_norm": 0.39312277258618483, "learning_rate": 5.352101736220499e-06, "loss": 0.371, "step": 5694 }, { "epoch": 1.8384519149279632, "grad_norm": 0.581998942186357, "learning_rate": 5.349544084685333e-06, "loss": 0.3159, "step": 5695 }, { "epoch": 1.838774768957585, "grad_norm": 0.42972507280985656, "learning_rate": 5.346986705609971e-06, "loss": 0.3275, "step": 5696 }, { "epoch": 1.839097622987207, "grad_norm": 0.3947606049877151, "learning_rate": 5.344429599318426e-06, "loss": 0.2536, "step": 5697 }, { "epoch": 1.8394204770168288, "grad_norm": 0.3420389832699977, "learning_rate": 5.341872766134686e-06, "loss": 0.2893, "step": 5698 }, { "epoch": 1.8397433310464506, "grad_norm": 0.3725806000474088, "learning_rate": 5.3393162063826924e-06, "loss": 0.3262, "step": 5699 }, { "epoch": 1.8400661850760724, "grad_norm": 0.3308016585760281, "learning_rate": 5.336759920386363e-06, "loss": 0.2675, "step": 5700 }, { "epoch": 1.8403890391056943, "grad_norm": 0.45893549851858617, "learning_rate": 5.334203908469569e-06, "loss": 0.3792, "step": 5701 }, { "epoch": 1.8407118931353161, "grad_norm": 0.45202460171681225, "learning_rate": 5.33164817095616e-06, "loss": 0.3396, "step": 5702 }, { "epoch": 1.841034747164938, "grad_norm": 0.36641948444350114, "learning_rate": 5.32909270816994e-06, "loss": 0.2851, "step": 5703 }, { "epoch": 1.8413576011945598, "grad_norm": 0.4371180859291097, "learning_rate": 5.326537520434686e-06, "loss": 0.3046, "step": 5704 }, { "epoch": 1.8416804552241817, "grad_norm": 0.44115370814953114, "learning_rate": 5.323982608074132e-06, "loss": 0.3679, "step": 5705 }, { "epoch": 1.8420033092538035, "grad_norm": 0.3822881382246787, "learning_rate": 5.3214279714119895e-06, "loss": 0.3191, "step": 5706 }, { "epoch": 1.8423261632834254, "grad_norm": 0.41436134272071695, "learning_rate": 5.318873610771919e-06, "loss": 0.301, "step": 5707 }, { "epoch": 1.8426490173130472, "grad_norm": 0.3487164764644751, "learning_rate": 5.3163195264775615e-06, "loss": 0.3152, "step": 5708 }, { "epoch": 1.842971871342669, "grad_norm": 0.4003784973758822, "learning_rate": 5.313765718852511e-06, "loss": 0.3549, "step": 5709 }, { "epoch": 1.8432947253722909, "grad_norm": 0.362331807035406, "learning_rate": 5.311212188220332e-06, "loss": 0.2678, "step": 5710 }, { "epoch": 1.843617579401913, "grad_norm": 0.39527150614422885, "learning_rate": 5.3086589349045575e-06, "loss": 0.2723, "step": 5711 }, { "epoch": 1.8439404334315348, "grad_norm": 0.3705715635164447, "learning_rate": 5.306105959228677e-06, "loss": 0.2944, "step": 5712 }, { "epoch": 1.8442632874611566, "grad_norm": 0.3415879527894206, "learning_rate": 5.30355326151615e-06, "loss": 0.2788, "step": 5713 }, { "epoch": 1.8445861414907785, "grad_norm": 0.39801769196212344, "learning_rate": 5.301000842090398e-06, "loss": 0.2865, "step": 5714 }, { "epoch": 1.8449089955204003, "grad_norm": 0.416737564029543, "learning_rate": 5.298448701274817e-06, "loss": 0.3225, "step": 5715 }, { "epoch": 1.8452318495500222, "grad_norm": 0.3488204112579676, "learning_rate": 5.2958968393927485e-06, "loss": 0.2325, "step": 5716 }, { "epoch": 1.845554703579644, "grad_norm": 0.39145952044372484, "learning_rate": 5.293345256767519e-06, "loss": 0.3049, "step": 5717 }, { "epoch": 1.8458775576092659, "grad_norm": 0.4265691134114498, "learning_rate": 5.290793953722404e-06, "loss": 0.317, "step": 5718 }, { "epoch": 1.846200411638888, "grad_norm": 0.40170448210808124, "learning_rate": 5.288242930580653e-06, "loss": 0.2781, "step": 5719 }, { "epoch": 1.8465232656685098, "grad_norm": 0.3679480622726426, "learning_rate": 5.285692187665476e-06, "loss": 0.3637, "step": 5720 }, { "epoch": 1.8468461196981316, "grad_norm": 0.39478468566673175, "learning_rate": 5.283141725300054e-06, "loss": 0.3094, "step": 5721 }, { "epoch": 1.8471689737277535, "grad_norm": 0.4588222300035253, "learning_rate": 5.280591543807517e-06, "loss": 0.2943, "step": 5722 }, { "epoch": 1.8474918277573753, "grad_norm": 0.40470960053486194, "learning_rate": 5.278041643510981e-06, "loss": 0.2771, "step": 5723 }, { "epoch": 1.8478146817869971, "grad_norm": 0.360176389065901, "learning_rate": 5.275492024733504e-06, "loss": 0.2802, "step": 5724 }, { "epoch": 1.848137535816619, "grad_norm": 0.4481021137206953, "learning_rate": 5.272942687798128e-06, "loss": 0.2732, "step": 5725 }, { "epoch": 1.8484603898462408, "grad_norm": 0.3980814678109174, "learning_rate": 5.2703936330278425e-06, "loss": 0.3114, "step": 5726 }, { "epoch": 1.8487832438758627, "grad_norm": 0.4338583163285777, "learning_rate": 5.267844860745616e-06, "loss": 0.3008, "step": 5727 }, { "epoch": 1.8491060979054845, "grad_norm": 0.33092280717386213, "learning_rate": 5.26529637127437e-06, "loss": 0.2388, "step": 5728 }, { "epoch": 1.8494289519351064, "grad_norm": 0.4025597549498649, "learning_rate": 5.262748164936996e-06, "loss": 0.3477, "step": 5729 }, { "epoch": 1.8497518059647282, "grad_norm": 0.40400603199483565, "learning_rate": 5.260200242056351e-06, "loss": 0.3322, "step": 5730 }, { "epoch": 1.85007465999435, "grad_norm": 0.3684915604144623, "learning_rate": 5.257652602955249e-06, "loss": 0.3247, "step": 5731 }, { "epoch": 1.850397514023972, "grad_norm": 0.5030587207302419, "learning_rate": 5.255105247956478e-06, "loss": 0.3043, "step": 5732 }, { "epoch": 1.8507203680535937, "grad_norm": 0.37555681190225854, "learning_rate": 5.2525581773827764e-06, "loss": 0.3148, "step": 5733 }, { "epoch": 1.8510432220832156, "grad_norm": 0.39099667839694424, "learning_rate": 5.250011391556862e-06, "loss": 0.2601, "step": 5734 }, { "epoch": 1.8513660761128374, "grad_norm": 0.44445579321011863, "learning_rate": 5.247464890801404e-06, "loss": 0.3324, "step": 5735 }, { "epoch": 1.8516889301424593, "grad_norm": 0.3900663813885393, "learning_rate": 5.244918675439042e-06, "loss": 0.3348, "step": 5736 }, { "epoch": 1.8520117841720811, "grad_norm": 0.396892974839875, "learning_rate": 5.242372745792379e-06, "loss": 0.3107, "step": 5737 }, { "epoch": 1.852334638201703, "grad_norm": 0.3615223190597369, "learning_rate": 5.239827102183982e-06, "loss": 0.3181, "step": 5738 }, { "epoch": 1.8526574922313248, "grad_norm": 0.4401262816111586, "learning_rate": 5.237281744936376e-06, "loss": 0.2697, "step": 5739 }, { "epoch": 1.8529803462609467, "grad_norm": 0.3976589929597033, "learning_rate": 5.234736674372059e-06, "loss": 0.3155, "step": 5740 }, { "epoch": 1.8533032002905685, "grad_norm": 0.35192264451354793, "learning_rate": 5.232191890813484e-06, "loss": 0.2825, "step": 5741 }, { "epoch": 1.8536260543201903, "grad_norm": 0.44510767184400224, "learning_rate": 5.229647394583074e-06, "loss": 0.3407, "step": 5742 }, { "epoch": 1.8539489083498122, "grad_norm": 0.40811599992800984, "learning_rate": 5.227103186003213e-06, "loss": 0.3347, "step": 5743 }, { "epoch": 1.8542717623794343, "grad_norm": 0.4213114745229587, "learning_rate": 5.2245592653962495e-06, "loss": 0.3495, "step": 5744 }, { "epoch": 1.854594616409056, "grad_norm": 0.3724082332350919, "learning_rate": 5.222015633084491e-06, "loss": 0.3634, "step": 5745 }, { "epoch": 1.854917470438678, "grad_norm": 0.41710815755310443, "learning_rate": 5.219472289390215e-06, "loss": 0.2795, "step": 5746 }, { "epoch": 1.8552403244682998, "grad_norm": 0.43384185076280696, "learning_rate": 5.216929234635663e-06, "loss": 0.273, "step": 5747 }, { "epoch": 1.8555631784979216, "grad_norm": 0.4047617921535596, "learning_rate": 5.214386469143029e-06, "loss": 0.3014, "step": 5748 }, { "epoch": 1.8558860325275435, "grad_norm": 0.6962540600351096, "learning_rate": 5.2118439932344854e-06, "loss": 0.3007, "step": 5749 }, { "epoch": 1.8562088865571653, "grad_norm": 0.3580514746118442, "learning_rate": 5.209301807232154e-06, "loss": 0.3094, "step": 5750 }, { "epoch": 1.8565317405867872, "grad_norm": 0.4007302012946608, "learning_rate": 5.20675991145813e-06, "loss": 0.3386, "step": 5751 }, { "epoch": 1.856854594616409, "grad_norm": 0.38899988584761314, "learning_rate": 5.204218306234467e-06, "loss": 0.3114, "step": 5752 }, { "epoch": 1.857177448646031, "grad_norm": 0.4478563903193188, "learning_rate": 5.201676991883185e-06, "loss": 0.3106, "step": 5753 }, { "epoch": 1.857500302675653, "grad_norm": 0.4512337487421453, "learning_rate": 5.19913596872626e-06, "loss": 0.3173, "step": 5754 }, { "epoch": 1.8578231567052748, "grad_norm": 0.40795615344297453, "learning_rate": 5.196595237085643e-06, "loss": 0.318, "step": 5755 }, { "epoch": 1.8581460107348966, "grad_norm": 0.3544369396372784, "learning_rate": 5.194054797283235e-06, "loss": 0.3219, "step": 5756 }, { "epoch": 1.8584688647645184, "grad_norm": 0.5136607867924216, "learning_rate": 5.191514649640912e-06, "loss": 0.3115, "step": 5757 }, { "epoch": 1.8587917187941403, "grad_norm": 0.3971347133939424, "learning_rate": 5.1889747944804985e-06, "loss": 0.2881, "step": 5758 }, { "epoch": 1.8591145728237621, "grad_norm": 0.3560391265123478, "learning_rate": 5.1864352321238e-06, "loss": 0.338, "step": 5759 }, { "epoch": 1.859437426853384, "grad_norm": 0.42627619386225146, "learning_rate": 5.183895962892568e-06, "loss": 0.3079, "step": 5760 }, { "epoch": 1.8597602808830058, "grad_norm": 0.3816496118170131, "learning_rate": 5.181356987108533e-06, "loss": 0.2813, "step": 5761 }, { "epoch": 1.8600831349126277, "grad_norm": 0.4532509698494958, "learning_rate": 5.178818305093371e-06, "loss": 0.3291, "step": 5762 }, { "epoch": 1.8604059889422495, "grad_norm": 0.4357920482525313, "learning_rate": 5.1762799171687355e-06, "loss": 0.2649, "step": 5763 }, { "epoch": 1.8607288429718714, "grad_norm": 0.40107643737629567, "learning_rate": 5.173741823656233e-06, "loss": 0.2741, "step": 5764 }, { "epoch": 1.8610516970014932, "grad_norm": 0.47440101625123066, "learning_rate": 5.171204024877436e-06, "loss": 0.2806, "step": 5765 }, { "epoch": 1.861374551031115, "grad_norm": 0.48140482630209397, "learning_rate": 5.168666521153886e-06, "loss": 0.3114, "step": 5766 }, { "epoch": 1.8616974050607369, "grad_norm": 0.37945623043166626, "learning_rate": 5.166129312807077e-06, "loss": 0.3279, "step": 5767 }, { "epoch": 1.8620202590903587, "grad_norm": 0.46146247180489247, "learning_rate": 5.163592400158468e-06, "loss": 0.2427, "step": 5768 }, { "epoch": 1.8623431131199806, "grad_norm": 0.36336514596966357, "learning_rate": 5.161055783529484e-06, "loss": 0.2866, "step": 5769 }, { "epoch": 1.8626659671496024, "grad_norm": 0.4354196320964906, "learning_rate": 5.158519463241515e-06, "loss": 0.2786, "step": 5770 }, { "epoch": 1.8629888211792243, "grad_norm": 0.36117965195253743, "learning_rate": 5.155983439615903e-06, "loss": 0.3037, "step": 5771 }, { "epoch": 1.863311675208846, "grad_norm": 0.39027840941290437, "learning_rate": 5.153447712973963e-06, "loss": 0.2926, "step": 5772 }, { "epoch": 1.863634529238468, "grad_norm": 0.42887396767947167, "learning_rate": 5.1509122836369636e-06, "loss": 0.2808, "step": 5773 }, { "epoch": 1.8639573832680898, "grad_norm": 0.4184584339163942, "learning_rate": 5.148377151926145e-06, "loss": 0.3135, "step": 5774 }, { "epoch": 1.8642802372977116, "grad_norm": 0.3521189990058923, "learning_rate": 5.145842318162701e-06, "loss": 0.3226, "step": 5775 }, { "epoch": 1.8646030913273335, "grad_norm": 0.39224624969954863, "learning_rate": 5.143307782667797e-06, "loss": 0.2768, "step": 5776 }, { "epoch": 1.8649259453569553, "grad_norm": 0.3881750405324514, "learning_rate": 5.140773545762547e-06, "loss": 0.2852, "step": 5777 }, { "epoch": 1.8652487993865774, "grad_norm": 0.36910024602130403, "learning_rate": 5.138239607768045e-06, "loss": 0.2641, "step": 5778 }, { "epoch": 1.8655716534161992, "grad_norm": 0.35224992403029093, "learning_rate": 5.135705969005328e-06, "loss": 0.2445, "step": 5779 }, { "epoch": 1.865894507445821, "grad_norm": 0.43133440592792494, "learning_rate": 5.133172629795413e-06, "loss": 0.2851, "step": 5780 }, { "epoch": 1.866217361475443, "grad_norm": 0.4401346943264821, "learning_rate": 5.130639590459262e-06, "loss": 0.304, "step": 5781 }, { "epoch": 1.8665402155050648, "grad_norm": 0.40849463448770623, "learning_rate": 5.128106851317815e-06, "loss": 0.3511, "step": 5782 }, { "epoch": 1.8668630695346866, "grad_norm": 0.43328382603216103, "learning_rate": 5.125574412691961e-06, "loss": 0.3027, "step": 5783 }, { "epoch": 1.8671859235643085, "grad_norm": 0.3978444307053763, "learning_rate": 5.123042274902559e-06, "loss": 0.2665, "step": 5784 }, { "epoch": 1.8675087775939303, "grad_norm": 0.4110974148274741, "learning_rate": 5.120510438270432e-06, "loss": 0.32, "step": 5785 }, { "epoch": 1.8678316316235521, "grad_norm": 0.4325530147508854, "learning_rate": 5.117978903116351e-06, "loss": 0.2407, "step": 5786 }, { "epoch": 1.8681544856531742, "grad_norm": 0.35457779611843976, "learning_rate": 5.115447669761068e-06, "loss": 0.3496, "step": 5787 }, { "epoch": 1.868477339682796, "grad_norm": 0.4356472548904754, "learning_rate": 5.112916738525277e-06, "loss": 0.3393, "step": 5788 }, { "epoch": 1.868800193712418, "grad_norm": 0.37928848950035904, "learning_rate": 5.110386109729649e-06, "loss": 0.3066, "step": 5789 }, { "epoch": 1.8691230477420397, "grad_norm": 0.3626413831465389, "learning_rate": 5.107855783694811e-06, "loss": 0.3122, "step": 5790 }, { "epoch": 1.8694459017716616, "grad_norm": 0.49509032902459305, "learning_rate": 5.1053257607413515e-06, "loss": 0.3283, "step": 5791 }, { "epoch": 1.8697687558012834, "grad_norm": 0.36304167576409596, "learning_rate": 5.1027960411898176e-06, "loss": 0.2998, "step": 5792 }, { "epoch": 1.8700916098309053, "grad_norm": 0.37339581137253514, "learning_rate": 5.100266625360728e-06, "loss": 0.2673, "step": 5793 }, { "epoch": 1.8704144638605271, "grad_norm": 0.3503359790900876, "learning_rate": 5.097737513574549e-06, "loss": 0.3104, "step": 5794 }, { "epoch": 1.870737317890149, "grad_norm": 0.42151046210176374, "learning_rate": 5.095208706151722e-06, "loss": 0.2674, "step": 5795 }, { "epoch": 1.8710601719197708, "grad_norm": 0.4171756972220318, "learning_rate": 5.092680203412637e-06, "loss": 0.3287, "step": 5796 }, { "epoch": 1.8713830259493927, "grad_norm": 0.34674528441174757, "learning_rate": 5.090152005677657e-06, "loss": 0.3334, "step": 5797 }, { "epoch": 1.8717058799790145, "grad_norm": 0.30146189207560387, "learning_rate": 5.087624113267098e-06, "loss": 0.2568, "step": 5798 }, { "epoch": 1.8720287340086363, "grad_norm": 0.3403969965511541, "learning_rate": 5.0850965265012454e-06, "loss": 0.2616, "step": 5799 }, { "epoch": 1.8723515880382582, "grad_norm": 0.3780475432448916, "learning_rate": 5.082569245700334e-06, "loss": 0.282, "step": 5800 }, { "epoch": 1.87267444206788, "grad_norm": 0.3396292630077181, "learning_rate": 5.080042271184573e-06, "loss": 0.3058, "step": 5801 }, { "epoch": 1.8729972960975019, "grad_norm": 0.3755932974368502, "learning_rate": 5.07751560327412e-06, "loss": 0.2633, "step": 5802 }, { "epoch": 1.8733201501271237, "grad_norm": 0.41469228268845126, "learning_rate": 5.0749892422891045e-06, "loss": 0.3029, "step": 5803 }, { "epoch": 1.8736430041567456, "grad_norm": 0.5487738726422055, "learning_rate": 5.072463188549617e-06, "loss": 0.2769, "step": 5804 }, { "epoch": 1.8739658581863674, "grad_norm": 0.365976129180861, "learning_rate": 5.069937442375696e-06, "loss": 0.2747, "step": 5805 }, { "epoch": 1.8742887122159892, "grad_norm": 0.4274968597422361, "learning_rate": 5.067412004087356e-06, "loss": 0.2982, "step": 5806 }, { "epoch": 1.874611566245611, "grad_norm": 0.3774995508924465, "learning_rate": 5.064886874004564e-06, "loss": 0.277, "step": 5807 }, { "epoch": 1.874934420275233, "grad_norm": 0.4353777407350452, "learning_rate": 5.0623620524472546e-06, "loss": 0.2878, "step": 5808 }, { "epoch": 1.8752572743048548, "grad_norm": 0.336305452892691, "learning_rate": 5.0598375397353145e-06, "loss": 0.3452, "step": 5809 }, { "epoch": 1.8755801283344766, "grad_norm": 0.31766942263027514, "learning_rate": 5.0573133361885994e-06, "loss": 0.3603, "step": 5810 }, { "epoch": 1.8759029823640985, "grad_norm": 0.38797630810139827, "learning_rate": 5.054789442126919e-06, "loss": 0.2731, "step": 5811 }, { "epoch": 1.8762258363937205, "grad_norm": 0.40139281949175576, "learning_rate": 5.052265857870051e-06, "loss": 0.3378, "step": 5812 }, { "epoch": 1.8765486904233424, "grad_norm": 0.35484958642163533, "learning_rate": 5.049742583737726e-06, "loss": 0.3676, "step": 5813 }, { "epoch": 1.8768715444529642, "grad_norm": 0.37023037775275847, "learning_rate": 5.047219620049641e-06, "loss": 0.3689, "step": 5814 }, { "epoch": 1.877194398482586, "grad_norm": 0.4386673768693593, "learning_rate": 5.044696967125454e-06, "loss": 0.2945, "step": 5815 }, { "epoch": 1.877517252512208, "grad_norm": 0.44215329690140565, "learning_rate": 5.042174625284781e-06, "loss": 0.3314, "step": 5816 }, { "epoch": 1.8778401065418298, "grad_norm": 0.34613983057952025, "learning_rate": 5.039652594847196e-06, "loss": 0.2488, "step": 5817 }, { "epoch": 1.8781629605714516, "grad_norm": 0.44562716295390564, "learning_rate": 5.0371308761322415e-06, "loss": 0.3242, "step": 5818 }, { "epoch": 1.8784858146010734, "grad_norm": 0.38713215567488196, "learning_rate": 5.034609469459412e-06, "loss": 0.2618, "step": 5819 }, { "epoch": 1.8788086686306955, "grad_norm": 0.4887100313986826, "learning_rate": 5.032088375148167e-06, "loss": 0.3132, "step": 5820 }, { "epoch": 1.8791315226603174, "grad_norm": 0.404323648395831, "learning_rate": 5.029567593517927e-06, "loss": 0.3146, "step": 5821 }, { "epoch": 1.8794543766899392, "grad_norm": 0.40083722871225747, "learning_rate": 5.0270471248880714e-06, "loss": 0.275, "step": 5822 }, { "epoch": 1.879777230719561, "grad_norm": 0.46074390144621363, "learning_rate": 5.024526969577939e-06, "loss": 0.3541, "step": 5823 }, { "epoch": 1.8801000847491829, "grad_norm": 0.3850412834721091, "learning_rate": 5.02200712790683e-06, "loss": 0.2724, "step": 5824 }, { "epoch": 1.8804229387788047, "grad_norm": 0.38786545496589414, "learning_rate": 5.0194876001940085e-06, "loss": 0.3505, "step": 5825 }, { "epoch": 1.8807457928084266, "grad_norm": 0.4418996354712008, "learning_rate": 5.016968386758688e-06, "loss": 0.3141, "step": 5826 }, { "epoch": 1.8810686468380484, "grad_norm": 0.3918854422688409, "learning_rate": 5.014449487920058e-06, "loss": 0.2301, "step": 5827 }, { "epoch": 1.8813915008676703, "grad_norm": 0.4099669618047917, "learning_rate": 5.011930903997252e-06, "loss": 0.3264, "step": 5828 }, { "epoch": 1.881714354897292, "grad_norm": 0.3601195340698508, "learning_rate": 5.009412635309375e-06, "loss": 0.3066, "step": 5829 }, { "epoch": 1.882037208926914, "grad_norm": 0.4740479482929272, "learning_rate": 5.006894682175485e-06, "loss": 0.2715, "step": 5830 }, { "epoch": 1.8823600629565358, "grad_norm": 0.4246570552035093, "learning_rate": 5.00437704491461e-06, "loss": 0.27, "step": 5831 }, { "epoch": 1.8826829169861576, "grad_norm": 0.39315648388432334, "learning_rate": 5.001859723845724e-06, "loss": 0.3227, "step": 5832 }, { "epoch": 1.8830057710157795, "grad_norm": 0.4066310649200941, "learning_rate": 4.999342719287774e-06, "loss": 0.2623, "step": 5833 }, { "epoch": 1.8833286250454013, "grad_norm": 0.39696458742574403, "learning_rate": 4.996826031559653e-06, "loss": 0.3378, "step": 5834 }, { "epoch": 1.8836514790750232, "grad_norm": 0.34706142800862677, "learning_rate": 4.994309660980231e-06, "loss": 0.274, "step": 5835 }, { "epoch": 1.883974333104645, "grad_norm": 0.33912867528100754, "learning_rate": 4.991793607868321e-06, "loss": 0.3329, "step": 5836 }, { "epoch": 1.8842971871342669, "grad_norm": 0.48037711825981994, "learning_rate": 4.989277872542709e-06, "loss": 0.2423, "step": 5837 }, { "epoch": 1.8846200411638887, "grad_norm": 0.48095959306537833, "learning_rate": 4.986762455322131e-06, "loss": 0.2817, "step": 5838 }, { "epoch": 1.8849428951935105, "grad_norm": 0.387758026401593, "learning_rate": 4.984247356525289e-06, "loss": 0.3002, "step": 5839 }, { "epoch": 1.8852657492231324, "grad_norm": 0.40529457687564385, "learning_rate": 4.981732576470846e-06, "loss": 0.3529, "step": 5840 }, { "epoch": 1.8855886032527542, "grad_norm": 0.4336292702108475, "learning_rate": 4.979218115477414e-06, "loss": 0.3488, "step": 5841 }, { "epoch": 1.885911457282376, "grad_norm": 0.3594661010730982, "learning_rate": 4.97670397386358e-06, "loss": 0.291, "step": 5842 }, { "epoch": 1.886234311311998, "grad_norm": 0.391374602380449, "learning_rate": 4.974190151947872e-06, "loss": 0.3072, "step": 5843 }, { "epoch": 1.8865571653416198, "grad_norm": 0.3367883742891917, "learning_rate": 4.9716766500487975e-06, "loss": 0.2932, "step": 5844 }, { "epoch": 1.8868800193712416, "grad_norm": 0.40644010298754296, "learning_rate": 4.9691634684848085e-06, "loss": 0.2852, "step": 5845 }, { "epoch": 1.8872028734008637, "grad_norm": 0.4174887224204152, "learning_rate": 4.9666506075743225e-06, "loss": 0.2605, "step": 5846 }, { "epoch": 1.8875257274304855, "grad_norm": 0.40018496619134547, "learning_rate": 4.964138067635716e-06, "loss": 0.3216, "step": 5847 }, { "epoch": 1.8878485814601074, "grad_norm": 0.3687705439263446, "learning_rate": 4.961625848987326e-06, "loss": 0.347, "step": 5848 }, { "epoch": 1.8881714354897292, "grad_norm": 0.33088145794720875, "learning_rate": 4.959113951947443e-06, "loss": 0.3034, "step": 5849 }, { "epoch": 1.888494289519351, "grad_norm": 0.40915656398986683, "learning_rate": 4.956602376834325e-06, "loss": 0.3065, "step": 5850 }, { "epoch": 1.888817143548973, "grad_norm": 0.46422063258116425, "learning_rate": 4.9540911239661815e-06, "loss": 0.2476, "step": 5851 }, { "epoch": 1.8891399975785947, "grad_norm": 0.44786151495390225, "learning_rate": 4.951580193661189e-06, "loss": 0.3265, "step": 5852 }, { "epoch": 1.8894628516082166, "grad_norm": 0.383970755831686, "learning_rate": 4.9490695862374735e-06, "loss": 0.3164, "step": 5853 }, { "epoch": 1.8897857056378387, "grad_norm": 0.3894124970755708, "learning_rate": 4.946559302013132e-06, "loss": 0.2901, "step": 5854 }, { "epoch": 1.8901085596674605, "grad_norm": 0.40532387507955125, "learning_rate": 4.944049341306208e-06, "loss": 0.3498, "step": 5855 }, { "epoch": 1.8904314136970823, "grad_norm": 0.3626008108856745, "learning_rate": 4.941539704434715e-06, "loss": 0.251, "step": 5856 }, { "epoch": 1.8907542677267042, "grad_norm": 0.36809468758503244, "learning_rate": 4.939030391716616e-06, "loss": 0.2996, "step": 5857 }, { "epoch": 1.891077121756326, "grad_norm": 0.3316759092550784, "learning_rate": 4.936521403469839e-06, "loss": 0.2866, "step": 5858 }, { "epoch": 1.8913999757859479, "grad_norm": 0.3478165842949655, "learning_rate": 4.934012740012275e-06, "loss": 0.2744, "step": 5859 }, { "epoch": 1.8917228298155697, "grad_norm": 0.4463522510707729, "learning_rate": 4.931504401661758e-06, "loss": 0.2488, "step": 5860 }, { "epoch": 1.8920456838451916, "grad_norm": 0.453918696906112, "learning_rate": 4.928996388736099e-06, "loss": 0.2736, "step": 5861 }, { "epoch": 1.8923685378748134, "grad_norm": 0.36220805056137023, "learning_rate": 4.926488701553054e-06, "loss": 0.2653, "step": 5862 }, { "epoch": 1.8926913919044352, "grad_norm": 0.4806376119175883, "learning_rate": 4.9239813404303515e-06, "loss": 0.2805, "step": 5863 }, { "epoch": 1.893014245934057, "grad_norm": 0.41339684708156554, "learning_rate": 4.921474305685662e-06, "loss": 0.3289, "step": 5864 }, { "epoch": 1.893337099963679, "grad_norm": 0.3578490590367465, "learning_rate": 4.918967597636631e-06, "loss": 0.3439, "step": 5865 }, { "epoch": 1.8936599539933008, "grad_norm": 0.3790229482704848, "learning_rate": 4.916461216600846e-06, "loss": 0.251, "step": 5866 }, { "epoch": 1.8939828080229226, "grad_norm": 0.3665017960013765, "learning_rate": 4.91395516289587e-06, "loss": 0.2709, "step": 5867 }, { "epoch": 1.8943056620525445, "grad_norm": 0.4142132601915649, "learning_rate": 4.9114494368392134e-06, "loss": 0.2999, "step": 5868 }, { "epoch": 1.8946285160821663, "grad_norm": 0.39702982915997975, "learning_rate": 4.908944038748348e-06, "loss": 0.3098, "step": 5869 }, { "epoch": 1.8949513701117882, "grad_norm": 0.4751017096461278, "learning_rate": 4.906438968940705e-06, "loss": 0.2706, "step": 5870 }, { "epoch": 1.89527422414141, "grad_norm": 0.3664170952226502, "learning_rate": 4.903934227733675e-06, "loss": 0.2921, "step": 5871 }, { "epoch": 1.8955970781710318, "grad_norm": 0.42408379603781937, "learning_rate": 4.901429815444599e-06, "loss": 0.2988, "step": 5872 }, { "epoch": 1.8959199322006537, "grad_norm": 0.4034077619940138, "learning_rate": 4.898925732390792e-06, "loss": 0.2598, "step": 5873 }, { "epoch": 1.8962427862302755, "grad_norm": 0.3588189387026229, "learning_rate": 4.896421978889509e-06, "loss": 0.3247, "step": 5874 }, { "epoch": 1.8965656402598974, "grad_norm": 0.34569740716058245, "learning_rate": 4.893918555257978e-06, "loss": 0.2896, "step": 5875 }, { "epoch": 1.8968884942895192, "grad_norm": 0.39720788553187103, "learning_rate": 4.891415461813375e-06, "loss": 0.2911, "step": 5876 }, { "epoch": 1.897211348319141, "grad_norm": 0.5152827010408777, "learning_rate": 4.888912698872842e-06, "loss": 0.3197, "step": 5877 }, { "epoch": 1.897534202348763, "grad_norm": 0.3494694937145261, "learning_rate": 4.886410266753475e-06, "loss": 0.2614, "step": 5878 }, { "epoch": 1.897857056378385, "grad_norm": 0.3595224376659663, "learning_rate": 4.883908165772326e-06, "loss": 0.2747, "step": 5879 }, { "epoch": 1.8981799104080068, "grad_norm": 0.33406494129718134, "learning_rate": 4.881406396246414e-06, "loss": 0.2869, "step": 5880 }, { "epoch": 1.8985027644376287, "grad_norm": 0.3443490052065079, "learning_rate": 4.878904958492702e-06, "loss": 0.3397, "step": 5881 }, { "epoch": 1.8988256184672505, "grad_norm": 0.6661296226141812, "learning_rate": 4.876403852828125e-06, "loss": 0.2727, "step": 5882 }, { "epoch": 1.8991484724968724, "grad_norm": 0.3880502476385932, "learning_rate": 4.873903079569564e-06, "loss": 0.3226, "step": 5883 }, { "epoch": 1.8994713265264942, "grad_norm": 0.397551320923415, "learning_rate": 4.871402639033868e-06, "loss": 0.2687, "step": 5884 }, { "epoch": 1.899794180556116, "grad_norm": 0.3433184736586299, "learning_rate": 4.868902531537836e-06, "loss": 0.2703, "step": 5885 }, { "epoch": 1.9001170345857379, "grad_norm": 0.45290512467214133, "learning_rate": 4.8664027573982345e-06, "loss": 0.3111, "step": 5886 }, { "epoch": 1.9004398886153597, "grad_norm": 0.4315562336397728, "learning_rate": 4.8639033169317725e-06, "loss": 0.34, "step": 5887 }, { "epoch": 1.9007627426449818, "grad_norm": 0.3445026441416464, "learning_rate": 4.861404210455133e-06, "loss": 0.2878, "step": 5888 }, { "epoch": 1.9010855966746036, "grad_norm": 0.37793411986107167, "learning_rate": 4.858905438284944e-06, "loss": 0.3186, "step": 5889 }, { "epoch": 1.9014084507042255, "grad_norm": 0.39381691971703975, "learning_rate": 4.8564070007378e-06, "loss": 0.2717, "step": 5890 }, { "epoch": 1.9017313047338473, "grad_norm": 0.324691558195642, "learning_rate": 4.853908898130249e-06, "loss": 0.2985, "step": 5891 }, { "epoch": 1.9020541587634692, "grad_norm": 0.3984100377333853, "learning_rate": 4.851411130778798e-06, "loss": 0.2872, "step": 5892 }, { "epoch": 1.902377012793091, "grad_norm": 0.36569816991758514, "learning_rate": 4.848913698999906e-06, "loss": 0.2679, "step": 5893 }, { "epoch": 1.9026998668227129, "grad_norm": 0.33457727340090654, "learning_rate": 4.846416603109999e-06, "loss": 0.3208, "step": 5894 }, { "epoch": 1.9030227208523347, "grad_norm": 0.3462303483905043, "learning_rate": 4.843919843425455e-06, "loss": 0.283, "step": 5895 }, { "epoch": 1.9033455748819565, "grad_norm": 0.37980836609434027, "learning_rate": 4.841423420262609e-06, "loss": 0.3706, "step": 5896 }, { "epoch": 1.9036684289115784, "grad_norm": 0.37022498483093025, "learning_rate": 4.8389273339377556e-06, "loss": 0.2919, "step": 5897 }, { "epoch": 1.9039912829412002, "grad_norm": 0.38925517708749574, "learning_rate": 4.836431584767142e-06, "loss": 0.2764, "step": 5898 }, { "epoch": 1.904314136970822, "grad_norm": 0.3080977486865388, "learning_rate": 4.83393617306698e-06, "loss": 0.2942, "step": 5899 }, { "epoch": 1.904636991000444, "grad_norm": 0.3584077406689359, "learning_rate": 4.831441099153434e-06, "loss": 0.3071, "step": 5900 }, { "epoch": 1.9049598450300658, "grad_norm": 0.41703711235401214, "learning_rate": 4.828946363342626e-06, "loss": 0.3158, "step": 5901 }, { "epoch": 1.9052826990596876, "grad_norm": 0.34747004681750393, "learning_rate": 4.826451965950633e-06, "loss": 0.2772, "step": 5902 }, { "epoch": 1.9056055530893095, "grad_norm": 0.48246743562141475, "learning_rate": 4.823957907293498e-06, "loss": 0.307, "step": 5903 }, { "epoch": 1.9059284071189313, "grad_norm": 0.4953133608180086, "learning_rate": 4.821464187687207e-06, "loss": 0.2375, "step": 5904 }, { "epoch": 1.9062512611485531, "grad_norm": 0.47801127658362763, "learning_rate": 4.818970807447719e-06, "loss": 0.264, "step": 5905 }, { "epoch": 1.906574115178175, "grad_norm": 0.4679899116276067, "learning_rate": 4.816477766890934e-06, "loss": 0.3128, "step": 5906 }, { "epoch": 1.9068969692077968, "grad_norm": 0.3610873382301477, "learning_rate": 4.813985066332724e-06, "loss": 0.2772, "step": 5907 }, { "epoch": 1.9072198232374187, "grad_norm": 0.3940174537219707, "learning_rate": 4.811492706088903e-06, "loss": 0.2894, "step": 5908 }, { "epoch": 1.9075426772670405, "grad_norm": 0.39282674707221554, "learning_rate": 4.8090006864752585e-06, "loss": 0.2968, "step": 5909 }, { "epoch": 1.9078655312966624, "grad_norm": 0.36895729492259516, "learning_rate": 4.806509007807517e-06, "loss": 0.276, "step": 5910 }, { "epoch": 1.9081883853262842, "grad_norm": 0.3415278104972898, "learning_rate": 4.804017670401379e-06, "loss": 0.2351, "step": 5911 }, { "epoch": 1.908511239355906, "grad_norm": 0.40391153406872793, "learning_rate": 4.801526674572486e-06, "loss": 0.2848, "step": 5912 }, { "epoch": 1.9088340933855281, "grad_norm": 0.46262255391447804, "learning_rate": 4.799036020636448e-06, "loss": 0.2699, "step": 5913 }, { "epoch": 1.90915694741515, "grad_norm": 0.4796090309489391, "learning_rate": 4.796545708908829e-06, "loss": 0.3375, "step": 5914 }, { "epoch": 1.9094798014447718, "grad_norm": 0.3820634572003754, "learning_rate": 4.794055739705144e-06, "loss": 0.2179, "step": 5915 }, { "epoch": 1.9098026554743937, "grad_norm": 0.4258121211853127, "learning_rate": 4.791566113340871e-06, "loss": 0.3484, "step": 5916 }, { "epoch": 1.9101255095040155, "grad_norm": 0.5034795217385863, "learning_rate": 4.7890768301314405e-06, "loss": 0.2745, "step": 5917 }, { "epoch": 1.9104483635336373, "grad_norm": 0.39927092927559793, "learning_rate": 4.786587890392245e-06, "loss": 0.2997, "step": 5918 }, { "epoch": 1.9107712175632592, "grad_norm": 0.41512978013797924, "learning_rate": 4.784099294438625e-06, "loss": 0.3029, "step": 5919 }, { "epoch": 1.911094071592881, "grad_norm": 0.36027602592614144, "learning_rate": 4.781611042585887e-06, "loss": 0.2837, "step": 5920 }, { "epoch": 1.9114169256225029, "grad_norm": 0.392496633420289, "learning_rate": 4.779123135149284e-06, "loss": 0.2713, "step": 5921 }, { "epoch": 1.911739779652125, "grad_norm": 0.42326791346234277, "learning_rate": 4.776635572444033e-06, "loss": 0.2701, "step": 5922 }, { "epoch": 1.9120626336817468, "grad_norm": 0.3436686554661593, "learning_rate": 4.774148354785305e-06, "loss": 0.2796, "step": 5923 }, { "epoch": 1.9123854877113686, "grad_norm": 0.4030754836588458, "learning_rate": 4.771661482488228e-06, "loss": 0.285, "step": 5924 }, { "epoch": 1.9127083417409905, "grad_norm": 0.41315277677567075, "learning_rate": 4.769174955867881e-06, "loss": 0.2756, "step": 5925 }, { "epoch": 1.9130311957706123, "grad_norm": 0.3401102725386696, "learning_rate": 4.766688775239311e-06, "loss": 0.2953, "step": 5926 }, { "epoch": 1.9133540498002342, "grad_norm": 0.37411962441957936, "learning_rate": 4.764202940917506e-06, "loss": 0.2961, "step": 5927 }, { "epoch": 1.913676903829856, "grad_norm": 0.38575252346718797, "learning_rate": 4.7617174532174225e-06, "loss": 0.3263, "step": 5928 }, { "epoch": 1.9139997578594778, "grad_norm": 0.42791781035777204, "learning_rate": 4.759232312453964e-06, "loss": 0.3201, "step": 5929 }, { "epoch": 1.9143226118890997, "grad_norm": 0.3993998684603823, "learning_rate": 4.756747518942e-06, "loss": 0.278, "step": 5930 }, { "epoch": 1.9146454659187215, "grad_norm": 0.3418408179198913, "learning_rate": 4.7542630729963465e-06, "loss": 0.3286, "step": 5931 }, { "epoch": 1.9149683199483434, "grad_norm": 0.40190918475838777, "learning_rate": 4.75177897493178e-06, "loss": 0.2806, "step": 5932 }, { "epoch": 1.9152911739779652, "grad_norm": 0.3829312586926099, "learning_rate": 4.749295225063035e-06, "loss": 0.2952, "step": 5933 }, { "epoch": 1.915614028007587, "grad_norm": 0.5478444436734288, "learning_rate": 4.746811823704794e-06, "loss": 0.2924, "step": 5934 }, { "epoch": 1.915936882037209, "grad_norm": 0.3809988469956266, "learning_rate": 4.744328771171708e-06, "loss": 0.269, "step": 5935 }, { "epoch": 1.9162597360668308, "grad_norm": 0.3580689522646092, "learning_rate": 4.741846067778368e-06, "loss": 0.2552, "step": 5936 }, { "epoch": 1.9165825900964526, "grad_norm": 0.40248455638696334, "learning_rate": 4.739363713839336e-06, "loss": 0.3566, "step": 5937 }, { "epoch": 1.9169054441260744, "grad_norm": 0.4439773688463119, "learning_rate": 4.736881709669117e-06, "loss": 0.2995, "step": 5938 }, { "epoch": 1.9172282981556963, "grad_norm": 0.42369552210694456, "learning_rate": 4.734400055582182e-06, "loss": 0.3002, "step": 5939 }, { "epoch": 1.9175511521853181, "grad_norm": 0.32216013415755773, "learning_rate": 4.731918751892949e-06, "loss": 0.2752, "step": 5940 }, { "epoch": 1.91787400621494, "grad_norm": 0.38294507671980615, "learning_rate": 4.729437798915802e-06, "loss": 0.2887, "step": 5941 }, { "epoch": 1.9181968602445618, "grad_norm": 0.36552596209455923, "learning_rate": 4.726957196965068e-06, "loss": 0.3309, "step": 5942 }, { "epoch": 1.9185197142741837, "grad_norm": 0.3695672759905108, "learning_rate": 4.724476946355042e-06, "loss": 0.3225, "step": 5943 }, { "epoch": 1.9188425683038055, "grad_norm": 0.45334931098307857, "learning_rate": 4.72199704739996e-06, "loss": 0.3079, "step": 5944 }, { "epoch": 1.9191654223334274, "grad_norm": 0.3069186657290515, "learning_rate": 4.719517500414029e-06, "loss": 0.2627, "step": 5945 }, { "epoch": 1.9194882763630492, "grad_norm": 0.4254224851619139, "learning_rate": 4.717038305711401e-06, "loss": 0.2936, "step": 5946 }, { "epoch": 1.9198111303926713, "grad_norm": 0.34468872856512434, "learning_rate": 4.714559463606188e-06, "loss": 0.3255, "step": 5947 }, { "epoch": 1.920133984422293, "grad_norm": 0.42740279516936375, "learning_rate": 4.712080974412454e-06, "loss": 0.272, "step": 5948 }, { "epoch": 1.920456838451915, "grad_norm": 0.4091994416180242, "learning_rate": 4.709602838444221e-06, "loss": 0.2859, "step": 5949 }, { "epoch": 1.9207796924815368, "grad_norm": 0.4300108019484763, "learning_rate": 4.707125056015468e-06, "loss": 0.3106, "step": 5950 }, { "epoch": 1.9211025465111586, "grad_norm": 0.5470045460827644, "learning_rate": 4.7046476274401224e-06, "loss": 0.3045, "step": 5951 }, { "epoch": 1.9214254005407805, "grad_norm": 0.4031434119325792, "learning_rate": 4.702170553032076e-06, "loss": 0.2903, "step": 5952 }, { "epoch": 1.9217482545704023, "grad_norm": 0.39060768232112925, "learning_rate": 4.6996938331051634e-06, "loss": 0.2775, "step": 5953 }, { "epoch": 1.9220711086000242, "grad_norm": 0.3708648332184379, "learning_rate": 4.697217467973189e-06, "loss": 0.2838, "step": 5954 }, { "epoch": 1.9223939626296462, "grad_norm": 0.38961051035602245, "learning_rate": 4.694741457949899e-06, "loss": 0.2459, "step": 5955 }, { "epoch": 1.922716816659268, "grad_norm": 0.41802727286521074, "learning_rate": 4.692265803349008e-06, "loss": 0.2958, "step": 5956 }, { "epoch": 1.92303967068889, "grad_norm": 0.47431187725497814, "learning_rate": 4.689790504484168e-06, "loss": 0.3334, "step": 5957 }, { "epoch": 1.9233625247185118, "grad_norm": 0.33210918695580716, "learning_rate": 4.687315561669005e-06, "loss": 0.3402, "step": 5958 }, { "epoch": 1.9236853787481336, "grad_norm": 0.46765222959378044, "learning_rate": 4.684840975217085e-06, "loss": 0.276, "step": 5959 }, { "epoch": 1.9240082327777555, "grad_norm": 0.3692987291361521, "learning_rate": 4.68236674544194e-06, "loss": 0.3396, "step": 5960 }, { "epoch": 1.9243310868073773, "grad_norm": 0.3371097813326958, "learning_rate": 4.679892872657043e-06, "loss": 0.3024, "step": 5961 }, { "epoch": 1.9246539408369991, "grad_norm": 0.43924205805772654, "learning_rate": 4.677419357175838e-06, "loss": 0.2954, "step": 5962 }, { "epoch": 1.924976794866621, "grad_norm": 0.4338429606630625, "learning_rate": 4.674946199311713e-06, "loss": 0.3145, "step": 5963 }, { "epoch": 1.9252996488962428, "grad_norm": 0.41558913709606665, "learning_rate": 4.672473399378017e-06, "loss": 0.2557, "step": 5964 }, { "epoch": 1.9256225029258647, "grad_norm": 0.3054302756074077, "learning_rate": 4.670000957688043e-06, "loss": 0.252, "step": 5965 }, { "epoch": 1.9259453569554865, "grad_norm": 0.45301332161247604, "learning_rate": 4.667528874555055e-06, "loss": 0.3223, "step": 5966 }, { "epoch": 1.9262682109851084, "grad_norm": 0.35468937266506595, "learning_rate": 4.6650571502922545e-06, "loss": 0.2988, "step": 5967 }, { "epoch": 1.9265910650147302, "grad_norm": 0.40256474633726813, "learning_rate": 4.66258578521281e-06, "loss": 0.2708, "step": 5968 }, { "epoch": 1.926913919044352, "grad_norm": 0.3547468176710823, "learning_rate": 4.66011477962984e-06, "loss": 0.2924, "step": 5969 }, { "epoch": 1.927236773073974, "grad_norm": 0.393363145986618, "learning_rate": 4.657644133856416e-06, "loss": 0.2547, "step": 5970 }, { "epoch": 1.9275596271035957, "grad_norm": 0.2934485599085621, "learning_rate": 4.655173848205566e-06, "loss": 0.2278, "step": 5971 }, { "epoch": 1.9278824811332176, "grad_norm": 0.40108051417331825, "learning_rate": 4.652703922990271e-06, "loss": 0.3605, "step": 5972 }, { "epoch": 1.9282053351628394, "grad_norm": 0.38287223436498885, "learning_rate": 4.650234358523472e-06, "loss": 0.3092, "step": 5973 }, { "epoch": 1.9285281891924613, "grad_norm": 0.3947285389660819, "learning_rate": 4.6477651551180515e-06, "loss": 0.3004, "step": 5974 }, { "epoch": 1.9288510432220831, "grad_norm": 0.34721384763661894, "learning_rate": 4.6452963130868606e-06, "loss": 0.2885, "step": 5975 }, { "epoch": 1.929173897251705, "grad_norm": 0.3635422570052204, "learning_rate": 4.642827832742693e-06, "loss": 0.2841, "step": 5976 }, { "epoch": 1.9294967512813268, "grad_norm": 0.4023786327693446, "learning_rate": 4.640359714398307e-06, "loss": 0.2726, "step": 5977 }, { "epoch": 1.9298196053109486, "grad_norm": 0.4413861647097366, "learning_rate": 4.637891958366406e-06, "loss": 0.3115, "step": 5978 }, { "epoch": 1.9301424593405705, "grad_norm": 0.38058940428289645, "learning_rate": 4.635424564959652e-06, "loss": 0.292, "step": 5979 }, { "epoch": 1.9304653133701923, "grad_norm": 0.37264095228748756, "learning_rate": 4.632957534490661e-06, "loss": 0.3402, "step": 5980 }, { "epoch": 1.9307881673998144, "grad_norm": 0.34239790553478083, "learning_rate": 4.630490867272004e-06, "loss": 0.3221, "step": 5981 }, { "epoch": 1.9311110214294362, "grad_norm": 0.41599484147931837, "learning_rate": 4.628024563616199e-06, "loss": 0.2397, "step": 5982 }, { "epoch": 1.931433875459058, "grad_norm": 0.38383833878878987, "learning_rate": 4.625558623835731e-06, "loss": 0.2681, "step": 5983 }, { "epoch": 1.93175672948868, "grad_norm": 0.34853896541937757, "learning_rate": 4.623093048243025e-06, "loss": 0.2777, "step": 5984 }, { "epoch": 1.9320795835183018, "grad_norm": 0.40190103178453057, "learning_rate": 4.620627837150468e-06, "loss": 0.2581, "step": 5985 }, { "epoch": 1.9324024375479236, "grad_norm": 0.4000859101890344, "learning_rate": 4.618162990870397e-06, "loss": 0.3166, "step": 5986 }, { "epoch": 1.9327252915775455, "grad_norm": 0.41394473782444446, "learning_rate": 4.615698509715107e-06, "loss": 0.2632, "step": 5987 }, { "epoch": 1.9330481456071673, "grad_norm": 0.3485993460699291, "learning_rate": 4.6132343939968475e-06, "loss": 0.3119, "step": 5988 }, { "epoch": 1.9333709996367894, "grad_norm": 0.4540851696509463, "learning_rate": 4.610770644027812e-06, "loss": 0.3281, "step": 5989 }, { "epoch": 1.9336938536664112, "grad_norm": 0.4491036689185855, "learning_rate": 4.608307260120159e-06, "loss": 0.3285, "step": 5990 }, { "epoch": 1.934016707696033, "grad_norm": 0.48438948513009017, "learning_rate": 4.605844242585993e-06, "loss": 0.2828, "step": 5991 }, { "epoch": 1.934339561725655, "grad_norm": 0.3695438430090427, "learning_rate": 4.603381591737377e-06, "loss": 0.2935, "step": 5992 }, { "epoch": 1.9346624157552768, "grad_norm": 0.4004920143476656, "learning_rate": 4.600919307886323e-06, "loss": 0.2562, "step": 5993 }, { "epoch": 1.9349852697848986, "grad_norm": 0.3952697374772494, "learning_rate": 4.598457391344801e-06, "loss": 0.3347, "step": 5994 }, { "epoch": 1.9353081238145204, "grad_norm": 0.39770027645921957, "learning_rate": 4.595995842424731e-06, "loss": 0.352, "step": 5995 }, { "epoch": 1.9356309778441423, "grad_norm": 0.3697281320941182, "learning_rate": 4.593534661437991e-06, "loss": 0.3256, "step": 5996 }, { "epoch": 1.9359538318737641, "grad_norm": 0.5259327835019579, "learning_rate": 4.591073848696404e-06, "loss": 0.268, "step": 5997 }, { "epoch": 1.936276685903386, "grad_norm": 0.39756830253188963, "learning_rate": 4.588613404511758e-06, "loss": 0.3068, "step": 5998 }, { "epoch": 1.9365995399330078, "grad_norm": 0.41902006264486125, "learning_rate": 4.586153329195781e-06, "loss": 0.3161, "step": 5999 }, { "epoch": 1.9369223939626297, "grad_norm": 0.41704988253628517, "learning_rate": 4.5836936230601665e-06, "loss": 0.374, "step": 6000 }, { "epoch": 1.9372452479922515, "grad_norm": 0.3734467052283309, "learning_rate": 4.581234286416554e-06, "loss": 0.2831, "step": 6001 }, { "epoch": 1.9375681020218734, "grad_norm": 0.30084179767914976, "learning_rate": 4.578775319576537e-06, "loss": 0.2789, "step": 6002 }, { "epoch": 1.9378909560514952, "grad_norm": 0.4388913077641206, "learning_rate": 4.5763167228516645e-06, "loss": 0.3525, "step": 6003 }, { "epoch": 1.938213810081117, "grad_norm": 0.4849084490063589, "learning_rate": 4.573858496553439e-06, "loss": 0.2718, "step": 6004 }, { "epoch": 1.9385366641107389, "grad_norm": 0.3617565629351912, "learning_rate": 4.571400640993309e-06, "loss": 0.2986, "step": 6005 }, { "epoch": 1.9388595181403607, "grad_norm": 0.3433957492016698, "learning_rate": 4.568943156482686e-06, "loss": 0.3404, "step": 6006 }, { "epoch": 1.9391823721699826, "grad_norm": 0.3575949636989237, "learning_rate": 4.566486043332931e-06, "loss": 0.2775, "step": 6007 }, { "epoch": 1.9395052261996044, "grad_norm": 0.43697614799805323, "learning_rate": 4.564029301855352e-06, "loss": 0.2579, "step": 6008 }, { "epoch": 1.9398280802292263, "grad_norm": 0.35176138891167885, "learning_rate": 4.561572932361218e-06, "loss": 0.3278, "step": 6009 }, { "epoch": 1.940150934258848, "grad_norm": 0.41913054569977853, "learning_rate": 4.559116935161746e-06, "loss": 0.2568, "step": 6010 }, { "epoch": 1.94047378828847, "grad_norm": 0.3887263066746564, "learning_rate": 4.556661310568113e-06, "loss": 0.345, "step": 6011 }, { "epoch": 1.9407966423180918, "grad_norm": 0.36342144646164043, "learning_rate": 4.554206058891434e-06, "loss": 0.2896, "step": 6012 }, { "epoch": 1.9411194963477136, "grad_norm": 0.3363772465639559, "learning_rate": 4.551751180442794e-06, "loss": 0.3172, "step": 6013 }, { "epoch": 1.9414423503773357, "grad_norm": 0.31839848342900395, "learning_rate": 4.549296675533216e-06, "loss": 0.2497, "step": 6014 }, { "epoch": 1.9417652044069575, "grad_norm": 0.387151606210158, "learning_rate": 4.54684254447369e-06, "loss": 0.3664, "step": 6015 }, { "epoch": 1.9420880584365794, "grad_norm": 0.39766643664955587, "learning_rate": 4.544388787575143e-06, "loss": 0.294, "step": 6016 }, { "epoch": 1.9424109124662012, "grad_norm": 0.41563055724807496, "learning_rate": 4.541935405148469e-06, "loss": 0.3251, "step": 6017 }, { "epoch": 1.942733766495823, "grad_norm": 0.33105004240426295, "learning_rate": 4.539482397504504e-06, "loss": 0.2714, "step": 6018 }, { "epoch": 1.943056620525445, "grad_norm": 0.40392489179833024, "learning_rate": 4.5370297649540444e-06, "loss": 0.319, "step": 6019 }, { "epoch": 1.9433794745550668, "grad_norm": 0.3688864491466621, "learning_rate": 4.53457750780783e-06, "loss": 0.3198, "step": 6020 }, { "epoch": 1.9437023285846886, "grad_norm": 0.3081170702273548, "learning_rate": 4.532125626376565e-06, "loss": 0.2983, "step": 6021 }, { "epoch": 1.9440251826143105, "grad_norm": 0.34963066243412605, "learning_rate": 4.529674120970894e-06, "loss": 0.2812, "step": 6022 }, { "epoch": 1.9443480366439325, "grad_norm": 0.37840539656622596, "learning_rate": 4.5272229919014215e-06, "loss": 0.3282, "step": 6023 }, { "epoch": 1.9446708906735544, "grad_norm": 0.4114998802875622, "learning_rate": 4.524772239478705e-06, "loss": 0.3321, "step": 6024 }, { "epoch": 1.9449937447031762, "grad_norm": 0.377744258930208, "learning_rate": 4.5223218640132455e-06, "loss": 0.3582, "step": 6025 }, { "epoch": 1.945316598732798, "grad_norm": 0.37106998396848806, "learning_rate": 4.519871865815507e-06, "loss": 0.2714, "step": 6026 }, { "epoch": 1.94563945276242, "grad_norm": 0.4652001763290255, "learning_rate": 4.517422245195897e-06, "loss": 0.3505, "step": 6027 }, { "epoch": 1.9459623067920417, "grad_norm": 0.33399743664292736, "learning_rate": 4.514973002464785e-06, "loss": 0.2809, "step": 6028 }, { "epoch": 1.9462851608216636, "grad_norm": 0.3996305021800167, "learning_rate": 4.512524137932479e-06, "loss": 0.2964, "step": 6029 }, { "epoch": 1.9466080148512854, "grad_norm": 0.40148493247432904, "learning_rate": 4.510075651909255e-06, "loss": 0.2935, "step": 6030 }, { "epoch": 1.9469308688809073, "grad_norm": 0.33912444702338157, "learning_rate": 4.507627544705326e-06, "loss": 0.2326, "step": 6031 }, { "epoch": 1.9472537229105291, "grad_norm": 0.391129765422458, "learning_rate": 4.505179816630865e-06, "loss": 0.2975, "step": 6032 }, { "epoch": 1.947576576940151, "grad_norm": 0.3867425519447816, "learning_rate": 4.502732467995998e-06, "loss": 0.2825, "step": 6033 }, { "epoch": 1.9478994309697728, "grad_norm": 0.4136734577999404, "learning_rate": 4.500285499110803e-06, "loss": 0.2855, "step": 6034 }, { "epoch": 1.9482222849993946, "grad_norm": 0.35466770744896525, "learning_rate": 4.4978389102853005e-06, "loss": 0.2941, "step": 6035 }, { "epoch": 1.9485451390290165, "grad_norm": 0.44424426373294557, "learning_rate": 4.495392701829477e-06, "loss": 0.3572, "step": 6036 }, { "epoch": 1.9488679930586383, "grad_norm": 0.31995125445586153, "learning_rate": 4.492946874053257e-06, "loss": 0.2749, "step": 6037 }, { "epoch": 1.9491908470882602, "grad_norm": 0.3773343364749395, "learning_rate": 4.490501427266531e-06, "loss": 0.2631, "step": 6038 }, { "epoch": 1.949513701117882, "grad_norm": 0.36220252713517587, "learning_rate": 4.488056361779127e-06, "loss": 0.2425, "step": 6039 }, { "epoch": 1.9498365551475039, "grad_norm": 0.3238914063728424, "learning_rate": 4.4856116779008346e-06, "loss": 0.273, "step": 6040 }, { "epoch": 1.9501594091771257, "grad_norm": 0.44121399254168964, "learning_rate": 4.483167375941391e-06, "loss": 0.2701, "step": 6041 }, { "epoch": 1.9504822632067476, "grad_norm": 0.3779839983916233, "learning_rate": 4.4807234562104865e-06, "loss": 0.2889, "step": 6042 }, { "epoch": 1.9508051172363694, "grad_norm": 0.3666265498678397, "learning_rate": 4.4782799190177645e-06, "loss": 0.313, "step": 6043 }, { "epoch": 1.9511279712659912, "grad_norm": 0.3433731903797776, "learning_rate": 4.475836764672813e-06, "loss": 0.3092, "step": 6044 }, { "epoch": 1.951450825295613, "grad_norm": 0.41624084198275363, "learning_rate": 4.473393993485182e-06, "loss": 0.2905, "step": 6045 }, { "epoch": 1.951773679325235, "grad_norm": 0.4210346667998559, "learning_rate": 4.470951605764361e-06, "loss": 0.3154, "step": 6046 }, { "epoch": 1.9520965333548568, "grad_norm": 0.4650061314311138, "learning_rate": 4.4685096018198034e-06, "loss": 0.295, "step": 6047 }, { "epoch": 1.9524193873844788, "grad_norm": 0.3970514595317175, "learning_rate": 4.466067981960902e-06, "loss": 0.2824, "step": 6048 }, { "epoch": 1.9527422414141007, "grad_norm": 0.32742198109275084, "learning_rate": 4.463626746497011e-06, "loss": 0.2771, "step": 6049 }, { "epoch": 1.9530650954437225, "grad_norm": 0.39100658724930915, "learning_rate": 4.461185895737428e-06, "loss": 0.2634, "step": 6050 }, { "epoch": 1.9533879494733444, "grad_norm": 0.3351900829364686, "learning_rate": 4.45874542999141e-06, "loss": 0.3064, "step": 6051 }, { "epoch": 1.9537108035029662, "grad_norm": 0.5557687046182005, "learning_rate": 4.456305349568155e-06, "loss": 0.3402, "step": 6052 }, { "epoch": 1.954033657532588, "grad_norm": 0.35499678321438133, "learning_rate": 4.453865654776824e-06, "loss": 0.324, "step": 6053 }, { "epoch": 1.95435651156221, "grad_norm": 0.34954061651525314, "learning_rate": 4.451426345926516e-06, "loss": 0.2843, "step": 6054 }, { "epoch": 1.9546793655918318, "grad_norm": 0.4362463203932769, "learning_rate": 4.448987423326295e-06, "loss": 0.3189, "step": 6055 }, { "epoch": 1.9550022196214536, "grad_norm": 0.37699769086183027, "learning_rate": 4.4465488872851635e-06, "loss": 0.2801, "step": 6056 }, { "epoch": 1.9553250736510757, "grad_norm": 0.33474845342187176, "learning_rate": 4.4441107381120855e-06, "loss": 0.2366, "step": 6057 }, { "epoch": 1.9556479276806975, "grad_norm": 0.3898448805648527, "learning_rate": 4.441672976115966e-06, "loss": 0.307, "step": 6058 }, { "epoch": 1.9559707817103194, "grad_norm": 0.41000516368725576, "learning_rate": 4.439235601605673e-06, "loss": 0.2849, "step": 6059 }, { "epoch": 1.9562936357399412, "grad_norm": 0.5221736882258582, "learning_rate": 4.43679861489001e-06, "loss": 0.2988, "step": 6060 }, { "epoch": 1.956616489769563, "grad_norm": 0.3701964481967969, "learning_rate": 4.434362016277746e-06, "loss": 0.2882, "step": 6061 }, { "epoch": 1.9569393437991849, "grad_norm": 0.37035839940907395, "learning_rate": 4.431925806077594e-06, "loss": 0.308, "step": 6062 }, { "epoch": 1.9572621978288067, "grad_norm": 0.3980443541240902, "learning_rate": 4.429489984598216e-06, "loss": 0.2565, "step": 6063 }, { "epoch": 1.9575850518584286, "grad_norm": 0.46388862954037263, "learning_rate": 4.427054552148229e-06, "loss": 0.3138, "step": 6064 }, { "epoch": 1.9579079058880504, "grad_norm": 0.3935967801528159, "learning_rate": 4.424619509036197e-06, "loss": 0.298, "step": 6065 }, { "epoch": 1.9582307599176723, "grad_norm": 0.42725190027439836, "learning_rate": 4.4221848555706414e-06, "loss": 0.3196, "step": 6066 }, { "epoch": 1.958553613947294, "grad_norm": 0.3335191549107625, "learning_rate": 4.419750592060023e-06, "loss": 0.3326, "step": 6067 }, { "epoch": 1.958876467976916, "grad_norm": 0.4646471879688329, "learning_rate": 4.417316718812766e-06, "loss": 0.3009, "step": 6068 }, { "epoch": 1.9591993220065378, "grad_norm": 0.44135544286073564, "learning_rate": 4.414883236137233e-06, "loss": 0.3213, "step": 6069 }, { "epoch": 1.9595221760361596, "grad_norm": 0.4405833596274114, "learning_rate": 4.412450144341748e-06, "loss": 0.3265, "step": 6070 }, { "epoch": 1.9598450300657815, "grad_norm": 0.41165703338815085, "learning_rate": 4.410017443734576e-06, "loss": 0.323, "step": 6071 }, { "epoch": 1.9601678840954033, "grad_norm": 0.31972969999375944, "learning_rate": 4.4075851346239385e-06, "loss": 0.2732, "step": 6072 }, { "epoch": 1.9604907381250252, "grad_norm": 0.3337609405234289, "learning_rate": 4.405153217318006e-06, "loss": 0.2912, "step": 6073 }, { "epoch": 1.960813592154647, "grad_norm": 0.40099441525598434, "learning_rate": 4.402721692124901e-06, "loss": 0.3202, "step": 6074 }, { "epoch": 1.9611364461842689, "grad_norm": 0.3894322132370424, "learning_rate": 4.4002905593526895e-06, "loss": 0.3587, "step": 6075 }, { "epoch": 1.9614593002138907, "grad_norm": 0.4212441136701599, "learning_rate": 4.397859819309398e-06, "loss": 0.2846, "step": 6076 }, { "epoch": 1.9617821542435125, "grad_norm": 0.40155603611367635, "learning_rate": 4.395429472302994e-06, "loss": 0.392, "step": 6077 }, { "epoch": 1.9621050082731344, "grad_norm": 0.4201881463810184, "learning_rate": 4.392999518641402e-06, "loss": 0.2984, "step": 6078 }, { "epoch": 1.9624278623027562, "grad_norm": 0.3488020460735956, "learning_rate": 4.39056995863249e-06, "loss": 0.27, "step": 6079 }, { "epoch": 1.962750716332378, "grad_norm": 0.40520543572847134, "learning_rate": 4.388140792584083e-06, "loss": 0.287, "step": 6080 }, { "epoch": 1.963073570362, "grad_norm": 0.3863258493869804, "learning_rate": 4.385712020803954e-06, "loss": 0.267, "step": 6081 }, { "epoch": 1.963396424391622, "grad_norm": 0.4275691580686077, "learning_rate": 4.383283643599821e-06, "loss": 0.2718, "step": 6082 }, { "epoch": 1.9637192784212438, "grad_norm": 0.4068913314741355, "learning_rate": 4.3808556612793615e-06, "loss": 0.2495, "step": 6083 }, { "epoch": 1.9640421324508657, "grad_norm": 0.43640715040962613, "learning_rate": 4.378428074150191e-06, "loss": 0.2944, "step": 6084 }, { "epoch": 1.9643649864804875, "grad_norm": 0.45088626769705786, "learning_rate": 4.376000882519888e-06, "loss": 0.3028, "step": 6085 }, { "epoch": 1.9646878405101094, "grad_norm": 0.35387068437893393, "learning_rate": 4.37357408669597e-06, "loss": 0.307, "step": 6086 }, { "epoch": 1.9650106945397312, "grad_norm": 0.351235827306591, "learning_rate": 4.371147686985911e-06, "loss": 0.2984, "step": 6087 }, { "epoch": 1.965333548569353, "grad_norm": 0.4964056964970726, "learning_rate": 4.368721683697128e-06, "loss": 0.3432, "step": 6088 }, { "epoch": 1.965656402598975, "grad_norm": 0.40404867846811043, "learning_rate": 4.366296077137e-06, "loss": 0.3134, "step": 6089 }, { "epoch": 1.965979256628597, "grad_norm": 0.37814931126239737, "learning_rate": 4.3638708676128395e-06, "loss": 0.3528, "step": 6090 }, { "epoch": 1.9663021106582188, "grad_norm": 0.34793039286414423, "learning_rate": 4.361446055431922e-06, "loss": 0.2656, "step": 6091 }, { "epoch": 1.9666249646878406, "grad_norm": 0.4136752112807135, "learning_rate": 4.359021640901467e-06, "loss": 0.3721, "step": 6092 }, { "epoch": 1.9669478187174625, "grad_norm": 0.3721762928305299, "learning_rate": 4.356597624328648e-06, "loss": 0.2518, "step": 6093 }, { "epoch": 1.9672706727470843, "grad_norm": 0.3417435851716997, "learning_rate": 4.3541740060205775e-06, "loss": 0.2962, "step": 6094 }, { "epoch": 1.9675935267767062, "grad_norm": 0.41577432898937194, "learning_rate": 4.351750786284333e-06, "loss": 0.287, "step": 6095 }, { "epoch": 1.967916380806328, "grad_norm": 0.44969124561529933, "learning_rate": 4.349327965426924e-06, "loss": 0.2849, "step": 6096 }, { "epoch": 1.9682392348359499, "grad_norm": 0.4261707805396616, "learning_rate": 4.346905543755323e-06, "loss": 0.249, "step": 6097 }, { "epoch": 1.9685620888655717, "grad_norm": 0.43437120014221164, "learning_rate": 4.344483521576451e-06, "loss": 0.293, "step": 6098 }, { "epoch": 1.9688849428951936, "grad_norm": 0.42308261396073077, "learning_rate": 4.342061899197169e-06, "loss": 0.2933, "step": 6099 }, { "epoch": 1.9692077969248154, "grad_norm": 0.3500861563986065, "learning_rate": 4.339640676924298e-06, "loss": 0.3009, "step": 6100 }, { "epoch": 1.9695306509544372, "grad_norm": 0.36674657290424867, "learning_rate": 4.337219855064598e-06, "loss": 0.2799, "step": 6101 }, { "epoch": 1.969853504984059, "grad_norm": 0.4719684544036612, "learning_rate": 4.334799433924791e-06, "loss": 0.2596, "step": 6102 }, { "epoch": 1.970176359013681, "grad_norm": 0.44819822730815995, "learning_rate": 4.332379413811533e-06, "loss": 0.3297, "step": 6103 }, { "epoch": 1.9704992130433028, "grad_norm": 0.4593440698174741, "learning_rate": 4.329959795031444e-06, "loss": 0.2989, "step": 6104 }, { "epoch": 1.9708220670729246, "grad_norm": 0.3841186109520847, "learning_rate": 4.327540577891081e-06, "loss": 0.3287, "step": 6105 }, { "epoch": 1.9711449211025465, "grad_norm": 0.4060472003727058, "learning_rate": 4.3251217626969585e-06, "loss": 0.2635, "step": 6106 }, { "epoch": 1.9714677751321683, "grad_norm": 0.45585992678116033, "learning_rate": 4.3227033497555385e-06, "loss": 0.324, "step": 6107 }, { "epoch": 1.9717906291617902, "grad_norm": 0.3732008028026646, "learning_rate": 4.320285339373227e-06, "loss": 0.3193, "step": 6108 }, { "epoch": 1.972113483191412, "grad_norm": 0.3783722438404593, "learning_rate": 4.317867731856383e-06, "loss": 0.2277, "step": 6109 }, { "epoch": 1.9724363372210338, "grad_norm": 0.3429470226575922, "learning_rate": 4.3154505275113184e-06, "loss": 0.261, "step": 6110 }, { "epoch": 1.9727591912506557, "grad_norm": 0.39788590929865525, "learning_rate": 4.313033726644284e-06, "loss": 0.2891, "step": 6111 }, { "epoch": 1.9730820452802775, "grad_norm": 0.3951572311764727, "learning_rate": 4.310617329561492e-06, "loss": 0.2851, "step": 6112 }, { "epoch": 1.9734048993098994, "grad_norm": 0.38380183528701023, "learning_rate": 4.308201336569087e-06, "loss": 0.2877, "step": 6113 }, { "epoch": 1.9737277533395212, "grad_norm": 0.3641847169193851, "learning_rate": 4.305785747973182e-06, "loss": 0.3193, "step": 6114 }, { "epoch": 1.974050607369143, "grad_norm": 0.33018082087280065, "learning_rate": 4.303370564079821e-06, "loss": 0.27, "step": 6115 }, { "epoch": 1.9743734613987651, "grad_norm": 0.3847251091149535, "learning_rate": 4.300955785195006e-06, "loss": 0.2711, "step": 6116 }, { "epoch": 1.974696315428387, "grad_norm": 0.37371368358661133, "learning_rate": 4.298541411624692e-06, "loss": 0.3024, "step": 6117 }, { "epoch": 1.9750191694580088, "grad_norm": 0.4059606554231443, "learning_rate": 4.2961274436747684e-06, "loss": 0.3459, "step": 6118 }, { "epoch": 1.9753420234876307, "grad_norm": 0.3801750492585734, "learning_rate": 4.2937138816510904e-06, "loss": 0.298, "step": 6119 }, { "epoch": 1.9756648775172525, "grad_norm": 0.3177355943572506, "learning_rate": 4.2913007258594435e-06, "loss": 0.2663, "step": 6120 }, { "epoch": 1.9759877315468743, "grad_norm": 0.3777099674260815, "learning_rate": 4.2888879766055796e-06, "loss": 0.2407, "step": 6121 }, { "epoch": 1.9763105855764962, "grad_norm": 0.5093406891306992, "learning_rate": 4.286475634195184e-06, "loss": 0.2741, "step": 6122 }, { "epoch": 1.976633439606118, "grad_norm": 0.42077485836524164, "learning_rate": 4.284063698933901e-06, "loss": 0.308, "step": 6123 }, { "epoch": 1.97695629363574, "grad_norm": 0.40788107033545085, "learning_rate": 4.281652171127318e-06, "loss": 0.3411, "step": 6124 }, { "epoch": 1.977279147665362, "grad_norm": 0.43037177302492524, "learning_rate": 4.279241051080976e-06, "loss": 0.3308, "step": 6125 }, { "epoch": 1.9776020016949838, "grad_norm": 0.4344441610542111, "learning_rate": 4.276830339100354e-06, "loss": 0.3395, "step": 6126 }, { "epoch": 1.9779248557246056, "grad_norm": 0.4301507140121326, "learning_rate": 4.274420035490895e-06, "loss": 0.3228, "step": 6127 }, { "epoch": 1.9782477097542275, "grad_norm": 0.4369039819251354, "learning_rate": 4.27201014055797e-06, "loss": 0.2634, "step": 6128 }, { "epoch": 1.9785705637838493, "grad_norm": 0.43006347086532193, "learning_rate": 4.269600654606919e-06, "loss": 0.2964, "step": 6129 }, { "epoch": 1.9788934178134712, "grad_norm": 0.37619319082666763, "learning_rate": 4.267191577943013e-06, "loss": 0.2746, "step": 6130 }, { "epoch": 1.979216271843093, "grad_norm": 0.3656282913740362, "learning_rate": 4.264782910871486e-06, "loss": 0.2453, "step": 6131 }, { "epoch": 1.9795391258727149, "grad_norm": 0.4041292617758698, "learning_rate": 4.262374653697506e-06, "loss": 0.2841, "step": 6132 }, { "epoch": 1.9798619799023367, "grad_norm": 0.4240006885181061, "learning_rate": 4.259966806726201e-06, "loss": 0.2718, "step": 6133 }, { "epoch": 1.9801848339319585, "grad_norm": 0.3964834075269621, "learning_rate": 4.257559370262637e-06, "loss": 0.2843, "step": 6134 }, { "epoch": 1.9805076879615804, "grad_norm": 0.3502412364223434, "learning_rate": 4.255152344611837e-06, "loss": 0.2696, "step": 6135 }, { "epoch": 1.9808305419912022, "grad_norm": 0.4000204547586704, "learning_rate": 4.252745730078768e-06, "loss": 0.2847, "step": 6136 }, { "epoch": 1.981153396020824, "grad_norm": 0.41613778416191904, "learning_rate": 4.25033952696834e-06, "loss": 0.2696, "step": 6137 }, { "epoch": 1.981476250050446, "grad_norm": 0.45784920884581076, "learning_rate": 4.2479337355854206e-06, "loss": 0.322, "step": 6138 }, { "epoch": 1.9817991040800678, "grad_norm": 0.36332593719498996, "learning_rate": 4.245528356234817e-06, "loss": 0.3544, "step": 6139 }, { "epoch": 1.9821219581096896, "grad_norm": 0.3803604195033847, "learning_rate": 4.2431233892212926e-06, "loss": 0.3341, "step": 6140 }, { "epoch": 1.9824448121393115, "grad_norm": 0.41909995319008503, "learning_rate": 4.240718834849547e-06, "loss": 0.2424, "step": 6141 }, { "epoch": 1.9827676661689333, "grad_norm": 0.36726556991275133, "learning_rate": 4.238314693424239e-06, "loss": 0.2925, "step": 6142 }, { "epoch": 1.9830905201985551, "grad_norm": 0.3632698698497695, "learning_rate": 4.235910965249966e-06, "loss": 0.3181, "step": 6143 }, { "epoch": 1.983413374228177, "grad_norm": 0.37898336805406824, "learning_rate": 4.233507650631281e-06, "loss": 0.2907, "step": 6144 }, { "epoch": 1.9837362282577988, "grad_norm": 0.4664976932972619, "learning_rate": 4.231104749872676e-06, "loss": 0.3013, "step": 6145 }, { "epoch": 1.9840590822874207, "grad_norm": 0.3944243236966305, "learning_rate": 4.2287022632786e-06, "loss": 0.3163, "step": 6146 }, { "epoch": 1.9843819363170425, "grad_norm": 0.35526374050971654, "learning_rate": 4.226300191153439e-06, "loss": 0.3119, "step": 6147 }, { "epoch": 1.9847047903466644, "grad_norm": 0.359115964181666, "learning_rate": 4.22389853380154e-06, "loss": 0.288, "step": 6148 }, { "epoch": 1.9850276443762864, "grad_norm": 0.43210386158799524, "learning_rate": 4.22149729152718e-06, "loss": 0.27, "step": 6149 }, { "epoch": 1.9853504984059083, "grad_norm": 0.4318933054364523, "learning_rate": 4.219096464634604e-06, "loss": 0.3082, "step": 6150 }, { "epoch": 1.9856733524355301, "grad_norm": 0.3836019269322181, "learning_rate": 4.216696053427984e-06, "loss": 0.3693, "step": 6151 }, { "epoch": 1.985996206465152, "grad_norm": 0.4084663769983039, "learning_rate": 4.214296058211452e-06, "loss": 0.25, "step": 6152 }, { "epoch": 1.9863190604947738, "grad_norm": 0.3785210840145202, "learning_rate": 4.211896479289088e-06, "loss": 0.2352, "step": 6153 }, { "epoch": 1.9866419145243956, "grad_norm": 0.34221406057734677, "learning_rate": 4.2094973169649095e-06, "loss": 0.2482, "step": 6154 }, { "epoch": 1.9869647685540175, "grad_norm": 0.3493033725731087, "learning_rate": 4.20709857154289e-06, "loss": 0.3553, "step": 6155 }, { "epoch": 1.9872876225836393, "grad_norm": 0.3621207693723672, "learning_rate": 4.204700243326946e-06, "loss": 0.302, "step": 6156 }, { "epoch": 1.9876104766132612, "grad_norm": 0.39925627822201626, "learning_rate": 4.202302332620946e-06, "loss": 0.2852, "step": 6157 }, { "epoch": 1.9879333306428832, "grad_norm": 0.46121221390739564, "learning_rate": 4.199904839728697e-06, "loss": 0.3541, "step": 6158 }, { "epoch": 1.988256184672505, "grad_norm": 0.39520585179584905, "learning_rate": 4.197507764953963e-06, "loss": 0.2574, "step": 6159 }, { "epoch": 1.988579038702127, "grad_norm": 0.3647547101188512, "learning_rate": 4.195111108600445e-06, "loss": 0.2877, "step": 6160 }, { "epoch": 1.9889018927317488, "grad_norm": 0.3826658577553163, "learning_rate": 4.1927148709718e-06, "loss": 0.2773, "step": 6161 }, { "epoch": 1.9892247467613706, "grad_norm": 0.33470400315421306, "learning_rate": 4.190319052371625e-06, "loss": 0.2552, "step": 6162 }, { "epoch": 1.9895476007909925, "grad_norm": 0.482353497128805, "learning_rate": 4.1879236531034705e-06, "loss": 0.2963, "step": 6163 }, { "epoch": 1.9898704548206143, "grad_norm": 0.35707237040933726, "learning_rate": 4.185528673470825e-06, "loss": 0.2438, "step": 6164 }, { "epoch": 1.9901933088502362, "grad_norm": 0.3923168148747038, "learning_rate": 4.183134113777136e-06, "loss": 0.3335, "step": 6165 }, { "epoch": 1.990516162879858, "grad_norm": 0.4636986492893046, "learning_rate": 4.180739974325783e-06, "loss": 0.3199, "step": 6166 }, { "epoch": 1.9908390169094798, "grad_norm": 0.5629122003404772, "learning_rate": 4.178346255420109e-06, "loss": 0.2871, "step": 6167 }, { "epoch": 1.9911618709391017, "grad_norm": 0.33466072163727273, "learning_rate": 4.175952957363386e-06, "loss": 0.3239, "step": 6168 }, { "epoch": 1.9914847249687235, "grad_norm": 0.3611343345962211, "learning_rate": 4.173560080458846e-06, "loss": 0.3049, "step": 6169 }, { "epoch": 1.9918075789983454, "grad_norm": 0.36725324396728576, "learning_rate": 4.1711676250096635e-06, "loss": 0.2663, "step": 6170 }, { "epoch": 1.9921304330279672, "grad_norm": 0.379869030367077, "learning_rate": 4.1687755913189594e-06, "loss": 0.2884, "step": 6171 }, { "epoch": 1.992453287057589, "grad_norm": 0.4608640156445111, "learning_rate": 4.166383979689802e-06, "loss": 0.2374, "step": 6172 }, { "epoch": 1.992776141087211, "grad_norm": 0.3669098243354426, "learning_rate": 4.163992790425203e-06, "loss": 0.2778, "step": 6173 }, { "epoch": 1.9930989951168328, "grad_norm": 0.41635120328126496, "learning_rate": 4.161602023828124e-06, "loss": 0.358, "step": 6174 }, { "epoch": 1.9934218491464546, "grad_norm": 0.45200985182268477, "learning_rate": 4.159211680201469e-06, "loss": 0.323, "step": 6175 }, { "epoch": 1.9937447031760764, "grad_norm": 0.5167556467679301, "learning_rate": 4.156821759848097e-06, "loss": 0.2988, "step": 6176 }, { "epoch": 1.9940675572056983, "grad_norm": 0.42877465245931656, "learning_rate": 4.1544322630708e-06, "loss": 0.321, "step": 6177 }, { "epoch": 1.9943904112353201, "grad_norm": 0.35434506991704867, "learning_rate": 4.152043190172333e-06, "loss": 0.2833, "step": 6178 }, { "epoch": 1.994713265264942, "grad_norm": 0.4157209252133006, "learning_rate": 4.14965454145538e-06, "loss": 0.3231, "step": 6179 }, { "epoch": 1.9950361192945638, "grad_norm": 0.3460866190237178, "learning_rate": 4.147266317222585e-06, "loss": 0.2939, "step": 6180 }, { "epoch": 1.9953589733241857, "grad_norm": 0.3773901291105398, "learning_rate": 4.144878517776527e-06, "loss": 0.3183, "step": 6181 }, { "epoch": 1.9956818273538075, "grad_norm": 0.42529980667998096, "learning_rate": 4.142491143419746e-06, "loss": 0.3222, "step": 6182 }, { "epoch": 1.9960046813834296, "grad_norm": 0.36521322351841445, "learning_rate": 4.140104194454709e-06, "loss": 0.2973, "step": 6183 }, { "epoch": 1.9963275354130514, "grad_norm": 0.4007176014982859, "learning_rate": 4.137717671183843e-06, "loss": 0.3802, "step": 6184 }, { "epoch": 1.9966503894426733, "grad_norm": 0.3646507539777823, "learning_rate": 4.135331573909523e-06, "loss": 0.2778, "step": 6185 }, { "epoch": 1.996973243472295, "grad_norm": 0.3511124493023407, "learning_rate": 4.132945902934054e-06, "loss": 0.3057, "step": 6186 }, { "epoch": 1.997296097501917, "grad_norm": 0.3686393759709811, "learning_rate": 4.130560658559703e-06, "loss": 0.2639, "step": 6187 }, { "epoch": 1.9976189515315388, "grad_norm": 0.390842793229146, "learning_rate": 4.128175841088681e-06, "loss": 0.3677, "step": 6188 }, { "epoch": 1.9979418055611606, "grad_norm": 0.35886269348592165, "learning_rate": 4.125791450823133e-06, "loss": 0.3616, "step": 6189 }, { "epoch": 1.9982646595907825, "grad_norm": 0.3692545407768144, "learning_rate": 4.123407488065161e-06, "loss": 0.3506, "step": 6190 }, { "epoch": 1.9985875136204043, "grad_norm": 0.32471720404922655, "learning_rate": 4.1210239531168134e-06, "loss": 0.2946, "step": 6191 }, { "epoch": 1.9989103676500264, "grad_norm": 0.4155011188860651, "learning_rate": 4.118640846280075e-06, "loss": 0.3511, "step": 6192 }, { "epoch": 1.9992332216796482, "grad_norm": 0.4164967805993283, "learning_rate": 4.1162581678568884e-06, "loss": 0.2775, "step": 6193 }, { "epoch": 1.99955607570927, "grad_norm": 0.49051249705823224, "learning_rate": 4.113875918149128e-06, "loss": 0.3131, "step": 6194 }, { "epoch": 1.999878929738892, "grad_norm": 0.41789930608712045, "learning_rate": 4.11149409745863e-06, "loss": 0.3338, "step": 6195 }, { "epoch": 2.0, "grad_norm": 0.8168735201431901, "learning_rate": 4.10911270608716e-06, "loss": 0.2634, "step": 6196 }, { "epoch": 2.000322854029622, "grad_norm": 0.3517316778974382, "learning_rate": 4.106731744336444e-06, "loss": 0.2978, "step": 6197 }, { "epoch": 2.0006457080592437, "grad_norm": 0.44503694589542697, "learning_rate": 4.104351212508141e-06, "loss": 0.24, "step": 6198 }, { "epoch": 2.0009685620888655, "grad_norm": 0.39132667226560724, "learning_rate": 4.101971110903864e-06, "loss": 0.284, "step": 6199 }, { "epoch": 2.0012914161184874, "grad_norm": 0.35389429210180123, "learning_rate": 4.0995914398251664e-06, "loss": 0.2103, "step": 6200 }, { "epoch": 2.001614270148109, "grad_norm": 0.4479397930107643, "learning_rate": 4.097212199573551e-06, "loss": 0.2493, "step": 6201 }, { "epoch": 2.001937124177731, "grad_norm": 0.3817254549883359, "learning_rate": 4.0948333904504635e-06, "loss": 0.3097, "step": 6202 }, { "epoch": 2.002259978207353, "grad_norm": 0.3231234803164643, "learning_rate": 4.092455012757299e-06, "loss": 0.2449, "step": 6203 }, { "epoch": 2.0025828322369748, "grad_norm": 0.37914794146117803, "learning_rate": 4.09007706679539e-06, "loss": 0.235, "step": 6204 }, { "epoch": 2.0029056862665966, "grad_norm": 0.4107963934839401, "learning_rate": 4.0876995528660225e-06, "loss": 0.2108, "step": 6205 }, { "epoch": 2.0032285402962184, "grad_norm": 0.4167469815081232, "learning_rate": 4.08532247127042e-06, "loss": 0.2521, "step": 6206 }, { "epoch": 2.0035513943258403, "grad_norm": 0.3663488432969641, "learning_rate": 4.082945822309761e-06, "loss": 0.2929, "step": 6207 }, { "epoch": 2.003874248355462, "grad_norm": 0.4372023267875795, "learning_rate": 4.080569606285158e-06, "loss": 0.2444, "step": 6208 }, { "epoch": 2.004197102385084, "grad_norm": 0.3706692769573013, "learning_rate": 4.078193823497677e-06, "loss": 0.2594, "step": 6209 }, { "epoch": 2.004519956414706, "grad_norm": 0.38093458433943006, "learning_rate": 4.0758184742483305e-06, "loss": 0.2817, "step": 6210 }, { "epoch": 2.0048428104443277, "grad_norm": 0.3756948423557709, "learning_rate": 4.073443558838063e-06, "loss": 0.2393, "step": 6211 }, { "epoch": 2.0051656644739495, "grad_norm": 0.3708755233505259, "learning_rate": 4.0710690775677825e-06, "loss": 0.2374, "step": 6212 }, { "epoch": 2.005488518503572, "grad_norm": 0.5235337991790867, "learning_rate": 4.068695030738324e-06, "loss": 0.2538, "step": 6213 }, { "epoch": 2.0058113725331936, "grad_norm": 0.413387779993061, "learning_rate": 4.066321418650484e-06, "loss": 0.3096, "step": 6214 }, { "epoch": 2.0061342265628155, "grad_norm": 0.36449160085145765, "learning_rate": 4.063948241604988e-06, "loss": 0.2645, "step": 6215 }, { "epoch": 2.0064570805924373, "grad_norm": 0.4649757183838215, "learning_rate": 4.061575499902518e-06, "loss": 0.2645, "step": 6216 }, { "epoch": 2.006779934622059, "grad_norm": 0.37849224374378193, "learning_rate": 4.0592031938436966e-06, "loss": 0.3441, "step": 6217 }, { "epoch": 2.007102788651681, "grad_norm": 0.3439284596890928, "learning_rate": 4.056831323729097e-06, "loss": 0.2687, "step": 6218 }, { "epoch": 2.007425642681303, "grad_norm": 0.33854949152074404, "learning_rate": 4.054459889859223e-06, "loss": 0.2226, "step": 6219 }, { "epoch": 2.0077484967109247, "grad_norm": 0.35630327386578053, "learning_rate": 4.052088892534539e-06, "loss": 0.2283, "step": 6220 }, { "epoch": 2.0080713507405465, "grad_norm": 0.41105402000095603, "learning_rate": 4.049718332055442e-06, "loss": 0.2889, "step": 6221 }, { "epoch": 2.0083942047701684, "grad_norm": 0.36075250605683884, "learning_rate": 4.047348208722283e-06, "loss": 0.2438, "step": 6222 }, { "epoch": 2.0087170587997902, "grad_norm": 0.4413946783939728, "learning_rate": 4.044978522835349e-06, "loss": 0.2315, "step": 6223 }, { "epoch": 2.009039912829412, "grad_norm": 0.39322745937348064, "learning_rate": 4.0426092746948826e-06, "loss": 0.247, "step": 6224 }, { "epoch": 2.009362766859034, "grad_norm": 0.48218858143038634, "learning_rate": 4.0402404646010555e-06, "loss": 0.3264, "step": 6225 }, { "epoch": 2.0096856208886558, "grad_norm": 0.4701419814124609, "learning_rate": 4.037872092853997e-06, "loss": 0.2218, "step": 6226 }, { "epoch": 2.0100084749182776, "grad_norm": 0.37491812733815455, "learning_rate": 4.03550415975378e-06, "loss": 0.2399, "step": 6227 }, { "epoch": 2.0103313289478995, "grad_norm": 0.3245184688528673, "learning_rate": 4.033136665600412e-06, "loss": 0.2537, "step": 6228 }, { "epoch": 2.0106541829775213, "grad_norm": 0.40176497910611114, "learning_rate": 4.030769610693858e-06, "loss": 0.241, "step": 6229 }, { "epoch": 2.010977037007143, "grad_norm": 0.37361025584485447, "learning_rate": 4.028402995334013e-06, "loss": 0.2732, "step": 6230 }, { "epoch": 2.011299891036765, "grad_norm": 0.32679060766980833, "learning_rate": 4.02603681982073e-06, "loss": 0.2898, "step": 6231 }, { "epoch": 2.011622745066387, "grad_norm": 0.3197104556309193, "learning_rate": 4.023671084453797e-06, "loss": 0.2187, "step": 6232 }, { "epoch": 2.0119455990960087, "grad_norm": 0.4364819959584814, "learning_rate": 4.02130578953295e-06, "loss": 0.2501, "step": 6233 }, { "epoch": 2.0122684531256305, "grad_norm": 0.36757648617339783, "learning_rate": 4.018940935357869e-06, "loss": 0.23, "step": 6234 }, { "epoch": 2.0125913071552524, "grad_norm": 0.42873475370078923, "learning_rate": 4.01657652222818e-06, "loss": 0.2339, "step": 6235 }, { "epoch": 2.012914161184874, "grad_norm": 0.3209495033822786, "learning_rate": 4.014212550443445e-06, "loss": 0.2735, "step": 6236 }, { "epoch": 2.013237015214496, "grad_norm": 0.4283523589728924, "learning_rate": 4.011849020303184e-06, "loss": 0.2461, "step": 6237 }, { "epoch": 2.013559869244118, "grad_norm": 0.3919056266079639, "learning_rate": 4.009485932106846e-06, "loss": 0.301, "step": 6238 }, { "epoch": 2.0138827232737397, "grad_norm": 0.330143912326063, "learning_rate": 4.007123286153836e-06, "loss": 0.2424, "step": 6239 }, { "epoch": 2.0142055773033616, "grad_norm": 0.4199192308808658, "learning_rate": 4.004761082743493e-06, "loss": 0.2396, "step": 6240 }, { "epoch": 2.0145284313329834, "grad_norm": 0.4259870918710433, "learning_rate": 4.0023993221751095e-06, "loss": 0.2772, "step": 6241 }, { "epoch": 2.0148512853626053, "grad_norm": 0.34750185287784596, "learning_rate": 4.000038004747913e-06, "loss": 0.2484, "step": 6242 }, { "epoch": 2.015174139392227, "grad_norm": 0.42102666355308166, "learning_rate": 3.997677130761085e-06, "loss": 0.287, "step": 6243 }, { "epoch": 2.015496993421849, "grad_norm": 0.419244692594865, "learning_rate": 3.995316700513739e-06, "loss": 0.2548, "step": 6244 }, { "epoch": 2.015819847451471, "grad_norm": 0.3966027186705409, "learning_rate": 3.9929567143049395e-06, "loss": 0.2284, "step": 6245 }, { "epoch": 2.0161427014810926, "grad_norm": 0.36579095038248116, "learning_rate": 3.9905971724336985e-06, "loss": 0.2855, "step": 6246 }, { "epoch": 2.016465555510715, "grad_norm": 0.36096929170269837, "learning_rate": 3.98823807519896e-06, "loss": 0.2974, "step": 6247 }, { "epoch": 2.016788409540337, "grad_norm": 0.3578235365356579, "learning_rate": 3.985879422899621e-06, "loss": 0.2743, "step": 6248 }, { "epoch": 2.0171112635699586, "grad_norm": 0.44003464147111243, "learning_rate": 3.98352121583452e-06, "loss": 0.2401, "step": 6249 }, { "epoch": 2.0174341175995805, "grad_norm": 0.4045116206310484, "learning_rate": 3.981163454302442e-06, "loss": 0.254, "step": 6250 }, { "epoch": 2.0177569716292023, "grad_norm": 0.40166746408181264, "learning_rate": 3.978806138602105e-06, "loss": 0.3411, "step": 6251 }, { "epoch": 2.018079825658824, "grad_norm": 0.4237415234701937, "learning_rate": 3.976449269032185e-06, "loss": 0.3187, "step": 6252 }, { "epoch": 2.018402679688446, "grad_norm": 0.4232029329663933, "learning_rate": 3.974092845891286e-06, "loss": 0.2638, "step": 6253 }, { "epoch": 2.018725533718068, "grad_norm": 0.9025722002989033, "learning_rate": 3.9717368694779725e-06, "loss": 0.2432, "step": 6254 }, { "epoch": 2.0190483877476897, "grad_norm": 0.368620011702583, "learning_rate": 3.969381340090736e-06, "loss": 0.2479, "step": 6255 }, { "epoch": 2.0193712417773115, "grad_norm": 0.37088232404094607, "learning_rate": 3.967026258028023e-06, "loss": 0.2461, "step": 6256 }, { "epoch": 2.0196940958069334, "grad_norm": 0.3638960323370632, "learning_rate": 3.964671623588216e-06, "loss": 0.2278, "step": 6257 }, { "epoch": 2.020016949836555, "grad_norm": 0.38111540345141565, "learning_rate": 3.962317437069649e-06, "loss": 0.2604, "step": 6258 }, { "epoch": 2.020339803866177, "grad_norm": 0.45546342186630023, "learning_rate": 3.959963698770588e-06, "loss": 0.2668, "step": 6259 }, { "epoch": 2.020662657895799, "grad_norm": 0.42958835044633287, "learning_rate": 3.9576104089892545e-06, "loss": 0.2679, "step": 6260 }, { "epoch": 2.0209855119254208, "grad_norm": 0.4032212461067635, "learning_rate": 3.955257568023801e-06, "loss": 0.2585, "step": 6261 }, { "epoch": 2.0213083659550426, "grad_norm": 0.41222668440844945, "learning_rate": 3.952905176172332e-06, "loss": 0.2317, "step": 6262 }, { "epoch": 2.0216312199846644, "grad_norm": 0.42531041749016835, "learning_rate": 3.950553233732893e-06, "loss": 0.2605, "step": 6263 }, { "epoch": 2.0219540740142863, "grad_norm": 0.4128994609282213, "learning_rate": 3.948201741003474e-06, "loss": 0.2595, "step": 6264 }, { "epoch": 2.022276928043908, "grad_norm": 0.35155000203306785, "learning_rate": 3.945850698282e-06, "loss": 0.2507, "step": 6265 }, { "epoch": 2.02259978207353, "grad_norm": 0.3393804447021201, "learning_rate": 3.943500105866349e-06, "loss": 0.2772, "step": 6266 }, { "epoch": 2.022922636103152, "grad_norm": 0.4053129956173491, "learning_rate": 3.941149964054339e-06, "loss": 0.3321, "step": 6267 }, { "epoch": 2.0232454901327737, "grad_norm": 0.6954312678471312, "learning_rate": 3.938800273143725e-06, "loss": 0.2831, "step": 6268 }, { "epoch": 2.0235683441623955, "grad_norm": 0.4063458261941138, "learning_rate": 3.9364510334322156e-06, "loss": 0.3121, "step": 6269 }, { "epoch": 2.0238911981920173, "grad_norm": 0.4436025692407472, "learning_rate": 3.93410224521745e-06, "loss": 0.2105, "step": 6270 }, { "epoch": 2.024214052221639, "grad_norm": 0.43138579921766906, "learning_rate": 3.931753908797023e-06, "loss": 0.2145, "step": 6271 }, { "epoch": 2.024536906251261, "grad_norm": 0.351721002620432, "learning_rate": 3.929406024468457e-06, "loss": 0.2797, "step": 6272 }, { "epoch": 2.024859760280883, "grad_norm": 0.40028228977694374, "learning_rate": 3.927058592529236e-06, "loss": 0.2919, "step": 6273 }, { "epoch": 2.0251826143105047, "grad_norm": 0.4369642684759009, "learning_rate": 3.924711613276767e-06, "loss": 0.2653, "step": 6274 }, { "epoch": 2.0255054683401266, "grad_norm": 0.40734244490654314, "learning_rate": 3.922365087008416e-06, "loss": 0.2904, "step": 6275 }, { "epoch": 2.0258283223697484, "grad_norm": 0.29955736189789844, "learning_rate": 3.9200190140214785e-06, "loss": 0.2236, "step": 6276 }, { "epoch": 2.0261511763993703, "grad_norm": 0.3747053940683399, "learning_rate": 3.917673394613206e-06, "loss": 0.2605, "step": 6277 }, { "epoch": 2.026474030428992, "grad_norm": 0.3918827504038926, "learning_rate": 3.9153282290807775e-06, "loss": 0.2489, "step": 6278 }, { "epoch": 2.026796884458614, "grad_norm": 0.35915245329335244, "learning_rate": 3.912983517721327e-06, "loss": 0.2461, "step": 6279 }, { "epoch": 2.027119738488236, "grad_norm": 0.4063131346293011, "learning_rate": 3.9106392608319236e-06, "loss": 0.2962, "step": 6280 }, { "epoch": 2.027442592517858, "grad_norm": 0.40801827954248204, "learning_rate": 3.908295458709587e-06, "loss": 0.2686, "step": 6281 }, { "epoch": 2.02776544654748, "grad_norm": 0.34765750651004135, "learning_rate": 3.905952111651267e-06, "loss": 0.2772, "step": 6282 }, { "epoch": 2.0280883005771018, "grad_norm": 0.37163742234733, "learning_rate": 3.903609219953864e-06, "loss": 0.2401, "step": 6283 }, { "epoch": 2.0284111546067236, "grad_norm": 0.35670645099559545, "learning_rate": 3.901266783914224e-06, "loss": 0.2786, "step": 6284 }, { "epoch": 2.0287340086363455, "grad_norm": 0.3603864934418854, "learning_rate": 3.898924803829124e-06, "loss": 0.2327, "step": 6285 }, { "epoch": 2.0290568626659673, "grad_norm": 0.4115563066412632, "learning_rate": 3.896583279995295e-06, "loss": 0.2681, "step": 6286 }, { "epoch": 2.029379716695589, "grad_norm": 0.34679279555913384, "learning_rate": 3.8942422127094e-06, "loss": 0.2595, "step": 6287 }, { "epoch": 2.029702570725211, "grad_norm": 0.39900257230525876, "learning_rate": 3.891901602268053e-06, "loss": 0.2544, "step": 6288 }, { "epoch": 2.030025424754833, "grad_norm": 0.3501838411252907, "learning_rate": 3.889561448967802e-06, "loss": 0.2915, "step": 6289 }, { "epoch": 2.0303482787844547, "grad_norm": 0.3804221101163068, "learning_rate": 3.887221753105147e-06, "loss": 0.2903, "step": 6290 }, { "epoch": 2.0306711328140765, "grad_norm": 0.3976741938035202, "learning_rate": 3.884882514976517e-06, "loss": 0.3317, "step": 6291 }, { "epoch": 2.0309939868436984, "grad_norm": 0.365831486936568, "learning_rate": 3.882543734878297e-06, "loss": 0.2523, "step": 6292 }, { "epoch": 2.03131684087332, "grad_norm": 0.3529082087722924, "learning_rate": 3.880205413106801e-06, "loss": 0.2703, "step": 6293 }, { "epoch": 2.031639694902942, "grad_norm": 0.3588479235347632, "learning_rate": 3.8778675499582955e-06, "loss": 0.2927, "step": 6294 }, { "epoch": 2.031962548932564, "grad_norm": 0.3815321117735894, "learning_rate": 3.875530145728983e-06, "loss": 0.2215, "step": 6295 }, { "epoch": 2.0322854029621857, "grad_norm": 0.369204078219184, "learning_rate": 3.873193200715013e-06, "loss": 0.2598, "step": 6296 }, { "epoch": 2.0326082569918076, "grad_norm": 0.39255467379330045, "learning_rate": 3.870856715212467e-06, "loss": 0.2761, "step": 6297 }, { "epoch": 2.0329311110214294, "grad_norm": 0.38189275762972474, "learning_rate": 3.86852068951738e-06, "loss": 0.2946, "step": 6298 }, { "epoch": 2.0332539650510513, "grad_norm": 0.385694354613102, "learning_rate": 3.866185123925718e-06, "loss": 0.2596, "step": 6299 }, { "epoch": 2.033576819080673, "grad_norm": 0.33701038534999717, "learning_rate": 3.863850018733396e-06, "loss": 0.2558, "step": 6300 }, { "epoch": 2.033899673110295, "grad_norm": 0.41499830854017, "learning_rate": 3.861515374236273e-06, "loss": 0.2913, "step": 6301 }, { "epoch": 2.034222527139917, "grad_norm": 0.3325722358883631, "learning_rate": 3.859181190730139e-06, "loss": 0.2491, "step": 6302 }, { "epoch": 2.0345453811695386, "grad_norm": 0.3909712118587173, "learning_rate": 3.856847468510738e-06, "loss": 0.3148, "step": 6303 }, { "epoch": 2.0348682351991605, "grad_norm": 0.39978871777324476, "learning_rate": 3.854514207873743e-06, "loss": 0.2447, "step": 6304 }, { "epoch": 2.0351910892287823, "grad_norm": 0.47754948198323477, "learning_rate": 3.852181409114779e-06, "loss": 0.3091, "step": 6305 }, { "epoch": 2.035513943258404, "grad_norm": 0.3506893059087084, "learning_rate": 3.849849072529404e-06, "loss": 0.1943, "step": 6306 }, { "epoch": 2.035836797288026, "grad_norm": 0.409306389591793, "learning_rate": 3.847517198413129e-06, "loss": 0.218, "step": 6307 }, { "epoch": 2.036159651317648, "grad_norm": 0.4179007631567929, "learning_rate": 3.845185787061393e-06, "loss": 0.2514, "step": 6308 }, { "epoch": 2.0364825053472697, "grad_norm": 0.39356254110777966, "learning_rate": 3.842854838769587e-06, "loss": 0.2777, "step": 6309 }, { "epoch": 2.0368053593768916, "grad_norm": 0.4023233956843697, "learning_rate": 3.840524353833034e-06, "loss": 0.2802, "step": 6310 }, { "epoch": 2.0371282134065134, "grad_norm": 0.3113069709502987, "learning_rate": 3.838194332547005e-06, "loss": 0.2448, "step": 6311 }, { "epoch": 2.0374510674361352, "grad_norm": 0.3834799801154306, "learning_rate": 3.835864775206712e-06, "loss": 0.2201, "step": 6312 }, { "epoch": 2.037773921465757, "grad_norm": 0.50585029740273, "learning_rate": 3.83353568210731e-06, "loss": 0.2258, "step": 6313 }, { "epoch": 2.0380967754953794, "grad_norm": 0.38058856076060044, "learning_rate": 3.831207053543883e-06, "loss": 0.273, "step": 6314 }, { "epoch": 2.038419629525001, "grad_norm": 0.37336910426930625, "learning_rate": 3.828878889811475e-06, "loss": 0.27, "step": 6315 }, { "epoch": 2.038742483554623, "grad_norm": 0.34931278360967544, "learning_rate": 3.826551191205053e-06, "loss": 0.2187, "step": 6316 }, { "epoch": 2.039065337584245, "grad_norm": 0.5257352780794992, "learning_rate": 3.824223958019538e-06, "loss": 0.2621, "step": 6317 }, { "epoch": 2.0393881916138668, "grad_norm": 0.29126422162526017, "learning_rate": 3.821897190549784e-06, "loss": 0.221, "step": 6318 }, { "epoch": 2.0397110456434886, "grad_norm": 0.36252444959657865, "learning_rate": 3.819570889090591e-06, "loss": 0.2595, "step": 6319 }, { "epoch": 2.0400338996731104, "grad_norm": 0.4387038833204259, "learning_rate": 3.8172450539367e-06, "loss": 0.2455, "step": 6320 }, { "epoch": 2.0403567537027323, "grad_norm": 0.39416945641024514, "learning_rate": 3.8149196853827876e-06, "loss": 0.2496, "step": 6321 }, { "epoch": 2.040679607732354, "grad_norm": 0.34243635500005426, "learning_rate": 3.8125947837234783e-06, "loss": 0.222, "step": 6322 }, { "epoch": 2.041002461761976, "grad_norm": 0.36277937752613276, "learning_rate": 3.810270349253329e-06, "loss": 0.259, "step": 6323 }, { "epoch": 2.041325315791598, "grad_norm": 0.3599679644055669, "learning_rate": 3.8079463822668487e-06, "loss": 0.2932, "step": 6324 }, { "epoch": 2.0416481698212197, "grad_norm": 0.4315088633058175, "learning_rate": 3.805622883058475e-06, "loss": 0.2482, "step": 6325 }, { "epoch": 2.0419710238508415, "grad_norm": 0.36037993980174915, "learning_rate": 3.8032998519225943e-06, "loss": 0.297, "step": 6326 }, { "epoch": 2.0422938778804633, "grad_norm": 0.35561677591281754, "learning_rate": 3.8009772891535317e-06, "loss": 0.2659, "step": 6327 }, { "epoch": 2.042616731910085, "grad_norm": 0.42365467503839116, "learning_rate": 3.7986551950455563e-06, "loss": 0.2361, "step": 6328 }, { "epoch": 2.042939585939707, "grad_norm": 0.3066014784405688, "learning_rate": 3.796333569892866e-06, "loss": 0.2173, "step": 6329 }, { "epoch": 2.043262439969329, "grad_norm": 0.3150613514689327, "learning_rate": 3.7940124139896166e-06, "loss": 0.2083, "step": 6330 }, { "epoch": 2.0435852939989507, "grad_norm": 0.37732142636253213, "learning_rate": 3.791691727629887e-06, "loss": 0.2722, "step": 6331 }, { "epoch": 2.0439081480285726, "grad_norm": 0.3637619377235598, "learning_rate": 3.7893715111077123e-06, "loss": 0.2464, "step": 6332 }, { "epoch": 2.0442310020581944, "grad_norm": 0.33383689406947326, "learning_rate": 3.7870517647170537e-06, "loss": 0.2236, "step": 6333 }, { "epoch": 2.0445538560878163, "grad_norm": 0.31819221527246927, "learning_rate": 3.784732488751828e-06, "loss": 0.2289, "step": 6334 }, { "epoch": 2.044876710117438, "grad_norm": 0.41277849940315187, "learning_rate": 3.7824136835058758e-06, "loss": 0.2275, "step": 6335 }, { "epoch": 2.04519956414706, "grad_norm": 0.3555112241686159, "learning_rate": 3.780095349272993e-06, "loss": 0.2746, "step": 6336 }, { "epoch": 2.045522418176682, "grad_norm": 0.4305457672394699, "learning_rate": 3.777777486346904e-06, "loss": 0.2485, "step": 6337 }, { "epoch": 2.0458452722063036, "grad_norm": 0.360554057787865, "learning_rate": 3.7754600950212824e-06, "loss": 0.2468, "step": 6338 }, { "epoch": 2.0461681262359255, "grad_norm": 0.35173760514915414, "learning_rate": 3.77314317558974e-06, "loss": 0.2671, "step": 6339 }, { "epoch": 2.0464909802655473, "grad_norm": 0.39837325989596906, "learning_rate": 3.770826728345823e-06, "loss": 0.249, "step": 6340 }, { "epoch": 2.046813834295169, "grad_norm": 0.44970926927403976, "learning_rate": 3.7685107535830237e-06, "loss": 0.248, "step": 6341 }, { "epoch": 2.047136688324791, "grad_norm": 0.371179757109403, "learning_rate": 3.766195251594776e-06, "loss": 0.2953, "step": 6342 }, { "epoch": 2.047459542354413, "grad_norm": 0.4190060007645388, "learning_rate": 3.763880222674445e-06, "loss": 0.247, "step": 6343 }, { "epoch": 2.0477823963840347, "grad_norm": 0.4219623982608124, "learning_rate": 3.761565667115345e-06, "loss": 0.2341, "step": 6344 }, { "epoch": 2.0481052504136565, "grad_norm": 0.4107016646436235, "learning_rate": 3.75925158521073e-06, "loss": 0.2372, "step": 6345 }, { "epoch": 2.0484281044432784, "grad_norm": 0.3849181041684448, "learning_rate": 3.7569379772537855e-06, "loss": 0.2404, "step": 6346 }, { "epoch": 2.0487509584729002, "grad_norm": 0.38673269157445456, "learning_rate": 3.7546248435376466e-06, "loss": 0.2409, "step": 6347 }, { "epoch": 2.049073812502522, "grad_norm": 0.4370430484467494, "learning_rate": 3.7523121843553797e-06, "loss": 0.2774, "step": 6348 }, { "epoch": 2.0493966665321444, "grad_norm": 0.4095163863649772, "learning_rate": 3.750000000000002e-06, "loss": 0.2661, "step": 6349 }, { "epoch": 2.049719520561766, "grad_norm": 0.30836646618642216, "learning_rate": 3.7476882907644563e-06, "loss": 0.2505, "step": 6350 }, { "epoch": 2.050042374591388, "grad_norm": 0.37619312499452523, "learning_rate": 3.74537705694164e-06, "loss": 0.265, "step": 6351 }, { "epoch": 2.05036522862101, "grad_norm": 0.38223584497702506, "learning_rate": 3.743066298824377e-06, "loss": 0.3174, "step": 6352 }, { "epoch": 2.0506880826506317, "grad_norm": 0.38627679544904486, "learning_rate": 3.740756016705444e-06, "loss": 0.271, "step": 6353 }, { "epoch": 2.0510109366802536, "grad_norm": 0.5698812465767727, "learning_rate": 3.7384462108775437e-06, "loss": 0.1998, "step": 6354 }, { "epoch": 2.0513337907098754, "grad_norm": 0.3406280479337855, "learning_rate": 3.736136881633331e-06, "loss": 0.237, "step": 6355 }, { "epoch": 2.0516566447394973, "grad_norm": 0.32285851166724494, "learning_rate": 3.73382802926539e-06, "loss": 0.209, "step": 6356 }, { "epoch": 2.051979498769119, "grad_norm": 0.36880831063099234, "learning_rate": 3.7315196540662516e-06, "loss": 0.313, "step": 6357 }, { "epoch": 2.052302352798741, "grad_norm": 0.34746698094745615, "learning_rate": 3.729211756328383e-06, "loss": 0.2322, "step": 6358 }, { "epoch": 2.052625206828363, "grad_norm": 0.41028803353002535, "learning_rate": 3.7269043363441917e-06, "loss": 0.2465, "step": 6359 }, { "epoch": 2.0529480608579846, "grad_norm": 0.3658362761151128, "learning_rate": 3.724597394406029e-06, "loss": 0.255, "step": 6360 }, { "epoch": 2.0532709148876065, "grad_norm": 0.3648093358158705, "learning_rate": 3.722290930806173e-06, "loss": 0.2656, "step": 6361 }, { "epoch": 2.0535937689172283, "grad_norm": 0.36775433828546494, "learning_rate": 3.7199849458368583e-06, "loss": 0.2629, "step": 6362 }, { "epoch": 2.05391662294685, "grad_norm": 0.35931633472187124, "learning_rate": 3.717679439790242e-06, "loss": 0.2319, "step": 6363 }, { "epoch": 2.054239476976472, "grad_norm": 0.36451910935657117, "learning_rate": 3.715374412958435e-06, "loss": 0.2389, "step": 6364 }, { "epoch": 2.054562331006094, "grad_norm": 0.3556237171173618, "learning_rate": 3.713069865633476e-06, "loss": 0.2542, "step": 6365 }, { "epoch": 2.0548851850357157, "grad_norm": 0.4161742741972861, "learning_rate": 3.7107657981073516e-06, "loss": 0.2442, "step": 6366 }, { "epoch": 2.0552080390653376, "grad_norm": 0.4090848854228829, "learning_rate": 3.7084622106719814e-06, "loss": 0.2794, "step": 6367 }, { "epoch": 2.0555308930949594, "grad_norm": 0.34668677956095056, "learning_rate": 3.7061591036192295e-06, "loss": 0.2423, "step": 6368 }, { "epoch": 2.0558537471245812, "grad_norm": 0.43347830200049825, "learning_rate": 3.703856477240892e-06, "loss": 0.2329, "step": 6369 }, { "epoch": 2.056176601154203, "grad_norm": 0.42510985575141946, "learning_rate": 3.701554331828714e-06, "loss": 0.3352, "step": 6370 }, { "epoch": 2.056499455183825, "grad_norm": 0.4532198215940069, "learning_rate": 3.699252667674369e-06, "loss": 0.2728, "step": 6371 }, { "epoch": 2.0568223092134468, "grad_norm": 0.40581008354628645, "learning_rate": 3.696951485069476e-06, "loss": 0.3093, "step": 6372 }, { "epoch": 2.0571451632430686, "grad_norm": 0.4006721465126181, "learning_rate": 3.6946507843055936e-06, "loss": 0.2524, "step": 6373 }, { "epoch": 2.0574680172726905, "grad_norm": 0.4286112194797119, "learning_rate": 3.6923505656742197e-06, "loss": 0.3007, "step": 6374 }, { "epoch": 2.0577908713023123, "grad_norm": 0.3940765475250732, "learning_rate": 3.6900508294667824e-06, "loss": 0.295, "step": 6375 }, { "epoch": 2.058113725331934, "grad_norm": 0.443799052829658, "learning_rate": 3.6877515759746587e-06, "loss": 0.2477, "step": 6376 }, { "epoch": 2.058436579361556, "grad_norm": 0.4260472521694487, "learning_rate": 3.6854528054891627e-06, "loss": 0.3115, "step": 6377 }, { "epoch": 2.058759433391178, "grad_norm": 0.3690663202160674, "learning_rate": 3.6831545183015413e-06, "loss": 0.2911, "step": 6378 }, { "epoch": 2.0590822874207997, "grad_norm": 0.3776377895343396, "learning_rate": 3.6808567147029887e-06, "loss": 0.3077, "step": 6379 }, { "epoch": 2.0594051414504215, "grad_norm": 0.3595209752797616, "learning_rate": 3.6785593949846276e-06, "loss": 0.2315, "step": 6380 }, { "epoch": 2.0597279954800434, "grad_norm": 0.604184062529061, "learning_rate": 3.6762625594375324e-06, "loss": 0.2632, "step": 6381 }, { "epoch": 2.0600508495096657, "grad_norm": 0.38642573287452353, "learning_rate": 3.673966208352702e-06, "loss": 0.3219, "step": 6382 }, { "epoch": 2.0603737035392875, "grad_norm": 0.3917582176434674, "learning_rate": 3.6716703420210883e-06, "loss": 0.2561, "step": 6383 }, { "epoch": 2.0606965575689093, "grad_norm": 0.3715137732155293, "learning_rate": 3.6693749607335663e-06, "loss": 0.2821, "step": 6384 }, { "epoch": 2.061019411598531, "grad_norm": 0.40348347860643724, "learning_rate": 3.6670800647809645e-06, "loss": 0.2993, "step": 6385 }, { "epoch": 2.061342265628153, "grad_norm": 0.36398235546327157, "learning_rate": 3.664785654454038e-06, "loss": 0.2663, "step": 6386 }, { "epoch": 2.061665119657775, "grad_norm": 0.4132188667757173, "learning_rate": 3.6624917300434875e-06, "loss": 0.295, "step": 6387 }, { "epoch": 2.0619879736873967, "grad_norm": 0.4032632432514926, "learning_rate": 3.660198291839953e-06, "loss": 0.2168, "step": 6388 }, { "epoch": 2.0623108277170186, "grad_norm": 0.35360828486563506, "learning_rate": 3.6579053401340045e-06, "loss": 0.2151, "step": 6389 }, { "epoch": 2.0626336817466404, "grad_norm": 0.31927971440894914, "learning_rate": 3.6556128752161577e-06, "loss": 0.2477, "step": 6390 }, { "epoch": 2.0629565357762623, "grad_norm": 0.371991306973343, "learning_rate": 3.6533208973768692e-06, "loss": 0.2922, "step": 6391 }, { "epoch": 2.063279389805884, "grad_norm": 0.3883292468685179, "learning_rate": 3.6510294069065235e-06, "loss": 0.2589, "step": 6392 }, { "epoch": 2.063602243835506, "grad_norm": 0.40694494797360037, "learning_rate": 3.6487384040954503e-06, "loss": 0.2805, "step": 6393 }, { "epoch": 2.063925097865128, "grad_norm": 0.44471398549663205, "learning_rate": 3.6464478892339216e-06, "loss": 0.2454, "step": 6394 }, { "epoch": 2.0642479518947496, "grad_norm": 0.3959199958470189, "learning_rate": 3.644157862612136e-06, "loss": 0.2431, "step": 6395 }, { "epoch": 2.0645708059243715, "grad_norm": 0.42948205211090834, "learning_rate": 3.6418683245202414e-06, "loss": 0.2238, "step": 6396 }, { "epoch": 2.0648936599539933, "grad_norm": 0.42254849682860723, "learning_rate": 3.6395792752483146e-06, "loss": 0.2278, "step": 6397 }, { "epoch": 2.065216513983615, "grad_norm": 0.4435347778831847, "learning_rate": 3.6372907150863806e-06, "loss": 0.2858, "step": 6398 }, { "epoch": 2.065539368013237, "grad_norm": 0.34872908664829855, "learning_rate": 3.6350026443243905e-06, "loss": 0.2622, "step": 6399 }, { "epoch": 2.065862222042859, "grad_norm": 0.3811084560126708, "learning_rate": 3.6327150632522456e-06, "loss": 0.2614, "step": 6400 }, { "epoch": 2.0661850760724807, "grad_norm": 0.43685147633694427, "learning_rate": 3.6304279721597744e-06, "loss": 0.2618, "step": 6401 }, { "epoch": 2.0665079301021025, "grad_norm": 0.3528590503482391, "learning_rate": 3.628141371336753e-06, "loss": 0.2757, "step": 6402 }, { "epoch": 2.0668307841317244, "grad_norm": 0.3295068507459964, "learning_rate": 3.625855261072885e-06, "loss": 0.2815, "step": 6403 }, { "epoch": 2.0671536381613462, "grad_norm": 0.33604441687006564, "learning_rate": 3.6235696416578204e-06, "loss": 0.2719, "step": 6404 }, { "epoch": 2.067476492190968, "grad_norm": 0.3962410146066375, "learning_rate": 3.621284513381145e-06, "loss": 0.2407, "step": 6405 }, { "epoch": 2.06779934622059, "grad_norm": 0.3180863383377985, "learning_rate": 3.6189998765323818e-06, "loss": 0.2501, "step": 6406 }, { "epoch": 2.0681222002502118, "grad_norm": 0.4539746953063594, "learning_rate": 3.616715731400988e-06, "loss": 0.2463, "step": 6407 }, { "epoch": 2.0684450542798336, "grad_norm": 0.3858190286657735, "learning_rate": 3.6144320782763654e-06, "loss": 0.3113, "step": 6408 }, { "epoch": 2.0687679083094554, "grad_norm": 0.3466862683516217, "learning_rate": 3.612148917447846e-06, "loss": 0.1963, "step": 6409 }, { "epoch": 2.0690907623390773, "grad_norm": 0.38995990168547723, "learning_rate": 3.6098662492047073e-06, "loss": 0.2339, "step": 6410 }, { "epoch": 2.069413616368699, "grad_norm": 0.40213254257456227, "learning_rate": 3.6075840738361546e-06, "loss": 0.2582, "step": 6411 }, { "epoch": 2.069736470398321, "grad_norm": 0.32271083127125805, "learning_rate": 3.6053023916313393e-06, "loss": 0.2731, "step": 6412 }, { "epoch": 2.070059324427943, "grad_norm": 0.3383384137800728, "learning_rate": 3.603021202879351e-06, "loss": 0.2992, "step": 6413 }, { "epoch": 2.0703821784575647, "grad_norm": 0.37065621464473314, "learning_rate": 3.6007405078692068e-06, "loss": 0.2728, "step": 6414 }, { "epoch": 2.0707050324871865, "grad_norm": 0.385845959080322, "learning_rate": 3.598460306889874e-06, "loss": 0.3113, "step": 6415 }, { "epoch": 2.071027886516809, "grad_norm": 0.38924024101733307, "learning_rate": 3.5961806002302442e-06, "loss": 0.2615, "step": 6416 }, { "epoch": 2.0713507405464306, "grad_norm": 0.399411435067338, "learning_rate": 3.5939013881791593e-06, "loss": 0.3047, "step": 6417 }, { "epoch": 2.0716735945760525, "grad_norm": 0.37582029503758935, "learning_rate": 3.591622671025387e-06, "loss": 0.2847, "step": 6418 }, { "epoch": 2.0719964486056743, "grad_norm": 0.3858635673199745, "learning_rate": 3.58934444905764e-06, "loss": 0.3128, "step": 6419 }, { "epoch": 2.072319302635296, "grad_norm": 0.4130705050018272, "learning_rate": 3.5870667225645688e-06, "loss": 0.1966, "step": 6420 }, { "epoch": 2.072642156664918, "grad_norm": 0.31707693580390606, "learning_rate": 3.5847894918347523e-06, "loss": 0.232, "step": 6421 }, { "epoch": 2.07296501069454, "grad_norm": 0.3647595498636123, "learning_rate": 3.5825127571567164e-06, "loss": 0.2199, "step": 6422 }, { "epoch": 2.0732878647241617, "grad_norm": 0.35742549113302996, "learning_rate": 3.5802365188189214e-06, "loss": 0.2321, "step": 6423 }, { "epoch": 2.0736107187537836, "grad_norm": 0.4190377715354331, "learning_rate": 3.577960777109759e-06, "loss": 0.265, "step": 6424 }, { "epoch": 2.0739335727834054, "grad_norm": 0.3431763140078803, "learning_rate": 3.5756855323175687e-06, "loss": 0.2729, "step": 6425 }, { "epoch": 2.0742564268130272, "grad_norm": 0.365724360744564, "learning_rate": 3.5734107847306135e-06, "loss": 0.2688, "step": 6426 }, { "epoch": 2.074579280842649, "grad_norm": 0.45273440939821114, "learning_rate": 3.571136534637108e-06, "loss": 0.2503, "step": 6427 }, { "epoch": 2.074902134872271, "grad_norm": 0.3172171311600849, "learning_rate": 3.568862782325191e-06, "loss": 0.2761, "step": 6428 }, { "epoch": 2.0752249889018928, "grad_norm": 0.3805773010116699, "learning_rate": 3.5665895280829474e-06, "loss": 0.2421, "step": 6429 }, { "epoch": 2.0755478429315146, "grad_norm": 0.33554760273741935, "learning_rate": 3.564316772198392e-06, "loss": 0.2768, "step": 6430 }, { "epoch": 2.0758706969611365, "grad_norm": 0.38482843606805983, "learning_rate": 3.562044514959482e-06, "loss": 0.2721, "step": 6431 }, { "epoch": 2.0761935509907583, "grad_norm": 0.41893094232657246, "learning_rate": 3.5597727566541115e-06, "loss": 0.2374, "step": 6432 }, { "epoch": 2.07651640502038, "grad_norm": 0.4278301633056026, "learning_rate": 3.557501497570104e-06, "loss": 0.2431, "step": 6433 }, { "epoch": 2.076839259050002, "grad_norm": 0.3727130445843983, "learning_rate": 3.5552307379952305e-06, "loss": 0.2556, "step": 6434 }, { "epoch": 2.077162113079624, "grad_norm": 0.34588798396775805, "learning_rate": 3.552960478217187e-06, "loss": 0.2696, "step": 6435 }, { "epoch": 2.0774849671092457, "grad_norm": 0.40022693133327863, "learning_rate": 3.550690718523616e-06, "loss": 0.2698, "step": 6436 }, { "epoch": 2.0778078211388675, "grad_norm": 0.43655880125580443, "learning_rate": 3.548421459202092e-06, "loss": 0.2782, "step": 6437 }, { "epoch": 2.0781306751684894, "grad_norm": 0.3671865943489859, "learning_rate": 3.546152700540131e-06, "loss": 0.2658, "step": 6438 }, { "epoch": 2.078453529198111, "grad_norm": 0.38429944444056074, "learning_rate": 3.5438844428251746e-06, "loss": 0.2712, "step": 6439 }, { "epoch": 2.078776383227733, "grad_norm": 0.5049210182536045, "learning_rate": 3.541616686344615e-06, "loss": 0.2805, "step": 6440 }, { "epoch": 2.079099237257355, "grad_norm": 0.334937028559868, "learning_rate": 3.5393494313857667e-06, "loss": 0.2622, "step": 6441 }, { "epoch": 2.0794220912869767, "grad_norm": 0.35251303767551506, "learning_rate": 3.5370826782358936e-06, "loss": 0.2406, "step": 6442 }, { "epoch": 2.0797449453165986, "grad_norm": 0.39334998109606906, "learning_rate": 3.534816427182186e-06, "loss": 0.2598, "step": 6443 }, { "epoch": 2.0800677993462204, "grad_norm": 0.41574072279853264, "learning_rate": 3.5325506785117793e-06, "loss": 0.2532, "step": 6444 }, { "epoch": 2.0803906533758423, "grad_norm": 0.3659849090349035, "learning_rate": 3.530285432511736e-06, "loss": 0.2591, "step": 6445 }, { "epoch": 2.080713507405464, "grad_norm": 0.38119271774544594, "learning_rate": 3.5280206894690635e-06, "loss": 0.3008, "step": 6446 }, { "epoch": 2.081036361435086, "grad_norm": 0.3497005189678645, "learning_rate": 3.5257564496706973e-06, "loss": 0.2151, "step": 6447 }, { "epoch": 2.081359215464708, "grad_norm": 0.3767300919953675, "learning_rate": 3.5234927134035163e-06, "loss": 0.2999, "step": 6448 }, { "epoch": 2.08168206949433, "grad_norm": 0.33411244653629996, "learning_rate": 3.521229480954335e-06, "loss": 0.2705, "step": 6449 }, { "epoch": 2.082004923523952, "grad_norm": 0.40269420087818364, "learning_rate": 3.518966752609897e-06, "loss": 0.2497, "step": 6450 }, { "epoch": 2.082327777553574, "grad_norm": 0.40416122020591516, "learning_rate": 3.516704528656889e-06, "loss": 0.2185, "step": 6451 }, { "epoch": 2.0826506315831956, "grad_norm": 0.3862995932216033, "learning_rate": 3.514442809381932e-06, "loss": 0.2634, "step": 6452 }, { "epoch": 2.0829734856128175, "grad_norm": 0.40881008624889675, "learning_rate": 3.5121815950715856e-06, "loss": 0.2762, "step": 6453 }, { "epoch": 2.0832963396424393, "grad_norm": 0.4490099331028226, "learning_rate": 3.5099208860123364e-06, "loss": 0.2646, "step": 6454 }, { "epoch": 2.083619193672061, "grad_norm": 0.4075504715410195, "learning_rate": 3.507660682490619e-06, "loss": 0.2903, "step": 6455 }, { "epoch": 2.083942047701683, "grad_norm": 0.35572912244461696, "learning_rate": 3.505400984792793e-06, "loss": 0.2386, "step": 6456 }, { "epoch": 2.084264901731305, "grad_norm": 0.442451481837929, "learning_rate": 3.503141793205164e-06, "loss": 0.2642, "step": 6457 }, { "epoch": 2.0845877557609267, "grad_norm": 0.40190954035648246, "learning_rate": 3.500883108013963e-06, "loss": 0.262, "step": 6458 }, { "epoch": 2.0849106097905485, "grad_norm": 0.37824103953915056, "learning_rate": 3.4986249295053678e-06, "loss": 0.2612, "step": 6459 }, { "epoch": 2.0852334638201704, "grad_norm": 0.3496843342812733, "learning_rate": 3.4963672579654806e-06, "loss": 0.305, "step": 6460 }, { "epoch": 2.0855563178497922, "grad_norm": 0.4860366553546847, "learning_rate": 3.4941100936803515e-06, "loss": 0.3019, "step": 6461 }, { "epoch": 2.085879171879414, "grad_norm": 0.6319343201001525, "learning_rate": 3.491853436935954e-06, "loss": 0.2608, "step": 6462 }, { "epoch": 2.086202025909036, "grad_norm": 0.4071506637424258, "learning_rate": 3.4895972880182103e-06, "loss": 0.2358, "step": 6463 }, { "epoch": 2.0865248799386578, "grad_norm": 0.40861642147242033, "learning_rate": 3.4873416472129644e-06, "loss": 0.2714, "step": 6464 }, { "epoch": 2.0868477339682796, "grad_norm": 0.41751790887277973, "learning_rate": 3.485086514806006e-06, "loss": 0.2686, "step": 6465 }, { "epoch": 2.0871705879979014, "grad_norm": 0.5810846002922949, "learning_rate": 3.48283189108306e-06, "loss": 0.2318, "step": 6466 }, { "epoch": 2.0874934420275233, "grad_norm": 0.33963674309663244, "learning_rate": 3.480577776329778e-06, "loss": 0.2075, "step": 6467 }, { "epoch": 2.087816296057145, "grad_norm": 0.41209498090393315, "learning_rate": 3.478324170831757e-06, "loss": 0.1795, "step": 6468 }, { "epoch": 2.088139150086767, "grad_norm": 0.32738860034535755, "learning_rate": 3.4760710748745263e-06, "loss": 0.2227, "step": 6469 }, { "epoch": 2.088462004116389, "grad_norm": 0.3466519048613188, "learning_rate": 3.473818488743551e-06, "loss": 0.2253, "step": 6470 }, { "epoch": 2.0887848581460107, "grad_norm": 0.4888462993330653, "learning_rate": 3.4715664127242268e-06, "loss": 0.2755, "step": 6471 }, { "epoch": 2.0891077121756325, "grad_norm": 0.42428249490726433, "learning_rate": 3.4693148471018927e-06, "loss": 0.2077, "step": 6472 }, { "epoch": 2.0894305662052544, "grad_norm": 0.3806711291099251, "learning_rate": 3.4670637921618144e-06, "loss": 0.2248, "step": 6473 }, { "epoch": 2.089753420234876, "grad_norm": 0.3831402612264089, "learning_rate": 3.4648132481892026e-06, "loss": 0.2571, "step": 6474 }, { "epoch": 2.090076274264498, "grad_norm": 0.3502651449973606, "learning_rate": 3.462563215469193e-06, "loss": 0.2494, "step": 6475 }, { "epoch": 2.09039912829412, "grad_norm": 0.4140837361555149, "learning_rate": 3.460313694286866e-06, "loss": 0.2885, "step": 6476 }, { "epoch": 2.0907219823237417, "grad_norm": 0.4060023807373396, "learning_rate": 3.4580646849272285e-06, "loss": 0.2553, "step": 6477 }, { "epoch": 2.0910448363533636, "grad_norm": 0.4096773542446071, "learning_rate": 3.455816187675233e-06, "loss": 0.2532, "step": 6478 }, { "epoch": 2.0913676903829854, "grad_norm": 0.4908857658847662, "learning_rate": 3.4535682028157536e-06, "loss": 0.2376, "step": 6479 }, { "epoch": 2.0916905444126073, "grad_norm": 0.37135986933870213, "learning_rate": 3.451320730633614e-06, "loss": 0.2345, "step": 6480 }, { "epoch": 2.092013398442229, "grad_norm": 0.35346098026988887, "learning_rate": 3.4490737714135596e-06, "loss": 0.2517, "step": 6481 }, { "epoch": 2.092336252471851, "grad_norm": 0.4534534439345481, "learning_rate": 3.4468273254402794e-06, "loss": 0.2565, "step": 6482 }, { "epoch": 2.092659106501473, "grad_norm": 0.45200193139581296, "learning_rate": 3.444581392998396e-06, "loss": 0.2619, "step": 6483 }, { "epoch": 2.092981960531095, "grad_norm": 0.39790209590102393, "learning_rate": 3.442335974372468e-06, "loss": 0.2795, "step": 6484 }, { "epoch": 2.093304814560717, "grad_norm": 0.36608671055961806, "learning_rate": 3.440091069846982e-06, "loss": 0.2308, "step": 6485 }, { "epoch": 2.0936276685903388, "grad_norm": 0.44717808803304254, "learning_rate": 3.437846679706366e-06, "loss": 0.2153, "step": 6486 }, { "epoch": 2.0939505226199606, "grad_norm": 0.3514310047590561, "learning_rate": 3.4356028042349845e-06, "loss": 0.2708, "step": 6487 }, { "epoch": 2.0942733766495825, "grad_norm": 0.4031021840474483, "learning_rate": 3.433359443717129e-06, "loss": 0.3218, "step": 6488 }, { "epoch": 2.0945962306792043, "grad_norm": 0.40059482148613096, "learning_rate": 3.4311165984370333e-06, "loss": 0.298, "step": 6489 }, { "epoch": 2.094919084708826, "grad_norm": 0.3462851962003587, "learning_rate": 3.4288742686788594e-06, "loss": 0.2783, "step": 6490 }, { "epoch": 2.095241938738448, "grad_norm": 0.37093237763079695, "learning_rate": 3.4266324547267125e-06, "loss": 0.2727, "step": 6491 }, { "epoch": 2.09556479276807, "grad_norm": 0.4042097981659733, "learning_rate": 3.4243911568646206e-06, "loss": 0.2219, "step": 6492 }, { "epoch": 2.0958876467976917, "grad_norm": 0.366960804304356, "learning_rate": 3.4221503753765596e-06, "loss": 0.2449, "step": 6493 }, { "epoch": 2.0962105008273135, "grad_norm": 0.36348577106186075, "learning_rate": 3.4199101105464283e-06, "loss": 0.2234, "step": 6494 }, { "epoch": 2.0965333548569354, "grad_norm": 0.4143597615164012, "learning_rate": 3.41767036265807e-06, "loss": 0.2136, "step": 6495 }, { "epoch": 2.096856208886557, "grad_norm": 0.32660133515996187, "learning_rate": 3.4154311319952513e-06, "loss": 0.2119, "step": 6496 }, { "epoch": 2.097179062916179, "grad_norm": 0.39802382444968604, "learning_rate": 3.413192418841684e-06, "loss": 0.269, "step": 6497 }, { "epoch": 2.097501916945801, "grad_norm": 0.39922389770162564, "learning_rate": 3.4109542234810128e-06, "loss": 0.2589, "step": 6498 }, { "epoch": 2.0978247709754227, "grad_norm": 0.3491590511030499, "learning_rate": 3.4087165461968066e-06, "loss": 0.2186, "step": 6499 }, { "epoch": 2.0981476250050446, "grad_norm": 0.48351614128757886, "learning_rate": 3.4064793872725813e-06, "loss": 0.2713, "step": 6500 }, { "epoch": 2.0984704790346664, "grad_norm": 0.37385499189912336, "learning_rate": 3.4042427469917826e-06, "loss": 0.2771, "step": 6501 }, { "epoch": 2.0987933330642883, "grad_norm": 0.3731414410579051, "learning_rate": 3.4020066256377845e-06, "loss": 0.3339, "step": 6502 }, { "epoch": 2.09911618709391, "grad_norm": 0.34688409453358093, "learning_rate": 3.3997710234939046e-06, "loss": 0.2472, "step": 6503 }, { "epoch": 2.099439041123532, "grad_norm": 0.5369821015223909, "learning_rate": 3.397535940843391e-06, "loss": 0.2603, "step": 6504 }, { "epoch": 2.099761895153154, "grad_norm": 0.3251820557171299, "learning_rate": 3.395301377969423e-06, "loss": 0.239, "step": 6505 }, { "epoch": 2.1000847491827757, "grad_norm": 0.36333138178690755, "learning_rate": 3.3930673351551205e-06, "loss": 0.22, "step": 6506 }, { "epoch": 2.1004076032123975, "grad_norm": 0.3411552002388359, "learning_rate": 3.3908338126835273e-06, "loss": 0.2632, "step": 6507 }, { "epoch": 2.1007304572420193, "grad_norm": 0.36466339598017616, "learning_rate": 3.388600810837634e-06, "loss": 0.3027, "step": 6508 }, { "epoch": 2.101053311271641, "grad_norm": 0.4097866509588009, "learning_rate": 3.386368329900355e-06, "loss": 0.2986, "step": 6509 }, { "epoch": 2.101376165301263, "grad_norm": 0.41780013234635827, "learning_rate": 3.3841363701545457e-06, "loss": 0.2571, "step": 6510 }, { "epoch": 2.101699019330885, "grad_norm": 0.4299302282788874, "learning_rate": 3.381904931882987e-06, "loss": 0.3156, "step": 6511 }, { "epoch": 2.1020218733605067, "grad_norm": 0.36588421548361283, "learning_rate": 3.3796740153684067e-06, "loss": 0.2933, "step": 6512 }, { "epoch": 2.1023447273901286, "grad_norm": 0.4018247636748991, "learning_rate": 3.3774436208934502e-06, "loss": 0.2654, "step": 6513 }, { "epoch": 2.1026675814197504, "grad_norm": 0.44431004579150685, "learning_rate": 3.375213748740711e-06, "loss": 0.2534, "step": 6514 }, { "epoch": 2.1029904354493723, "grad_norm": 0.3649871479459793, "learning_rate": 3.3729843991927097e-06, "loss": 0.2987, "step": 6515 }, { "epoch": 2.103313289478994, "grad_norm": 0.42043827769104114, "learning_rate": 3.370755572531904e-06, "loss": 0.2451, "step": 6516 }, { "epoch": 2.1036361435086164, "grad_norm": 0.385090532241775, "learning_rate": 3.3685272690406785e-06, "loss": 0.3123, "step": 6517 }, { "epoch": 2.1039589975382382, "grad_norm": 0.4582029123161218, "learning_rate": 3.3662994890013608e-06, "loss": 0.2769, "step": 6518 }, { "epoch": 2.10428185156786, "grad_norm": 0.5086783823034572, "learning_rate": 3.3640722326962037e-06, "loss": 0.2595, "step": 6519 }, { "epoch": 2.104604705597482, "grad_norm": 0.45277259340330495, "learning_rate": 3.361845500407401e-06, "loss": 0.2431, "step": 6520 }, { "epoch": 2.1049275596271038, "grad_norm": 0.468035109815967, "learning_rate": 3.3596192924170728e-06, "loss": 0.2495, "step": 6521 }, { "epoch": 2.1052504136567256, "grad_norm": 0.42298610025033884, "learning_rate": 3.3573936090072784e-06, "loss": 0.2428, "step": 6522 }, { "epoch": 2.1055732676863474, "grad_norm": 0.4923967352104994, "learning_rate": 3.355168450460011e-06, "loss": 0.2596, "step": 6523 }, { "epoch": 2.1058961217159693, "grad_norm": 0.41348819041455925, "learning_rate": 3.352943817057192e-06, "loss": 0.266, "step": 6524 }, { "epoch": 2.106218975745591, "grad_norm": 0.35841529904894476, "learning_rate": 3.350719709080682e-06, "loss": 0.2947, "step": 6525 }, { "epoch": 2.106541829775213, "grad_norm": 0.53165912561655, "learning_rate": 3.348496126812268e-06, "loss": 0.2084, "step": 6526 }, { "epoch": 2.106864683804835, "grad_norm": 0.6956311571093476, "learning_rate": 3.3462730705336816e-06, "loss": 0.2706, "step": 6527 }, { "epoch": 2.1071875378344567, "grad_norm": 0.4185084949300772, "learning_rate": 3.3440505405265737e-06, "loss": 0.2289, "step": 6528 }, { "epoch": 2.1075103918640785, "grad_norm": 0.4031579341929214, "learning_rate": 3.3418285370725395e-06, "loss": 0.2739, "step": 6529 }, { "epoch": 2.1078332458937004, "grad_norm": 0.3155745860835582, "learning_rate": 3.3396070604531033e-06, "loss": 0.2812, "step": 6530 }, { "epoch": 2.108156099923322, "grad_norm": 0.3873428154606321, "learning_rate": 3.3373861109497258e-06, "loss": 0.2496, "step": 6531 }, { "epoch": 2.108478953952944, "grad_norm": 0.3993415538189755, "learning_rate": 3.3351656888437934e-06, "loss": 0.2148, "step": 6532 }, { "epoch": 2.108801807982566, "grad_norm": 0.4475702731215273, "learning_rate": 3.332945794416636e-06, "loss": 0.259, "step": 6533 }, { "epoch": 2.1091246620121877, "grad_norm": 0.418548102865764, "learning_rate": 3.330726427949505e-06, "loss": 0.2651, "step": 6534 }, { "epoch": 2.1094475160418096, "grad_norm": 0.4003863862662002, "learning_rate": 3.328507589723597e-06, "loss": 0.2962, "step": 6535 }, { "epoch": 2.1097703700714314, "grad_norm": 0.3829054609621165, "learning_rate": 3.32628928002003e-06, "loss": 0.2935, "step": 6536 }, { "epoch": 2.1100932241010533, "grad_norm": 0.37666909538498, "learning_rate": 3.3240714991198667e-06, "loss": 0.2548, "step": 6537 }, { "epoch": 2.110416078130675, "grad_norm": 0.368874533049811, "learning_rate": 3.3218542473040904e-06, "loss": 0.2552, "step": 6538 }, { "epoch": 2.110738932160297, "grad_norm": 0.36318295211123675, "learning_rate": 3.319637524853631e-06, "loss": 0.2729, "step": 6539 }, { "epoch": 2.111061786189919, "grad_norm": 0.4094849307830524, "learning_rate": 3.317421332049337e-06, "loss": 0.2135, "step": 6540 }, { "epoch": 2.1113846402195406, "grad_norm": 0.35126678320557353, "learning_rate": 3.3152056691720012e-06, "loss": 0.2382, "step": 6541 }, { "epoch": 2.1117074942491625, "grad_norm": 0.3713449062331212, "learning_rate": 3.312990536502346e-06, "loss": 0.2396, "step": 6542 }, { "epoch": 2.1120303482787843, "grad_norm": 0.36429694239693905, "learning_rate": 3.3107759343210215e-06, "loss": 0.2377, "step": 6543 }, { "epoch": 2.112353202308406, "grad_norm": 0.4858509841674782, "learning_rate": 3.308561862908621e-06, "loss": 0.2163, "step": 6544 }, { "epoch": 2.112676056338028, "grad_norm": 0.45343669291231936, "learning_rate": 3.3063483225456567e-06, "loss": 0.2957, "step": 6545 }, { "epoch": 2.11299891036765, "grad_norm": 0.4408620318721642, "learning_rate": 3.3041353135125857e-06, "loss": 0.2452, "step": 6546 }, { "epoch": 2.1133217643972717, "grad_norm": 0.4427942742158222, "learning_rate": 3.301922836089792e-06, "loss": 0.2145, "step": 6547 }, { "epoch": 2.1136446184268936, "grad_norm": 0.48364333077799604, "learning_rate": 3.299710890557597e-06, "loss": 0.2683, "step": 6548 }, { "epoch": 2.1139674724565154, "grad_norm": 0.37204880077053387, "learning_rate": 3.2974994771962447e-06, "loss": 0.2629, "step": 6549 }, { "epoch": 2.1142903264861372, "grad_norm": 0.3648159875047617, "learning_rate": 3.295288596285924e-06, "loss": 0.2637, "step": 6550 }, { "epoch": 2.1146131805157595, "grad_norm": 0.3730086469386542, "learning_rate": 3.293078248106746e-06, "loss": 0.2647, "step": 6551 }, { "epoch": 2.1149360345453814, "grad_norm": 0.3408685095077833, "learning_rate": 3.290868432938763e-06, "loss": 0.2223, "step": 6552 }, { "epoch": 2.115258888575003, "grad_norm": 0.3313251458580872, "learning_rate": 3.28865915106195e-06, "loss": 0.2704, "step": 6553 }, { "epoch": 2.115581742604625, "grad_norm": 0.3415382680051558, "learning_rate": 3.286450402756227e-06, "loss": 0.2542, "step": 6554 }, { "epoch": 2.115904596634247, "grad_norm": 0.4162391951711725, "learning_rate": 3.2842421883014327e-06, "loss": 0.253, "step": 6555 }, { "epoch": 2.1162274506638687, "grad_norm": 0.39473816344388435, "learning_rate": 3.28203450797735e-06, "loss": 0.2918, "step": 6556 }, { "epoch": 2.1165503046934906, "grad_norm": 0.3952084635416939, "learning_rate": 3.2798273620636836e-06, "loss": 0.2214, "step": 6557 }, { "epoch": 2.1168731587231124, "grad_norm": 0.35264194861190956, "learning_rate": 3.2776207508400825e-06, "loss": 0.2483, "step": 6558 }, { "epoch": 2.1171960127527343, "grad_norm": 0.43482391120513886, "learning_rate": 3.275414674586114e-06, "loss": 0.2366, "step": 6559 }, { "epoch": 2.117518866782356, "grad_norm": 0.36312431587683797, "learning_rate": 3.2732091335812893e-06, "loss": 0.2271, "step": 6560 }, { "epoch": 2.117841720811978, "grad_norm": 0.3960658130455595, "learning_rate": 3.2710041281050463e-06, "loss": 0.2294, "step": 6561 }, { "epoch": 2.1181645748416, "grad_norm": 0.4037933237547759, "learning_rate": 3.2687996584367566e-06, "loss": 0.2313, "step": 6562 }, { "epoch": 2.1184874288712217, "grad_norm": 0.29906086017591904, "learning_rate": 3.2665957248557264e-06, "loss": 0.2147, "step": 6563 }, { "epoch": 2.1188102829008435, "grad_norm": 0.3910399689519069, "learning_rate": 3.2643923276411856e-06, "loss": 0.2258, "step": 6564 }, { "epoch": 2.1191331369304653, "grad_norm": 0.3338240611048421, "learning_rate": 3.262189467072307e-06, "loss": 0.251, "step": 6565 }, { "epoch": 2.119455990960087, "grad_norm": 0.38442482352341256, "learning_rate": 3.259987143428184e-06, "loss": 0.2434, "step": 6566 }, { "epoch": 2.119778844989709, "grad_norm": 0.44924003251667594, "learning_rate": 3.2577853569878536e-06, "loss": 0.2725, "step": 6567 }, { "epoch": 2.120101699019331, "grad_norm": 0.3736748899466993, "learning_rate": 3.2555841080302744e-06, "loss": 0.2821, "step": 6568 }, { "epoch": 2.1204245530489527, "grad_norm": 0.543960932265314, "learning_rate": 3.253383396834346e-06, "loss": 0.2572, "step": 6569 }, { "epoch": 2.1207474070785746, "grad_norm": 0.31110254744769283, "learning_rate": 3.2511832236788907e-06, "loss": 0.2568, "step": 6570 }, { "epoch": 2.1210702611081964, "grad_norm": 0.3853062166405314, "learning_rate": 3.248983588842673e-06, "loss": 0.2329, "step": 6571 }, { "epoch": 2.1213931151378183, "grad_norm": 0.3749073488835467, "learning_rate": 3.2467844926043775e-06, "loss": 0.2075, "step": 6572 }, { "epoch": 2.12171596916744, "grad_norm": 0.3334580331617082, "learning_rate": 3.2445859352426327e-06, "loss": 0.2229, "step": 6573 }, { "epoch": 2.122038823197062, "grad_norm": 0.3846024793702628, "learning_rate": 3.242387917035987e-06, "loss": 0.2447, "step": 6574 }, { "epoch": 2.122361677226684, "grad_norm": 0.4023019540195276, "learning_rate": 3.240190438262929e-06, "loss": 0.2645, "step": 6575 }, { "epoch": 2.1226845312563056, "grad_norm": 0.3681005309711709, "learning_rate": 3.2379934992018767e-06, "loss": 0.2464, "step": 6576 }, { "epoch": 2.1230073852859275, "grad_norm": 0.41355546442777696, "learning_rate": 3.235797100131182e-06, "loss": 0.2674, "step": 6577 }, { "epoch": 2.1233302393155493, "grad_norm": 0.3962047353914292, "learning_rate": 3.233601241329119e-06, "loss": 0.2898, "step": 6578 }, { "epoch": 2.123653093345171, "grad_norm": 0.40410622225128184, "learning_rate": 3.2314059230739037e-06, "loss": 0.2421, "step": 6579 }, { "epoch": 2.123975947374793, "grad_norm": 0.43746269587211095, "learning_rate": 3.229211145643683e-06, "loss": 0.3029, "step": 6580 }, { "epoch": 2.124298801404415, "grad_norm": 0.3949984273536127, "learning_rate": 3.2270169093165256e-06, "loss": 0.2074, "step": 6581 }, { "epoch": 2.1246216554340367, "grad_norm": 0.36715363824183644, "learning_rate": 3.224823214370444e-06, "loss": 0.2012, "step": 6582 }, { "epoch": 2.1249445094636585, "grad_norm": 0.41016971712600064, "learning_rate": 3.222630061083372e-06, "loss": 0.3055, "step": 6583 }, { "epoch": 2.125267363493281, "grad_norm": 0.5046933711645297, "learning_rate": 3.220437449733184e-06, "loss": 0.2396, "step": 6584 }, { "epoch": 2.1255902175229027, "grad_norm": 0.38853587782594184, "learning_rate": 3.218245380597674e-06, "loss": 0.2467, "step": 6585 }, { "epoch": 2.1259130715525245, "grad_norm": 0.422224635110712, "learning_rate": 3.2160538539545814e-06, "loss": 0.2211, "step": 6586 }, { "epoch": 2.1262359255821464, "grad_norm": 0.43754210572738766, "learning_rate": 3.213862870081563e-06, "loss": 0.2572, "step": 6587 }, { "epoch": 2.126558779611768, "grad_norm": 0.3520761554796257, "learning_rate": 3.2116724292562193e-06, "loss": 0.2496, "step": 6588 }, { "epoch": 2.12688163364139, "grad_norm": 0.40372447524394145, "learning_rate": 3.2094825317560707e-06, "loss": 0.2555, "step": 6589 }, { "epoch": 2.127204487671012, "grad_norm": 0.34605870969030694, "learning_rate": 3.20729317785858e-06, "loss": 0.3135, "step": 6590 }, { "epoch": 2.1275273417006337, "grad_norm": 0.4977974787641495, "learning_rate": 3.2051043678411294e-06, "loss": 0.2439, "step": 6591 }, { "epoch": 2.1278501957302556, "grad_norm": 0.3589664486796184, "learning_rate": 3.2029161019810417e-06, "loss": 0.2459, "step": 6592 }, { "epoch": 2.1281730497598774, "grad_norm": 0.3682235286178219, "learning_rate": 3.2007283805555653e-06, "loss": 0.26, "step": 6593 }, { "epoch": 2.1284959037894993, "grad_norm": 0.4159890689978388, "learning_rate": 3.198541203841886e-06, "loss": 0.2597, "step": 6594 }, { "epoch": 2.128818757819121, "grad_norm": 0.4796524776326051, "learning_rate": 3.196354572117109e-06, "loss": 0.2835, "step": 6595 }, { "epoch": 2.129141611848743, "grad_norm": 0.3529489037200264, "learning_rate": 3.1941684856582824e-06, "loss": 0.2376, "step": 6596 }, { "epoch": 2.129464465878365, "grad_norm": 0.3417569018020133, "learning_rate": 3.1919829447423812e-06, "loss": 0.2758, "step": 6597 }, { "epoch": 2.1297873199079866, "grad_norm": 0.4074233318953555, "learning_rate": 3.189797949646306e-06, "loss": 0.2497, "step": 6598 }, { "epoch": 2.1301101739376085, "grad_norm": 0.3425126841170033, "learning_rate": 3.187613500646898e-06, "loss": 0.2527, "step": 6599 }, { "epoch": 2.1304330279672303, "grad_norm": 0.45581003370012463, "learning_rate": 3.1854295980209184e-06, "loss": 0.2454, "step": 6600 }, { "epoch": 2.130755881996852, "grad_norm": 0.3911307605246889, "learning_rate": 3.1832462420450695e-06, "loss": 0.2656, "step": 6601 }, { "epoch": 2.131078736026474, "grad_norm": 0.41256228508548737, "learning_rate": 3.1810634329959746e-06, "loss": 0.245, "step": 6602 }, { "epoch": 2.131401590056096, "grad_norm": 0.381782731110204, "learning_rate": 3.1788811711501973e-06, "loss": 0.2576, "step": 6603 }, { "epoch": 2.1317244440857177, "grad_norm": 0.4293502313488053, "learning_rate": 3.176699456784223e-06, "loss": 0.2285, "step": 6604 }, { "epoch": 2.1320472981153396, "grad_norm": 0.3914588540876559, "learning_rate": 3.174518290174477e-06, "loss": 0.3085, "step": 6605 }, { "epoch": 2.1323701521449614, "grad_norm": 0.39485235837729377, "learning_rate": 3.172337671597304e-06, "loss": 0.3181, "step": 6606 }, { "epoch": 2.1326930061745832, "grad_norm": 0.38084284925383227, "learning_rate": 3.170157601328988e-06, "loss": 0.2752, "step": 6607 }, { "epoch": 2.133015860204205, "grad_norm": 0.40160374168541285, "learning_rate": 3.1679780796457417e-06, "loss": 0.2638, "step": 6608 }, { "epoch": 2.133338714233827, "grad_norm": 0.345272546296267, "learning_rate": 3.16579910682371e-06, "loss": 0.3015, "step": 6609 }, { "epoch": 2.1336615682634488, "grad_norm": 0.3997313624268582, "learning_rate": 3.1636206831389602e-06, "loss": 0.2874, "step": 6610 }, { "epoch": 2.1339844222930706, "grad_norm": 0.3676842480779818, "learning_rate": 3.161442808867501e-06, "loss": 0.2802, "step": 6611 }, { "epoch": 2.1343072763226925, "grad_norm": 0.34463589220684004, "learning_rate": 3.159265484285261e-06, "loss": 0.2072, "step": 6612 }, { "epoch": 2.1346301303523143, "grad_norm": 0.35020177613000775, "learning_rate": 3.1570887096681093e-06, "loss": 0.269, "step": 6613 }, { "epoch": 2.134952984381936, "grad_norm": 0.38261525774810323, "learning_rate": 3.1549124852918337e-06, "loss": 0.2807, "step": 6614 }, { "epoch": 2.135275838411558, "grad_norm": 0.3300156466020379, "learning_rate": 3.152736811432164e-06, "loss": 0.2438, "step": 6615 }, { "epoch": 2.13559869244118, "grad_norm": 0.39537198872028523, "learning_rate": 3.150561688364756e-06, "loss": 0.2169, "step": 6616 }, { "epoch": 2.1359215464708017, "grad_norm": 0.4358661211182034, "learning_rate": 3.14838711636519e-06, "loss": 0.2908, "step": 6617 }, { "epoch": 2.1362444005004235, "grad_norm": 0.38689709568807595, "learning_rate": 3.146213095708986e-06, "loss": 0.2322, "step": 6618 }, { "epoch": 2.1365672545300454, "grad_norm": 0.3316448400649025, "learning_rate": 3.144039626671585e-06, "loss": 0.2473, "step": 6619 }, { "epoch": 2.1368901085596677, "grad_norm": 0.38793203076601523, "learning_rate": 3.141866709528367e-06, "loss": 0.3163, "step": 6620 }, { "epoch": 2.1372129625892895, "grad_norm": 0.38037600858327225, "learning_rate": 3.1396943445546334e-06, "loss": 0.2945, "step": 6621 }, { "epoch": 2.1375358166189113, "grad_norm": 0.3476220507837679, "learning_rate": 3.1375225320256236e-06, "loss": 0.2966, "step": 6622 }, { "epoch": 2.137858670648533, "grad_norm": 0.3936464717270943, "learning_rate": 3.1353512722165e-06, "loss": 0.2319, "step": 6623 }, { "epoch": 2.138181524678155, "grad_norm": 0.3460983086637927, "learning_rate": 3.133180565402359e-06, "loss": 0.2637, "step": 6624 }, { "epoch": 2.138504378707777, "grad_norm": 0.40948449946539794, "learning_rate": 3.1310104118582273e-06, "loss": 0.2677, "step": 6625 }, { "epoch": 2.1388272327373987, "grad_norm": 0.3853535449941315, "learning_rate": 3.128840811859062e-06, "loss": 0.2744, "step": 6626 }, { "epoch": 2.1391500867670206, "grad_norm": 0.3778844633396288, "learning_rate": 3.126671765679744e-06, "loss": 0.3064, "step": 6627 }, { "epoch": 2.1394729407966424, "grad_norm": 0.4622190325396975, "learning_rate": 3.124503273595094e-06, "loss": 0.2972, "step": 6628 }, { "epoch": 2.1397957948262643, "grad_norm": 0.39997519089882844, "learning_rate": 3.12233533587985e-06, "loss": 0.2447, "step": 6629 }, { "epoch": 2.140118648855886, "grad_norm": 0.386833850860323, "learning_rate": 3.120167952808694e-06, "loss": 0.2327, "step": 6630 }, { "epoch": 2.140441502885508, "grad_norm": 0.3537951006093607, "learning_rate": 3.1180011246562247e-06, "loss": 0.2497, "step": 6631 }, { "epoch": 2.14076435691513, "grad_norm": 0.4248012168236928, "learning_rate": 3.1158348516969813e-06, "loss": 0.2602, "step": 6632 }, { "epoch": 2.1410872109447516, "grad_norm": 0.3944268513839338, "learning_rate": 3.113669134205421e-06, "loss": 0.239, "step": 6633 }, { "epoch": 2.1414100649743735, "grad_norm": 0.4282302514426748, "learning_rate": 3.111503972455942e-06, "loss": 0.2567, "step": 6634 }, { "epoch": 2.1417329190039953, "grad_norm": 0.38459753136329355, "learning_rate": 3.1093393667228684e-06, "loss": 0.2625, "step": 6635 }, { "epoch": 2.142055773033617, "grad_norm": 0.3847124697426859, "learning_rate": 3.107175317280449e-06, "loss": 0.2599, "step": 6636 }, { "epoch": 2.142378627063239, "grad_norm": 0.3939612825464625, "learning_rate": 3.1050118244028697e-06, "loss": 0.3286, "step": 6637 }, { "epoch": 2.142701481092861, "grad_norm": 0.3599282987699305, "learning_rate": 3.102848888364237e-06, "loss": 0.2271, "step": 6638 }, { "epoch": 2.1430243351224827, "grad_norm": 0.3779744562734972, "learning_rate": 3.100686509438595e-06, "loss": 0.2599, "step": 6639 }, { "epoch": 2.1433471891521045, "grad_norm": 0.31829100425860724, "learning_rate": 3.0985246878999144e-06, "loss": 0.2787, "step": 6640 }, { "epoch": 2.1436700431817264, "grad_norm": 0.36992203633308346, "learning_rate": 3.0963634240220974e-06, "loss": 0.2611, "step": 6641 }, { "epoch": 2.1439928972113482, "grad_norm": 0.3634866019263156, "learning_rate": 3.0942027180789675e-06, "loss": 0.2861, "step": 6642 }, { "epoch": 2.14431575124097, "grad_norm": 0.4072291916475156, "learning_rate": 3.092042570344289e-06, "loss": 0.2346, "step": 6643 }, { "epoch": 2.144638605270592, "grad_norm": 0.33880987066753654, "learning_rate": 3.0898829810917437e-06, "loss": 0.1941, "step": 6644 }, { "epoch": 2.1449614593002138, "grad_norm": 0.400758986124198, "learning_rate": 3.087723950594954e-06, "loss": 0.2251, "step": 6645 }, { "epoch": 2.1452843133298356, "grad_norm": 0.41175480057649505, "learning_rate": 3.085565479127462e-06, "loss": 0.2331, "step": 6646 }, { "epoch": 2.1456071673594574, "grad_norm": 0.3719203898444663, "learning_rate": 3.0834075669627467e-06, "loss": 0.2917, "step": 6647 }, { "epoch": 2.1459300213890793, "grad_norm": 0.3737559155269683, "learning_rate": 3.0812502143742082e-06, "loss": 0.2392, "step": 6648 }, { "epoch": 2.146252875418701, "grad_norm": 0.4036689067340389, "learning_rate": 3.079093421635185e-06, "loss": 0.2923, "step": 6649 }, { "epoch": 2.146575729448323, "grad_norm": 0.3532754441638372, "learning_rate": 3.0769371890189344e-06, "loss": 0.2902, "step": 6650 }, { "epoch": 2.1468985834779453, "grad_norm": 0.3358402795670458, "learning_rate": 3.074781516798652e-06, "loss": 0.2276, "step": 6651 }, { "epoch": 2.147221437507567, "grad_norm": 0.4906042985926392, "learning_rate": 3.07262640524746e-06, "loss": 0.2738, "step": 6652 }, { "epoch": 2.147544291537189, "grad_norm": 0.35805726377812774, "learning_rate": 3.070471854638402e-06, "loss": 0.2504, "step": 6653 }, { "epoch": 2.147867145566811, "grad_norm": 0.3859493886102868, "learning_rate": 3.068317865244461e-06, "loss": 0.3477, "step": 6654 }, { "epoch": 2.1481899995964326, "grad_norm": 0.451756890132384, "learning_rate": 3.066164437338546e-06, "loss": 0.2392, "step": 6655 }, { "epoch": 2.1485128536260545, "grad_norm": 0.39036112382321975, "learning_rate": 3.064011571193488e-06, "loss": 0.2385, "step": 6656 }, { "epoch": 2.1488357076556763, "grad_norm": 0.3306648691748502, "learning_rate": 3.0618592670820568e-06, "loss": 0.2436, "step": 6657 }, { "epoch": 2.149158561685298, "grad_norm": 0.35177602703604494, "learning_rate": 3.0597075252769457e-06, "loss": 0.2194, "step": 6658 }, { "epoch": 2.14948141571492, "grad_norm": 0.38430551083672937, "learning_rate": 3.057556346050775e-06, "loss": 0.2678, "step": 6659 }, { "epoch": 2.149804269744542, "grad_norm": 0.3791278539342074, "learning_rate": 3.0554057296760995e-06, "loss": 0.2454, "step": 6660 }, { "epoch": 2.1501271237741637, "grad_norm": 0.3861165129119351, "learning_rate": 3.053255676425395e-06, "loss": 0.2765, "step": 6661 }, { "epoch": 2.1504499778037856, "grad_norm": 0.38836991279527067, "learning_rate": 3.0511061865710756e-06, "loss": 0.2635, "step": 6662 }, { "epoch": 2.1507728318334074, "grad_norm": 0.3816522496245352, "learning_rate": 3.048957260385473e-06, "loss": 0.2487, "step": 6663 }, { "epoch": 2.1510956858630292, "grad_norm": 0.38868935969916907, "learning_rate": 3.0468088981408587e-06, "loss": 0.2534, "step": 6664 }, { "epoch": 2.151418539892651, "grad_norm": 0.44341372027171266, "learning_rate": 3.044661100109422e-06, "loss": 0.3211, "step": 6665 }, { "epoch": 2.151741393922273, "grad_norm": 0.3914868815174183, "learning_rate": 3.0425138665632915e-06, "loss": 0.25, "step": 6666 }, { "epoch": 2.1520642479518948, "grad_norm": 0.40690408840676073, "learning_rate": 3.0403671977745134e-06, "loss": 0.2453, "step": 6667 }, { "epoch": 2.1523871019815166, "grad_norm": 0.3937643177253394, "learning_rate": 3.038221094015072e-06, "loss": 0.2781, "step": 6668 }, { "epoch": 2.1527099560111385, "grad_norm": 0.38941867107777783, "learning_rate": 3.0360755555568716e-06, "loss": 0.2448, "step": 6669 }, { "epoch": 2.1530328100407603, "grad_norm": 0.3982277488618108, "learning_rate": 3.033930582671751e-06, "loss": 0.2476, "step": 6670 }, { "epoch": 2.153355664070382, "grad_norm": 0.3843731482057428, "learning_rate": 3.0317861756314748e-06, "loss": 0.2868, "step": 6671 }, { "epoch": 2.153678518100004, "grad_norm": 0.41876726125729014, "learning_rate": 3.0296423347077378e-06, "loss": 0.2938, "step": 6672 }, { "epoch": 2.154001372129626, "grad_norm": 0.33890142243620686, "learning_rate": 3.0274990601721635e-06, "loss": 0.2542, "step": 6673 }, { "epoch": 2.1543242261592477, "grad_norm": 0.35108010767367787, "learning_rate": 3.0253563522962963e-06, "loss": 0.2644, "step": 6674 }, { "epoch": 2.1546470801888695, "grad_norm": 0.41451961541617577, "learning_rate": 3.0232142113516204e-06, "loss": 0.2429, "step": 6675 }, { "epoch": 2.1549699342184914, "grad_norm": 0.38585455878515473, "learning_rate": 3.021072637609536e-06, "loss": 0.2451, "step": 6676 }, { "epoch": 2.155292788248113, "grad_norm": 0.3246302462693811, "learning_rate": 3.0189316313413827e-06, "loss": 0.264, "step": 6677 }, { "epoch": 2.155615642277735, "grad_norm": 0.3634468785896254, "learning_rate": 3.0167911928184196e-06, "loss": 0.2499, "step": 6678 }, { "epoch": 2.155938496307357, "grad_norm": 0.4932795198945838, "learning_rate": 3.0146513223118406e-06, "loss": 0.2066, "step": 6679 }, { "epoch": 2.1562613503369787, "grad_norm": 0.40301120268500157, "learning_rate": 3.0125120200927594e-06, "loss": 0.2197, "step": 6680 }, { "epoch": 2.1565842043666006, "grad_norm": 0.3720464964798706, "learning_rate": 3.010373286432229e-06, "loss": 0.2442, "step": 6681 }, { "epoch": 2.1569070583962224, "grad_norm": 0.36724876347391894, "learning_rate": 3.0082351216012173e-06, "loss": 0.2969, "step": 6682 }, { "epoch": 2.1572299124258443, "grad_norm": 0.3796204153832132, "learning_rate": 3.006097525870633e-06, "loss": 0.2675, "step": 6683 }, { "epoch": 2.157552766455466, "grad_norm": 0.3566390021936305, "learning_rate": 3.0039604995113008e-06, "loss": 0.2411, "step": 6684 }, { "epoch": 2.157875620485088, "grad_norm": 0.35550439215230145, "learning_rate": 3.001824042793982e-06, "loss": 0.2479, "step": 6685 }, { "epoch": 2.15819847451471, "grad_norm": 0.44790149932863094, "learning_rate": 2.9996881559893626e-06, "loss": 0.2628, "step": 6686 }, { "epoch": 2.158521328544332, "grad_norm": 0.3996132443420178, "learning_rate": 2.99755283936806e-06, "loss": 0.3037, "step": 6687 }, { "epoch": 2.158844182573954, "grad_norm": 0.3713065444530172, "learning_rate": 2.9954180932006088e-06, "loss": 0.2005, "step": 6688 }, { "epoch": 2.159167036603576, "grad_norm": 0.30904871955590063, "learning_rate": 2.9932839177574827e-06, "loss": 0.2644, "step": 6689 }, { "epoch": 2.1594898906331976, "grad_norm": 0.3956091490339918, "learning_rate": 2.99115031330908e-06, "loss": 0.2631, "step": 6690 }, { "epoch": 2.1598127446628195, "grad_norm": 0.46576110142502886, "learning_rate": 2.989017280125722e-06, "loss": 0.2438, "step": 6691 }, { "epoch": 2.1601355986924413, "grad_norm": 0.3572026062930505, "learning_rate": 2.9868848184776642e-06, "loss": 0.2702, "step": 6692 }, { "epoch": 2.160458452722063, "grad_norm": 0.38453168806971555, "learning_rate": 2.984752928635083e-06, "loss": 0.2113, "step": 6693 }, { "epoch": 2.160781306751685, "grad_norm": 0.3817595444410544, "learning_rate": 2.9826216108680896e-06, "loss": 0.2723, "step": 6694 }, { "epoch": 2.161104160781307, "grad_norm": 0.38240350546596286, "learning_rate": 2.9804908654467153e-06, "loss": 0.302, "step": 6695 }, { "epoch": 2.1614270148109287, "grad_norm": 0.46980826264307696, "learning_rate": 2.9783606926409277e-06, "loss": 0.2346, "step": 6696 }, { "epoch": 2.1617498688405505, "grad_norm": 0.40131913796875873, "learning_rate": 2.976231092720611e-06, "loss": 0.2721, "step": 6697 }, { "epoch": 2.1620727228701724, "grad_norm": 0.3407149115653341, "learning_rate": 2.9741020659555872e-06, "loss": 0.2359, "step": 6698 }, { "epoch": 2.1623955768997942, "grad_norm": 0.39142970188819204, "learning_rate": 2.9719736126155965e-06, "loss": 0.2425, "step": 6699 }, { "epoch": 2.162718430929416, "grad_norm": 0.39816676027711095, "learning_rate": 2.9698457329703137e-06, "loss": 0.2957, "step": 6700 }, { "epoch": 2.163041284959038, "grad_norm": 0.3573493055925116, "learning_rate": 2.9677184272893414e-06, "loss": 0.2495, "step": 6701 }, { "epoch": 2.1633641389886598, "grad_norm": 0.4566427612772299, "learning_rate": 2.965591695842199e-06, "loss": 0.2331, "step": 6702 }, { "epoch": 2.1636869930182816, "grad_norm": 0.3474340462134879, "learning_rate": 2.963465538898346e-06, "loss": 0.2691, "step": 6703 }, { "epoch": 2.1640098470479034, "grad_norm": 0.4026772452631942, "learning_rate": 2.9613399567271643e-06, "loss": 0.2555, "step": 6704 }, { "epoch": 2.1643327010775253, "grad_norm": 0.3724516832305408, "learning_rate": 2.9592149495979564e-06, "loss": 0.2838, "step": 6705 }, { "epoch": 2.164655555107147, "grad_norm": 0.35111634463590163, "learning_rate": 2.957090517779962e-06, "loss": 0.2431, "step": 6706 }, { "epoch": 2.164978409136769, "grad_norm": 0.3304287374543579, "learning_rate": 2.9549666615423465e-06, "loss": 0.273, "step": 6707 }, { "epoch": 2.165301263166391, "grad_norm": 0.43089155468468104, "learning_rate": 2.9528433811541922e-06, "loss": 0.2899, "step": 6708 }, { "epoch": 2.1656241171960127, "grad_norm": 0.34957162032152816, "learning_rate": 2.950720676884523e-06, "loss": 0.3184, "step": 6709 }, { "epoch": 2.1659469712256345, "grad_norm": 0.34667359276659554, "learning_rate": 2.948598549002276e-06, "loss": 0.2192, "step": 6710 }, { "epoch": 2.1662698252552564, "grad_norm": 0.6899110849401748, "learning_rate": 2.9464769977763285e-06, "loss": 0.2486, "step": 6711 }, { "epoch": 2.166592679284878, "grad_norm": 0.35007338495047075, "learning_rate": 2.944356023475473e-06, "loss": 0.2859, "step": 6712 }, { "epoch": 2.1669155333145, "grad_norm": 0.4190415828709978, "learning_rate": 2.9422356263684367e-06, "loss": 0.2312, "step": 6713 }, { "epoch": 2.167238387344122, "grad_norm": 0.44624991001727876, "learning_rate": 2.940115806723869e-06, "loss": 0.2211, "step": 6714 }, { "epoch": 2.1675612413737437, "grad_norm": 0.4782852800741727, "learning_rate": 2.937996564810351e-06, "loss": 0.2404, "step": 6715 }, { "epoch": 2.1678840954033656, "grad_norm": 0.42129937571695253, "learning_rate": 2.935877900896383e-06, "loss": 0.2358, "step": 6716 }, { "epoch": 2.1682069494329874, "grad_norm": 0.38977874790053, "learning_rate": 2.933759815250401e-06, "loss": 0.238, "step": 6717 }, { "epoch": 2.1685298034626093, "grad_norm": 0.6601620792766856, "learning_rate": 2.931642308140761e-06, "loss": 0.2234, "step": 6718 }, { "epoch": 2.1688526574922316, "grad_norm": 0.3685232345778636, "learning_rate": 2.9295253798357525e-06, "loss": 0.2224, "step": 6719 }, { "epoch": 2.1691755115218534, "grad_norm": 0.46334320484143754, "learning_rate": 2.9274090306035826e-06, "loss": 0.3245, "step": 6720 }, { "epoch": 2.1694983655514752, "grad_norm": 0.4110949281213264, "learning_rate": 2.925293260712394e-06, "loss": 0.2608, "step": 6721 }, { "epoch": 2.169821219581097, "grad_norm": 0.35911102108304577, "learning_rate": 2.9231780704302467e-06, "loss": 0.2473, "step": 6722 }, { "epoch": 2.170144073610719, "grad_norm": 0.34515938390454065, "learning_rate": 2.921063460025138e-06, "loss": 0.2464, "step": 6723 }, { "epoch": 2.1704669276403408, "grad_norm": 0.29665120805249356, "learning_rate": 2.9189494297649806e-06, "loss": 0.237, "step": 6724 }, { "epoch": 2.1707897816699626, "grad_norm": 0.39026197670610463, "learning_rate": 2.9168359799176216e-06, "loss": 0.2633, "step": 6725 }, { "epoch": 2.1711126356995845, "grad_norm": 0.4116822369543991, "learning_rate": 2.914723110750836e-06, "loss": 0.3258, "step": 6726 }, { "epoch": 2.1714354897292063, "grad_norm": 0.37923390184976014, "learning_rate": 2.912610822532315e-06, "loss": 0.2175, "step": 6727 }, { "epoch": 2.171758343758828, "grad_norm": 0.3341551650813858, "learning_rate": 2.910499115529688e-06, "loss": 0.2863, "step": 6728 }, { "epoch": 2.17208119778845, "grad_norm": 0.46328499822081326, "learning_rate": 2.9083879900105e-06, "loss": 0.2581, "step": 6729 }, { "epoch": 2.172404051818072, "grad_norm": 0.41625687344346884, "learning_rate": 2.906277446242234e-06, "loss": 0.2434, "step": 6730 }, { "epoch": 2.1727269058476937, "grad_norm": 0.37200902138600084, "learning_rate": 2.9041674844922865e-06, "loss": 0.2501, "step": 6731 }, { "epoch": 2.1730497598773155, "grad_norm": 0.3761956666315352, "learning_rate": 2.9020581050279903e-06, "loss": 0.2477, "step": 6732 }, { "epoch": 2.1733726139069374, "grad_norm": 0.35108116687871427, "learning_rate": 2.8999493081166027e-06, "loss": 0.2424, "step": 6733 }, { "epoch": 2.173695467936559, "grad_norm": 0.45024772467862567, "learning_rate": 2.8978410940253003e-06, "loss": 0.3032, "step": 6734 }, { "epoch": 2.174018321966181, "grad_norm": 0.3560920777681693, "learning_rate": 2.8957334630211945e-06, "loss": 0.2111, "step": 6735 }, { "epoch": 2.174341175995803, "grad_norm": 0.40315922213922106, "learning_rate": 2.8936264153713205e-06, "loss": 0.2277, "step": 6736 }, { "epoch": 2.1746640300254247, "grad_norm": 0.34331102558307885, "learning_rate": 2.8915199513426335e-06, "loss": 0.2658, "step": 6737 }, { "epoch": 2.1749868840550466, "grad_norm": 0.3801657157967613, "learning_rate": 2.889414071202025e-06, "loss": 0.2522, "step": 6738 }, { "epoch": 2.1753097380846684, "grad_norm": 0.3529568041535124, "learning_rate": 2.8873087752163035e-06, "loss": 0.2335, "step": 6739 }, { "epoch": 2.1756325921142903, "grad_norm": 0.40826163196346155, "learning_rate": 2.885204063652209e-06, "loss": 0.2553, "step": 6740 }, { "epoch": 2.175955446143912, "grad_norm": 0.4780166561695624, "learning_rate": 2.8830999367764024e-06, "loss": 0.2146, "step": 6741 }, { "epoch": 2.176278300173534, "grad_norm": 0.3789275333800805, "learning_rate": 2.8809963948554786e-06, "loss": 0.2582, "step": 6742 }, { "epoch": 2.176601154203156, "grad_norm": 0.37795160063580635, "learning_rate": 2.878893438155949e-06, "loss": 0.3159, "step": 6743 }, { "epoch": 2.1769240082327777, "grad_norm": 0.34592609724130086, "learning_rate": 2.8767910669442564e-06, "loss": 0.2372, "step": 6744 }, { "epoch": 2.1772468622623995, "grad_norm": 0.3853450854992875, "learning_rate": 2.874689281486772e-06, "loss": 0.2869, "step": 6745 }, { "epoch": 2.1775697162920213, "grad_norm": 0.46351940600103225, "learning_rate": 2.872588082049783e-06, "loss": 0.2696, "step": 6746 }, { "epoch": 2.177892570321643, "grad_norm": 0.41589528417054233, "learning_rate": 2.870487468899515e-06, "loss": 0.2438, "step": 6747 }, { "epoch": 2.178215424351265, "grad_norm": 0.4185632616932483, "learning_rate": 2.868387442302107e-06, "loss": 0.2984, "step": 6748 }, { "epoch": 2.178538278380887, "grad_norm": 0.35811348458846254, "learning_rate": 2.8662880025236314e-06, "loss": 0.2516, "step": 6749 }, { "epoch": 2.1788611324105087, "grad_norm": 0.3878611622034584, "learning_rate": 2.864189149830085e-06, "loss": 0.2946, "step": 6750 }, { "epoch": 2.1791839864401306, "grad_norm": 0.40309327703751024, "learning_rate": 2.8620908844873913e-06, "loss": 0.3027, "step": 6751 }, { "epoch": 2.1795068404697524, "grad_norm": 0.3865434266513562, "learning_rate": 2.859993206761394e-06, "loss": 0.2673, "step": 6752 }, { "epoch": 2.1798296944993742, "grad_norm": 0.35365798144389354, "learning_rate": 2.8578961169178704e-06, "loss": 0.2251, "step": 6753 }, { "epoch": 2.180152548528996, "grad_norm": 0.3288674703962352, "learning_rate": 2.8557996152225132e-06, "loss": 0.243, "step": 6754 }, { "epoch": 2.1804754025586184, "grad_norm": 0.3687739102914283, "learning_rate": 2.8537037019409525e-06, "loss": 0.2244, "step": 6755 }, { "epoch": 2.1807982565882402, "grad_norm": 0.45080690198972584, "learning_rate": 2.8516083773387316e-06, "loss": 0.291, "step": 6756 }, { "epoch": 2.181121110617862, "grad_norm": 0.36700557672941964, "learning_rate": 2.8495136416813307e-06, "loss": 0.2711, "step": 6757 }, { "epoch": 2.181443964647484, "grad_norm": 0.3523634769598697, "learning_rate": 2.847419495234144e-06, "loss": 0.2878, "step": 6758 }, { "epoch": 2.1817668186771058, "grad_norm": 0.43342265693764814, "learning_rate": 2.845325938262504e-06, "loss": 0.2556, "step": 6759 }, { "epoch": 2.1820896727067276, "grad_norm": 0.3939293517503846, "learning_rate": 2.8432329710316536e-06, "loss": 0.2093, "step": 6760 }, { "epoch": 2.1824125267363494, "grad_norm": 0.372227362596973, "learning_rate": 2.841140593806777e-06, "loss": 0.2444, "step": 6761 }, { "epoch": 2.1827353807659713, "grad_norm": 0.43175254613887104, "learning_rate": 2.8390488068529686e-06, "loss": 0.2666, "step": 6762 }, { "epoch": 2.183058234795593, "grad_norm": 0.4202950073229839, "learning_rate": 2.8369576104352583e-06, "loss": 0.2944, "step": 6763 }, { "epoch": 2.183381088825215, "grad_norm": 0.46060230431091753, "learning_rate": 2.8348670048185967e-06, "loss": 0.2556, "step": 6764 }, { "epoch": 2.183703942854837, "grad_norm": 0.5666552998955067, "learning_rate": 2.8327769902678636e-06, "loss": 0.2628, "step": 6765 }, { "epoch": 2.1840267968844587, "grad_norm": 0.45908036936656205, "learning_rate": 2.8306875670478567e-06, "loss": 0.2753, "step": 6766 }, { "epoch": 2.1843496509140805, "grad_norm": 0.4728153134384067, "learning_rate": 2.828598735423305e-06, "loss": 0.26, "step": 6767 }, { "epoch": 2.1846725049437024, "grad_norm": 0.36420495208560694, "learning_rate": 2.826510495658863e-06, "loss": 0.3118, "step": 6768 }, { "epoch": 2.184995358973324, "grad_norm": 0.3536494610396174, "learning_rate": 2.824422848019103e-06, "loss": 0.2709, "step": 6769 }, { "epoch": 2.185318213002946, "grad_norm": 0.42902186442003265, "learning_rate": 2.8223357927685315e-06, "loss": 0.2932, "step": 6770 }, { "epoch": 2.185641067032568, "grad_norm": 0.3198865986086284, "learning_rate": 2.820249330171572e-06, "loss": 0.2373, "step": 6771 }, { "epoch": 2.1859639210621897, "grad_norm": 0.36068332254600227, "learning_rate": 2.81816346049258e-06, "loss": 0.2554, "step": 6772 }, { "epoch": 2.1862867750918116, "grad_norm": 0.41288416477071915, "learning_rate": 2.816078183995828e-06, "loss": 0.2484, "step": 6773 }, { "epoch": 2.1866096291214334, "grad_norm": 0.3869667971319213, "learning_rate": 2.813993500945522e-06, "loss": 0.2885, "step": 6774 }, { "epoch": 2.1869324831510553, "grad_norm": 0.4022705498177602, "learning_rate": 2.811909411605784e-06, "loss": 0.2453, "step": 6775 }, { "epoch": 2.187255337180677, "grad_norm": 0.41435906362998626, "learning_rate": 2.80982591624067e-06, "loss": 0.2748, "step": 6776 }, { "epoch": 2.187578191210299, "grad_norm": 0.37701236663143434, "learning_rate": 2.8077430151141518e-06, "loss": 0.2933, "step": 6777 }, { "epoch": 2.187901045239921, "grad_norm": 0.3876453232931542, "learning_rate": 2.8056607084901317e-06, "loss": 0.2314, "step": 6778 }, { "epoch": 2.1882238992695426, "grad_norm": 0.39123367101229817, "learning_rate": 2.8035789966324373e-06, "loss": 0.2254, "step": 6779 }, { "epoch": 2.1885467532991645, "grad_norm": 0.42294473854102754, "learning_rate": 2.8014978798048154e-06, "loss": 0.2514, "step": 6780 }, { "epoch": 2.1888696073287863, "grad_norm": 0.406540108071257, "learning_rate": 2.7994173582709408e-06, "loss": 0.2193, "step": 6781 }, { "epoch": 2.189192461358408, "grad_norm": 0.39383556898618355, "learning_rate": 2.7973374322944135e-06, "loss": 0.2131, "step": 6782 }, { "epoch": 2.18951531538803, "grad_norm": 0.3263872478538738, "learning_rate": 2.795258102138761e-06, "loss": 0.2795, "step": 6783 }, { "epoch": 2.189838169417652, "grad_norm": 0.40288928428999926, "learning_rate": 2.7931793680674254e-06, "loss": 0.2453, "step": 6784 }, { "epoch": 2.1901610234472737, "grad_norm": 0.42239216167648896, "learning_rate": 2.791101230343785e-06, "loss": 0.3114, "step": 6785 }, { "epoch": 2.190483877476896, "grad_norm": 1.273108750092124, "learning_rate": 2.789023689231131e-06, "loss": 0.261, "step": 6786 }, { "epoch": 2.190806731506518, "grad_norm": 0.4026780235760824, "learning_rate": 2.7869467449926905e-06, "loss": 0.2936, "step": 6787 }, { "epoch": 2.1911295855361397, "grad_norm": 0.4860600932330264, "learning_rate": 2.7848703978916054e-06, "loss": 0.2656, "step": 6788 }, { "epoch": 2.1914524395657615, "grad_norm": 0.42446116687988883, "learning_rate": 2.7827946481909504e-06, "loss": 0.2594, "step": 6789 }, { "epoch": 2.1917752935953834, "grad_norm": 0.37083874651420096, "learning_rate": 2.7807194961537148e-06, "loss": 0.2133, "step": 6790 }, { "epoch": 2.192098147625005, "grad_norm": 0.41937069152486095, "learning_rate": 2.778644942042822e-06, "loss": 0.261, "step": 6791 }, { "epoch": 2.192421001654627, "grad_norm": 0.3361040572416573, "learning_rate": 2.776570986121111e-06, "loss": 0.2484, "step": 6792 }, { "epoch": 2.192743855684249, "grad_norm": 0.3522783204073223, "learning_rate": 2.774497628651355e-06, "loss": 0.2883, "step": 6793 }, { "epoch": 2.1930667097138707, "grad_norm": 0.38956208355635064, "learning_rate": 2.7724248698962386e-06, "loss": 0.2643, "step": 6794 }, { "epoch": 2.1933895637434926, "grad_norm": 0.332169943039374, "learning_rate": 2.770352710118381e-06, "loss": 0.2724, "step": 6795 }, { "epoch": 2.1937124177731144, "grad_norm": 0.37304731221421733, "learning_rate": 2.7682811495803225e-06, "loss": 0.2192, "step": 6796 }, { "epoch": 2.1940352718027363, "grad_norm": 0.351582178616276, "learning_rate": 2.7662101885445287e-06, "loss": 0.2775, "step": 6797 }, { "epoch": 2.194358125832358, "grad_norm": 0.35650695338659605, "learning_rate": 2.764139827273382e-06, "loss": 0.2831, "step": 6798 }, { "epoch": 2.19468097986198, "grad_norm": 0.40225781443294206, "learning_rate": 2.7620700660291983e-06, "loss": 0.2586, "step": 6799 }, { "epoch": 2.195003833891602, "grad_norm": 0.41030968979290205, "learning_rate": 2.760000905074215e-06, "loss": 0.2219, "step": 6800 }, { "epoch": 2.1953266879212237, "grad_norm": 0.37826537001374494, "learning_rate": 2.7579323446705874e-06, "loss": 0.243, "step": 6801 }, { "epoch": 2.1956495419508455, "grad_norm": 0.6076386050472704, "learning_rate": 2.755864385080403e-06, "loss": 0.2581, "step": 6802 }, { "epoch": 2.1959723959804673, "grad_norm": 0.3602653607564256, "learning_rate": 2.7537970265656666e-06, "loss": 0.2366, "step": 6803 }, { "epoch": 2.196295250010089, "grad_norm": 0.3826775385789886, "learning_rate": 2.751730269388313e-06, "loss": 0.2714, "step": 6804 }, { "epoch": 2.196618104039711, "grad_norm": 0.3387104995560595, "learning_rate": 2.7496641138101936e-06, "loss": 0.2285, "step": 6805 }, { "epoch": 2.196940958069333, "grad_norm": 0.37580137840270894, "learning_rate": 2.7475985600930906e-06, "loss": 0.3115, "step": 6806 }, { "epoch": 2.1972638120989547, "grad_norm": 0.34670635166610375, "learning_rate": 2.745533608498704e-06, "loss": 0.2581, "step": 6807 }, { "epoch": 2.1975866661285766, "grad_norm": 0.3073733759384738, "learning_rate": 2.7434692592886646e-06, "loss": 0.2372, "step": 6808 }, { "epoch": 2.1979095201581984, "grad_norm": 0.35858984649289605, "learning_rate": 2.7414055127245175e-06, "loss": 0.2728, "step": 6809 }, { "epoch": 2.1982323741878202, "grad_norm": 0.3776507514732668, "learning_rate": 2.7393423690677377e-06, "loss": 0.2566, "step": 6810 }, { "epoch": 2.198555228217442, "grad_norm": 0.35399042654371726, "learning_rate": 2.737279828579728e-06, "loss": 0.2623, "step": 6811 }, { "epoch": 2.198878082247064, "grad_norm": 0.5384074552214538, "learning_rate": 2.7352178915218014e-06, "loss": 0.2705, "step": 6812 }, { "epoch": 2.199200936276686, "grad_norm": 0.35833161063048896, "learning_rate": 2.733156558155208e-06, "loss": 0.2286, "step": 6813 }, { "epoch": 2.1995237903063076, "grad_norm": 0.39789962815040236, "learning_rate": 2.7310958287411154e-06, "loss": 0.2999, "step": 6814 }, { "epoch": 2.1998466443359295, "grad_norm": 0.41946841814433705, "learning_rate": 2.7290357035406125e-06, "loss": 0.2608, "step": 6815 }, { "epoch": 2.2001694983655513, "grad_norm": 0.5471457202366465, "learning_rate": 2.726976182814718e-06, "loss": 0.252, "step": 6816 }, { "epoch": 2.200492352395173, "grad_norm": 0.4219338320041602, "learning_rate": 2.724917266824366e-06, "loss": 0.2184, "step": 6817 }, { "epoch": 2.200815206424795, "grad_norm": 0.4204349300035774, "learning_rate": 2.722858955830421e-06, "loss": 0.2341, "step": 6818 }, { "epoch": 2.201138060454417, "grad_norm": 0.34989036316157207, "learning_rate": 2.72080125009367e-06, "loss": 0.2793, "step": 6819 }, { "epoch": 2.2014609144840387, "grad_norm": 0.38908659301935083, "learning_rate": 2.718744149874816e-06, "loss": 0.2377, "step": 6820 }, { "epoch": 2.2017837685136605, "grad_norm": 0.510496102312081, "learning_rate": 2.716687655434498e-06, "loss": 0.2845, "step": 6821 }, { "epoch": 2.202106622543283, "grad_norm": 0.3846960435023124, "learning_rate": 2.7146317670332638e-06, "loss": 0.2312, "step": 6822 }, { "epoch": 2.2024294765729047, "grad_norm": 0.4119859944352473, "learning_rate": 2.712576484931597e-06, "loss": 0.268, "step": 6823 }, { "epoch": 2.2027523306025265, "grad_norm": 0.34048767167536326, "learning_rate": 2.7105218093898947e-06, "loss": 0.2081, "step": 6824 }, { "epoch": 2.2030751846321484, "grad_norm": 0.33790755094603425, "learning_rate": 2.7084677406684863e-06, "loss": 0.2089, "step": 6825 }, { "epoch": 2.20339803866177, "grad_norm": 0.34190587250472093, "learning_rate": 2.706414279027614e-06, "loss": 0.2158, "step": 6826 }, { "epoch": 2.203720892691392, "grad_norm": 0.3807101649260789, "learning_rate": 2.704361424727452e-06, "loss": 0.2453, "step": 6827 }, { "epoch": 2.204043746721014, "grad_norm": 0.36675028296531487, "learning_rate": 2.7023091780280927e-06, "loss": 0.2961, "step": 6828 }, { "epoch": 2.2043666007506357, "grad_norm": 0.4034703697109923, "learning_rate": 2.7002575391895567e-06, "loss": 0.284, "step": 6829 }, { "epoch": 2.2046894547802576, "grad_norm": 0.31734404074651246, "learning_rate": 2.698206508471778e-06, "loss": 0.2447, "step": 6830 }, { "epoch": 2.2050123088098794, "grad_norm": 0.4906167634319693, "learning_rate": 2.696156086134625e-06, "loss": 0.2769, "step": 6831 }, { "epoch": 2.2053351628395013, "grad_norm": 0.32392721835370314, "learning_rate": 2.694106272437878e-06, "loss": 0.2528, "step": 6832 }, { "epoch": 2.205658016869123, "grad_norm": 0.4193056449942069, "learning_rate": 2.6920570676412508e-06, "loss": 0.2425, "step": 6833 }, { "epoch": 2.205980870898745, "grad_norm": 0.4039546569471564, "learning_rate": 2.6900084720043697e-06, "loss": 0.2439, "step": 6834 }, { "epoch": 2.206303724928367, "grad_norm": 0.36726441288278056, "learning_rate": 2.687960485786794e-06, "loss": 0.2511, "step": 6835 }, { "epoch": 2.2066265789579886, "grad_norm": 0.32374411696575817, "learning_rate": 2.6859131092479956e-06, "loss": 0.2292, "step": 6836 }, { "epoch": 2.2069494329876105, "grad_norm": 0.41287839891745426, "learning_rate": 2.683866342647376e-06, "loss": 0.2147, "step": 6837 }, { "epoch": 2.2072722870172323, "grad_norm": 0.5351158980423035, "learning_rate": 2.6818201862442628e-06, "loss": 0.2778, "step": 6838 }, { "epoch": 2.207595141046854, "grad_norm": 0.39997165217932373, "learning_rate": 2.6797746402978935e-06, "loss": 0.2792, "step": 6839 }, { "epoch": 2.207917995076476, "grad_norm": 0.3984156245142191, "learning_rate": 2.677729705067442e-06, "loss": 0.2344, "step": 6840 }, { "epoch": 2.208240849106098, "grad_norm": 0.4399561295360334, "learning_rate": 2.675685380811993e-06, "loss": 0.2479, "step": 6841 }, { "epoch": 2.2085637031357197, "grad_norm": 0.44435812696189697, "learning_rate": 2.673641667790564e-06, "loss": 0.2598, "step": 6842 }, { "epoch": 2.2088865571653415, "grad_norm": 0.36942175515304143, "learning_rate": 2.6715985662620886e-06, "loss": 0.2933, "step": 6843 }, { "epoch": 2.2092094111949634, "grad_norm": 0.432538890666046, "learning_rate": 2.6695560764854284e-06, "loss": 0.2822, "step": 6844 }, { "epoch": 2.2095322652245852, "grad_norm": 0.38165658583365597, "learning_rate": 2.6675141987193587e-06, "loss": 0.2575, "step": 6845 }, { "epoch": 2.209855119254207, "grad_norm": 0.3321734279166454, "learning_rate": 2.6654729332225885e-06, "loss": 0.3162, "step": 6846 }, { "epoch": 2.210177973283829, "grad_norm": 0.3511916219998523, "learning_rate": 2.6634322802537376e-06, "loss": 0.2483, "step": 6847 }, { "epoch": 2.2105008273134508, "grad_norm": 0.42362017296663773, "learning_rate": 2.661392240071358e-06, "loss": 0.2412, "step": 6848 }, { "epoch": 2.2108236813430726, "grad_norm": 0.336919257945274, "learning_rate": 2.6593528129339168e-06, "loss": 0.2537, "step": 6849 }, { "epoch": 2.2111465353726945, "grad_norm": 0.34486982286964063, "learning_rate": 2.65731399909981e-06, "loss": 0.2522, "step": 6850 }, { "epoch": 2.2114693894023163, "grad_norm": 0.4022993101129789, "learning_rate": 2.6552757988273495e-06, "loss": 0.2927, "step": 6851 }, { "epoch": 2.211792243431938, "grad_norm": 0.3842096049315157, "learning_rate": 2.6532382123747752e-06, "loss": 0.2433, "step": 6852 }, { "epoch": 2.21211509746156, "grad_norm": 0.4330593918038612, "learning_rate": 2.6512012400002426e-06, "loss": 0.2816, "step": 6853 }, { "epoch": 2.2124379514911823, "grad_norm": 0.357717874653846, "learning_rate": 2.649164881961835e-06, "loss": 0.2875, "step": 6854 }, { "epoch": 2.212760805520804, "grad_norm": 0.3680720941162848, "learning_rate": 2.6471291385175605e-06, "loss": 0.2821, "step": 6855 }, { "epoch": 2.213083659550426, "grad_norm": 0.4383425762763814, "learning_rate": 2.6450940099253387e-06, "loss": 0.2608, "step": 6856 }, { "epoch": 2.213406513580048, "grad_norm": 0.41157340779830254, "learning_rate": 2.6430594964430226e-06, "loss": 0.2318, "step": 6857 }, { "epoch": 2.2137293676096697, "grad_norm": 0.3982929676206104, "learning_rate": 2.641025598328377e-06, "loss": 0.2439, "step": 6858 }, { "epoch": 2.2140522216392915, "grad_norm": 0.3788931047865837, "learning_rate": 2.6389923158390973e-06, "loss": 0.2448, "step": 6859 }, { "epoch": 2.2143750756689133, "grad_norm": 0.37050322035357625, "learning_rate": 2.636959649232797e-06, "loss": 0.2813, "step": 6860 }, { "epoch": 2.214697929698535, "grad_norm": 0.39974891773816845, "learning_rate": 2.634927598767015e-06, "loss": 0.2812, "step": 6861 }, { "epoch": 2.215020783728157, "grad_norm": 0.4830146628612394, "learning_rate": 2.632896164699204e-06, "loss": 0.2666, "step": 6862 }, { "epoch": 2.215343637757779, "grad_norm": 0.4108625611414314, "learning_rate": 2.630865347286749e-06, "loss": 0.2964, "step": 6863 }, { "epoch": 2.2156664917874007, "grad_norm": 0.4605126688219696, "learning_rate": 2.6288351467869473e-06, "loss": 0.3084, "step": 6864 }, { "epoch": 2.2159893458170226, "grad_norm": 0.38210280183580414, "learning_rate": 2.626805563457026e-06, "loss": 0.2206, "step": 6865 }, { "epoch": 2.2163121998466444, "grad_norm": 0.37542719846011646, "learning_rate": 2.6247765975541272e-06, "loss": 0.2363, "step": 6866 }, { "epoch": 2.2166350538762662, "grad_norm": 0.3513890843005004, "learning_rate": 2.622748249335323e-06, "loss": 0.2642, "step": 6867 }, { "epoch": 2.216957907905888, "grad_norm": 0.3843311629021371, "learning_rate": 2.6207205190575954e-06, "loss": 0.2431, "step": 6868 }, { "epoch": 2.21728076193551, "grad_norm": 0.31916067198457315, "learning_rate": 2.618693406977862e-06, "loss": 0.2633, "step": 6869 }, { "epoch": 2.217603615965132, "grad_norm": 0.3560563448386159, "learning_rate": 2.61666691335295e-06, "loss": 0.2437, "step": 6870 }, { "epoch": 2.2179264699947536, "grad_norm": 0.36614236970268493, "learning_rate": 2.614641038439617e-06, "loss": 0.2234, "step": 6871 }, { "epoch": 2.2182493240243755, "grad_norm": 0.3887062158702349, "learning_rate": 2.6126157824945347e-06, "loss": 0.2676, "step": 6872 }, { "epoch": 2.2185721780539973, "grad_norm": 0.3705676699383131, "learning_rate": 2.610591145774302e-06, "loss": 0.2848, "step": 6873 }, { "epoch": 2.218895032083619, "grad_norm": 0.40751150154185883, "learning_rate": 2.608567128535439e-06, "loss": 0.2717, "step": 6874 }, { "epoch": 2.219217886113241, "grad_norm": 0.372607186941362, "learning_rate": 2.606543731034384e-06, "loss": 0.2811, "step": 6875 }, { "epoch": 2.219540740142863, "grad_norm": 0.316431475242907, "learning_rate": 2.6045209535275028e-06, "loss": 0.2355, "step": 6876 }, { "epoch": 2.2198635941724847, "grad_norm": 0.36825382571101906, "learning_rate": 2.6024987962710725e-06, "loss": 0.286, "step": 6877 }, { "epoch": 2.2201864482021065, "grad_norm": 0.4375589637966896, "learning_rate": 2.6004772595213023e-06, "loss": 0.2769, "step": 6878 }, { "epoch": 2.2205093022317284, "grad_norm": 0.3851920777975648, "learning_rate": 2.5984563435343146e-06, "loss": 0.3234, "step": 6879 }, { "epoch": 2.22083215626135, "grad_norm": 0.4986048842765633, "learning_rate": 2.5964360485661604e-06, "loss": 0.2988, "step": 6880 }, { "epoch": 2.221155010290972, "grad_norm": 0.32687646615542865, "learning_rate": 2.5944163748728043e-06, "loss": 0.2561, "step": 6881 }, { "epoch": 2.221477864320594, "grad_norm": 0.4695654204476028, "learning_rate": 2.5923973227101403e-06, "loss": 0.2684, "step": 6882 }, { "epoch": 2.2218007183502158, "grad_norm": 0.4134993390790513, "learning_rate": 2.590378892333975e-06, "loss": 0.2622, "step": 6883 }, { "epoch": 2.2221235723798376, "grad_norm": 0.3711646509702942, "learning_rate": 2.588361084000045e-06, "loss": 0.3024, "step": 6884 }, { "epoch": 2.2224464264094594, "grad_norm": 0.37678472827221166, "learning_rate": 2.586343897964e-06, "loss": 0.2767, "step": 6885 }, { "epoch": 2.2227692804390813, "grad_norm": 0.4637585292284578, "learning_rate": 2.584327334481419e-06, "loss": 0.247, "step": 6886 }, { "epoch": 2.223092134468703, "grad_norm": 0.36351598817367214, "learning_rate": 2.582311393807793e-06, "loss": 0.2653, "step": 6887 }, { "epoch": 2.223414988498325, "grad_norm": 0.38855789774875654, "learning_rate": 2.5802960761985414e-06, "loss": 0.2087, "step": 6888 }, { "epoch": 2.223737842527947, "grad_norm": 0.4050603791546673, "learning_rate": 2.578281381909004e-06, "loss": 0.2061, "step": 6889 }, { "epoch": 2.224060696557569, "grad_norm": 0.40517805883068575, "learning_rate": 2.576267311194436e-06, "loss": 0.2368, "step": 6890 }, { "epoch": 2.224383550587191, "grad_norm": 0.3762012635820736, "learning_rate": 2.574253864310019e-06, "loss": 0.21, "step": 6891 }, { "epoch": 2.224706404616813, "grad_norm": 0.4776373120971292, "learning_rate": 2.5722410415108543e-06, "loss": 0.2711, "step": 6892 }, { "epoch": 2.2250292586464346, "grad_norm": 0.35844064909632056, "learning_rate": 2.5702288430519664e-06, "loss": 0.2195, "step": 6893 }, { "epoch": 2.2253521126760565, "grad_norm": 0.3966713885785837, "learning_rate": 2.5682172691882926e-06, "loss": 0.2611, "step": 6894 }, { "epoch": 2.2256749667056783, "grad_norm": 0.3468220934532249, "learning_rate": 2.566206320174702e-06, "loss": 0.2445, "step": 6895 }, { "epoch": 2.2259978207353, "grad_norm": 0.31574696473819724, "learning_rate": 2.5641959962659745e-06, "loss": 0.2503, "step": 6896 }, { "epoch": 2.226320674764922, "grad_norm": 0.37556391402286954, "learning_rate": 2.5621862977168185e-06, "loss": 0.2321, "step": 6897 }, { "epoch": 2.226643528794544, "grad_norm": 0.3539251210005554, "learning_rate": 2.5601772247818568e-06, "loss": 0.2095, "step": 6898 }, { "epoch": 2.2269663828241657, "grad_norm": 0.38372459388094127, "learning_rate": 2.5581687777156417e-06, "loss": 0.2847, "step": 6899 }, { "epoch": 2.2272892368537875, "grad_norm": 0.3333657547405769, "learning_rate": 2.5561609567726334e-06, "loss": 0.2148, "step": 6900 }, { "epoch": 2.2276120908834094, "grad_norm": 0.42692401770561117, "learning_rate": 2.554153762207227e-06, "loss": 0.2999, "step": 6901 }, { "epoch": 2.2279349449130312, "grad_norm": 0.4704807702213246, "learning_rate": 2.552147194273725e-06, "loss": 0.3344, "step": 6902 }, { "epoch": 2.228257798942653, "grad_norm": 0.3294337169750007, "learning_rate": 2.5501412532263628e-06, "loss": 0.2413, "step": 6903 }, { "epoch": 2.228580652972275, "grad_norm": 0.3030654582673265, "learning_rate": 2.5481359393192847e-06, "loss": 0.2367, "step": 6904 }, { "epoch": 2.2289035070018968, "grad_norm": 0.37833458091372035, "learning_rate": 2.5461312528065638e-06, "loss": 0.2303, "step": 6905 }, { "epoch": 2.2292263610315186, "grad_norm": 0.3728116901591539, "learning_rate": 2.544127193942192e-06, "loss": 0.2794, "step": 6906 }, { "epoch": 2.2295492150611405, "grad_norm": 0.411495416142167, "learning_rate": 2.5421237629800817e-06, "loss": 0.2623, "step": 6907 }, { "epoch": 2.2298720690907623, "grad_norm": 0.32102943256216143, "learning_rate": 2.5401209601740613e-06, "loss": 0.2208, "step": 6908 }, { "epoch": 2.230194923120384, "grad_norm": 0.35002027885516346, "learning_rate": 2.5381187857778877e-06, "loss": 0.2838, "step": 6909 }, { "epoch": 2.230517777150006, "grad_norm": 0.4089689826713001, "learning_rate": 2.5361172400452286e-06, "loss": 0.2003, "step": 6910 }, { "epoch": 2.230840631179628, "grad_norm": 0.35567725899233626, "learning_rate": 2.5341163232296785e-06, "loss": 0.2618, "step": 6911 }, { "epoch": 2.2311634852092497, "grad_norm": 0.37728252817568086, "learning_rate": 2.532116035584755e-06, "loss": 0.2518, "step": 6912 }, { "epoch": 2.2314863392388715, "grad_norm": 0.31140501283481903, "learning_rate": 2.5301163773638868e-06, "loss": 0.2488, "step": 6913 }, { "epoch": 2.2318091932684934, "grad_norm": 0.580706037469492, "learning_rate": 2.5281173488204317e-06, "loss": 0.2325, "step": 6914 }, { "epoch": 2.232132047298115, "grad_norm": 0.38523240653824076, "learning_rate": 2.5261189502076588e-06, "loss": 0.2578, "step": 6915 }, { "epoch": 2.232454901327737, "grad_norm": 0.38348210890422024, "learning_rate": 2.5241211817787686e-06, "loss": 0.2303, "step": 6916 }, { "epoch": 2.232777755357359, "grad_norm": 0.3821061020100672, "learning_rate": 2.522124043786869e-06, "loss": 0.2795, "step": 6917 }, { "epoch": 2.2331006093869807, "grad_norm": 0.3813191683776032, "learning_rate": 2.520127536485001e-06, "loss": 0.2184, "step": 6918 }, { "epoch": 2.2334234634166026, "grad_norm": 0.3711404617100866, "learning_rate": 2.518131660126115e-06, "loss": 0.2037, "step": 6919 }, { "epoch": 2.2337463174462244, "grad_norm": 0.40669834065347776, "learning_rate": 2.516136414963086e-06, "loss": 0.279, "step": 6920 }, { "epoch": 2.2340691714758467, "grad_norm": 0.3492729134082394, "learning_rate": 2.51414180124871e-06, "loss": 0.2583, "step": 6921 }, { "epoch": 2.2343920255054686, "grad_norm": 0.34818414580914553, "learning_rate": 2.512147819235705e-06, "loss": 0.2356, "step": 6922 }, { "epoch": 2.2347148795350904, "grad_norm": 0.40862197212168716, "learning_rate": 2.5101544691766996e-06, "loss": 0.2552, "step": 6923 }, { "epoch": 2.2350377335647122, "grad_norm": 0.3517524134173481, "learning_rate": 2.508161751324255e-06, "loss": 0.2573, "step": 6924 }, { "epoch": 2.235360587594334, "grad_norm": 0.3460622791858407, "learning_rate": 2.5061696659308394e-06, "loss": 0.2198, "step": 6925 }, { "epoch": 2.235683441623956, "grad_norm": 0.3454036172619398, "learning_rate": 2.504178213248852e-06, "loss": 0.2728, "step": 6926 }, { "epoch": 2.236006295653578, "grad_norm": 0.3772719988034209, "learning_rate": 2.5021873935306037e-06, "loss": 0.2389, "step": 6927 }, { "epoch": 2.2363291496831996, "grad_norm": 0.37019633664091517, "learning_rate": 2.5001972070283294e-06, "loss": 0.2441, "step": 6928 }, { "epoch": 2.2366520037128215, "grad_norm": 0.33906232643669876, "learning_rate": 2.498207653994187e-06, "loss": 0.2443, "step": 6929 }, { "epoch": 2.2369748577424433, "grad_norm": 0.3771090712470478, "learning_rate": 2.4962187346802444e-06, "loss": 0.2544, "step": 6930 }, { "epoch": 2.237297711772065, "grad_norm": 0.36708912958756723, "learning_rate": 2.4942304493384998e-06, "loss": 0.2628, "step": 6931 }, { "epoch": 2.237620565801687, "grad_norm": 0.46638139077975377, "learning_rate": 2.4922427982208614e-06, "loss": 0.239, "step": 6932 }, { "epoch": 2.237943419831309, "grad_norm": 0.3839761774583142, "learning_rate": 2.490255781579167e-06, "loss": 0.2309, "step": 6933 }, { "epoch": 2.2382662738609307, "grad_norm": 0.3443386373674109, "learning_rate": 2.4882693996651636e-06, "loss": 0.2638, "step": 6934 }, { "epoch": 2.2385891278905525, "grad_norm": 0.363497022860873, "learning_rate": 2.4862836527305284e-06, "loss": 0.2627, "step": 6935 }, { "epoch": 2.2389119819201744, "grad_norm": 0.4014561012929826, "learning_rate": 2.4842985410268477e-06, "loss": 0.2496, "step": 6936 }, { "epoch": 2.239234835949796, "grad_norm": 0.3719650618316768, "learning_rate": 2.4823140648056343e-06, "loss": 0.2591, "step": 6937 }, { "epoch": 2.239557689979418, "grad_norm": 0.4171336387286256, "learning_rate": 2.4803302243183194e-06, "loss": 0.2995, "step": 6938 }, { "epoch": 2.23988054400904, "grad_norm": 0.3430289614967815, "learning_rate": 2.478347019816255e-06, "loss": 0.2742, "step": 6939 }, { "epoch": 2.2402033980386618, "grad_norm": 0.3674999800376065, "learning_rate": 2.4763644515507057e-06, "loss": 0.2306, "step": 6940 }, { "epoch": 2.2405262520682836, "grad_norm": 0.39054468670618314, "learning_rate": 2.4743825197728652e-06, "loss": 0.2813, "step": 6941 }, { "epoch": 2.2408491060979054, "grad_norm": 0.38217724645311807, "learning_rate": 2.4724012247338355e-06, "loss": 0.2131, "step": 6942 }, { "epoch": 2.2411719601275273, "grad_norm": 0.3587564037367285, "learning_rate": 2.47042056668465e-06, "loss": 0.2691, "step": 6943 }, { "epoch": 2.241494814157149, "grad_norm": 0.4818197454064042, "learning_rate": 2.4684405458762505e-06, "loss": 0.2824, "step": 6944 }, { "epoch": 2.241817668186771, "grad_norm": 0.3255712402368205, "learning_rate": 2.4664611625595063e-06, "loss": 0.2481, "step": 6945 }, { "epoch": 2.242140522216393, "grad_norm": 0.3543444102486457, "learning_rate": 2.4644824169852e-06, "loss": 0.2699, "step": 6946 }, { "epoch": 2.2424633762460147, "grad_norm": 0.4278185170835332, "learning_rate": 2.4625043094040364e-06, "loss": 0.2467, "step": 6947 }, { "epoch": 2.2427862302756365, "grad_norm": 0.46199543168475193, "learning_rate": 2.4605268400666424e-06, "loss": 0.2198, "step": 6948 }, { "epoch": 2.2431090843052583, "grad_norm": 0.3963796299761853, "learning_rate": 2.4585500092235556e-06, "loss": 0.2769, "step": 6949 }, { "epoch": 2.24343193833488, "grad_norm": 0.4435400111335402, "learning_rate": 2.4565738171252424e-06, "loss": 0.2482, "step": 6950 }, { "epoch": 2.243754792364502, "grad_norm": 0.39715742062896936, "learning_rate": 2.4545982640220794e-06, "loss": 0.2693, "step": 6951 }, { "epoch": 2.244077646394124, "grad_norm": 0.3553923491312915, "learning_rate": 2.4526233501643683e-06, "loss": 0.2539, "step": 6952 }, { "epoch": 2.2444005004237457, "grad_norm": 0.368523098126061, "learning_rate": 2.4506490758023275e-06, "loss": 0.263, "step": 6953 }, { "epoch": 2.2447233544533676, "grad_norm": 0.31646980577261863, "learning_rate": 2.448675441186099e-06, "loss": 0.2566, "step": 6954 }, { "epoch": 2.2450462084829894, "grad_norm": 0.33741073825386403, "learning_rate": 2.4467024465657324e-06, "loss": 0.2714, "step": 6955 }, { "epoch": 2.2453690625126113, "grad_norm": 0.3568295199066921, "learning_rate": 2.44473009219121e-06, "loss": 0.2677, "step": 6956 }, { "epoch": 2.2456919165422335, "grad_norm": 0.3910824435706705, "learning_rate": 2.4427583783124206e-06, "loss": 0.3108, "step": 6957 }, { "epoch": 2.2460147705718554, "grad_norm": 0.41523473941583766, "learning_rate": 2.4407873051791835e-06, "loss": 0.2458, "step": 6958 }, { "epoch": 2.2463376246014772, "grad_norm": 0.3893902137418487, "learning_rate": 2.438816873041225e-06, "loss": 0.2224, "step": 6959 }, { "epoch": 2.246660478631099, "grad_norm": 0.45879991427376315, "learning_rate": 2.436847082148202e-06, "loss": 0.2204, "step": 6960 }, { "epoch": 2.246983332660721, "grad_norm": 0.37749050122364153, "learning_rate": 2.434877932749678e-06, "loss": 0.22, "step": 6961 }, { "epoch": 2.2473061866903428, "grad_norm": 0.38167488493641083, "learning_rate": 2.432909425095148e-06, "loss": 0.2583, "step": 6962 }, { "epoch": 2.2476290407199646, "grad_norm": 0.41797125570990395, "learning_rate": 2.4309415594340133e-06, "loss": 0.2489, "step": 6963 }, { "epoch": 2.2479518947495865, "grad_norm": 0.4543766395087378, "learning_rate": 2.4289743360156045e-06, "loss": 0.2328, "step": 6964 }, { "epoch": 2.2482747487792083, "grad_norm": 0.3575732898416859, "learning_rate": 2.4270077550891626e-06, "loss": 0.2955, "step": 6965 }, { "epoch": 2.24859760280883, "grad_norm": 0.5310681472720264, "learning_rate": 2.4250418169038516e-06, "loss": 0.2727, "step": 6966 }, { "epoch": 2.248920456838452, "grad_norm": 0.4156507017183807, "learning_rate": 2.423076521708753e-06, "loss": 0.2764, "step": 6967 }, { "epoch": 2.249243310868074, "grad_norm": 0.37585364148650185, "learning_rate": 2.421111869752872e-06, "loss": 0.2146, "step": 6968 }, { "epoch": 2.2495661648976957, "grad_norm": 0.39429142888104646, "learning_rate": 2.419147861285119e-06, "loss": 0.2668, "step": 6969 }, { "epoch": 2.2498890189273175, "grad_norm": 0.33338396174019763, "learning_rate": 2.417184496554336e-06, "loss": 0.2302, "step": 6970 }, { "epoch": 2.2502118729569394, "grad_norm": 0.41029971409946475, "learning_rate": 2.41522177580928e-06, "loss": 0.2204, "step": 6971 }, { "epoch": 2.250534726986561, "grad_norm": 0.4542413247455814, "learning_rate": 2.413259699298621e-06, "loss": 0.3313, "step": 6972 }, { "epoch": 2.250857581016183, "grad_norm": 0.3509235243532192, "learning_rate": 2.411298267270955e-06, "loss": 0.2762, "step": 6973 }, { "epoch": 2.251180435045805, "grad_norm": 0.3814648810910828, "learning_rate": 2.4093374799747893e-06, "loss": 0.2298, "step": 6974 }, { "epoch": 2.2515032890754267, "grad_norm": 0.3900717479922621, "learning_rate": 2.4073773376585566e-06, "loss": 0.2462, "step": 6975 }, { "epoch": 2.2518261431050486, "grad_norm": 0.40851249253355987, "learning_rate": 2.4054178405705993e-06, "loss": 0.2664, "step": 6976 }, { "epoch": 2.2521489971346704, "grad_norm": 0.3663973118630681, "learning_rate": 2.4034589889591893e-06, "loss": 0.264, "step": 6977 }, { "epoch": 2.2524718511642923, "grad_norm": 0.3554472708103709, "learning_rate": 2.4015007830725037e-06, "loss": 0.278, "step": 6978 }, { "epoch": 2.252794705193914, "grad_norm": 0.3582683761858415, "learning_rate": 2.3995432231586512e-06, "loss": 0.2103, "step": 6979 }, { "epoch": 2.253117559223536, "grad_norm": 0.6367553382117651, "learning_rate": 2.397586309465646e-06, "loss": 0.296, "step": 6980 }, { "epoch": 2.253440413253158, "grad_norm": 0.38460280751327036, "learning_rate": 2.3956300422414314e-06, "loss": 0.2377, "step": 6981 }, { "epoch": 2.2537632672827796, "grad_norm": 0.35506183372687805, "learning_rate": 2.3936744217338582e-06, "loss": 0.2628, "step": 6982 }, { "epoch": 2.2540861213124015, "grad_norm": 0.3758329687516019, "learning_rate": 2.391719448190704e-06, "loss": 0.2364, "step": 6983 }, { "epoch": 2.2544089753420233, "grad_norm": 0.5085168956108281, "learning_rate": 2.3897651218596618e-06, "loss": 0.2579, "step": 6984 }, { "epoch": 2.254731829371645, "grad_norm": 0.37524527871777824, "learning_rate": 2.387811442988341e-06, "loss": 0.257, "step": 6985 }, { "epoch": 2.255054683401267, "grad_norm": 0.39639250509498264, "learning_rate": 2.385858411824273e-06, "loss": 0.2476, "step": 6986 }, { "epoch": 2.255377537430889, "grad_norm": 0.4632058864735315, "learning_rate": 2.383906028614899e-06, "loss": 0.2049, "step": 6987 }, { "epoch": 2.255700391460511, "grad_norm": 0.350335009503596, "learning_rate": 2.3819542936075894e-06, "loss": 0.2541, "step": 6988 }, { "epoch": 2.256023245490133, "grad_norm": 0.3745342450335603, "learning_rate": 2.3800032070496187e-06, "loss": 0.25, "step": 6989 }, { "epoch": 2.256346099519755, "grad_norm": 0.40790379866594373, "learning_rate": 2.3780527691881944e-06, "loss": 0.2211, "step": 6990 }, { "epoch": 2.2566689535493767, "grad_norm": 0.4230336116906536, "learning_rate": 2.3761029802704286e-06, "loss": 0.2891, "step": 6991 }, { "epoch": 2.2569918075789985, "grad_norm": 0.3473955595616448, "learning_rate": 2.374153840543361e-06, "loss": 0.2585, "step": 6992 }, { "epoch": 2.2573146616086204, "grad_norm": 0.38045592568705977, "learning_rate": 2.3722053502539414e-06, "loss": 0.2582, "step": 6993 }, { "epoch": 2.257637515638242, "grad_norm": 0.37015019426963147, "learning_rate": 2.370257509649045e-06, "loss": 0.2686, "step": 6994 }, { "epoch": 2.257960369667864, "grad_norm": 0.6234563253716616, "learning_rate": 2.368310318975455e-06, "loss": 0.2579, "step": 6995 }, { "epoch": 2.258283223697486, "grad_norm": 0.39736399438143016, "learning_rate": 2.3663637784798836e-06, "loss": 0.2636, "step": 6996 }, { "epoch": 2.2586060777271078, "grad_norm": 0.37670794212764824, "learning_rate": 2.3644178884089495e-06, "loss": 0.2547, "step": 6997 }, { "epoch": 2.2589289317567296, "grad_norm": 0.389513159732303, "learning_rate": 2.3624726490091967e-06, "loss": 0.2249, "step": 6998 }, { "epoch": 2.2592517857863514, "grad_norm": 0.41836305227507165, "learning_rate": 2.3605280605270843e-06, "loss": 0.3146, "step": 6999 }, { "epoch": 2.2595746398159733, "grad_norm": 0.3687753219806622, "learning_rate": 2.358584123208992e-06, "loss": 0.2957, "step": 7000 }, { "epoch": 2.259897493845595, "grad_norm": 0.3575133149529602, "learning_rate": 2.356640837301208e-06, "loss": 0.2683, "step": 7001 }, { "epoch": 2.260220347875217, "grad_norm": 0.35821158275595966, "learning_rate": 2.3546982030499483e-06, "loss": 0.2456, "step": 7002 }, { "epoch": 2.260543201904839, "grad_norm": 0.36542757391585245, "learning_rate": 2.352756220701342e-06, "loss": 0.3402, "step": 7003 }, { "epoch": 2.2608660559344607, "grad_norm": 0.4170516362655452, "learning_rate": 2.3508148905014334e-06, "loss": 0.2718, "step": 7004 }, { "epoch": 2.2611889099640825, "grad_norm": 0.40830592004405913, "learning_rate": 2.3488742126961884e-06, "loss": 0.2459, "step": 7005 }, { "epoch": 2.2615117639937043, "grad_norm": 0.3856448094545653, "learning_rate": 2.3469341875314848e-06, "loss": 0.2657, "step": 7006 }, { "epoch": 2.261834618023326, "grad_norm": 0.3414043453432668, "learning_rate": 2.3449948152531264e-06, "loss": 0.2102, "step": 7007 }, { "epoch": 2.262157472052948, "grad_norm": 0.3546763710831729, "learning_rate": 2.343056096106824e-06, "loss": 0.2741, "step": 7008 }, { "epoch": 2.26248032608257, "grad_norm": 0.3472881168747727, "learning_rate": 2.341118030338214e-06, "loss": 0.2198, "step": 7009 }, { "epoch": 2.2628031801121917, "grad_norm": 0.45224606910830484, "learning_rate": 2.3391806181928433e-06, "loss": 0.2403, "step": 7010 }, { "epoch": 2.2631260341418136, "grad_norm": 0.4796867147869485, "learning_rate": 2.3372438599161845e-06, "loss": 0.2149, "step": 7011 }, { "epoch": 2.2634488881714354, "grad_norm": 0.37000850309653893, "learning_rate": 2.3353077557536156e-06, "loss": 0.3045, "step": 7012 }, { "epoch": 2.2637717422010573, "grad_norm": 0.3337936889865001, "learning_rate": 2.3333723059504453e-06, "loss": 0.2401, "step": 7013 }, { "epoch": 2.264094596230679, "grad_norm": 0.42927597461701333, "learning_rate": 2.3314375107518863e-06, "loss": 0.2762, "step": 7014 }, { "epoch": 2.264417450260301, "grad_norm": 0.3900003399667806, "learning_rate": 2.3295033704030768e-06, "loss": 0.2669, "step": 7015 }, { "epoch": 2.264740304289923, "grad_norm": 0.44159826891820947, "learning_rate": 2.3275698851490705e-06, "loss": 0.2673, "step": 7016 }, { "epoch": 2.2650631583195446, "grad_norm": 0.3573097095801262, "learning_rate": 2.3256370552348383e-06, "loss": 0.2404, "step": 7017 }, { "epoch": 2.2653860123491665, "grad_norm": 0.4293502803325641, "learning_rate": 2.3237048809052646e-06, "loss": 0.2453, "step": 7018 }, { "epoch": 2.2657088663787883, "grad_norm": 0.3858691984630929, "learning_rate": 2.3217733624051563e-06, "loss": 0.2364, "step": 7019 }, { "epoch": 2.26603172040841, "grad_norm": 0.34384617728594463, "learning_rate": 2.319842499979229e-06, "loss": 0.2342, "step": 7020 }, { "epoch": 2.266354574438032, "grad_norm": 0.41896343321241397, "learning_rate": 2.317912293872124e-06, "loss": 0.2643, "step": 7021 }, { "epoch": 2.266677428467654, "grad_norm": 0.39853521385208845, "learning_rate": 2.3159827443283972e-06, "loss": 0.3114, "step": 7022 }, { "epoch": 2.2670002824972757, "grad_norm": 0.39644378342282427, "learning_rate": 2.314053851592516e-06, "loss": 0.245, "step": 7023 }, { "epoch": 2.2673231365268975, "grad_norm": 0.39272215726252274, "learning_rate": 2.3121256159088733e-06, "loss": 0.2257, "step": 7024 }, { "epoch": 2.2676459905565194, "grad_norm": 0.42144170999161135, "learning_rate": 2.3101980375217675e-06, "loss": 0.2424, "step": 7025 }, { "epoch": 2.2679688445861417, "grad_norm": 0.386741548294005, "learning_rate": 2.3082711166754264e-06, "loss": 0.2358, "step": 7026 }, { "epoch": 2.2682916986157635, "grad_norm": 0.41042556393147195, "learning_rate": 2.3063448536139827e-06, "loss": 0.2258, "step": 7027 }, { "epoch": 2.2686145526453854, "grad_norm": 0.3985278591915253, "learning_rate": 2.3044192485814964e-06, "loss": 0.2302, "step": 7028 }, { "epoch": 2.268937406675007, "grad_norm": 0.35456495690000045, "learning_rate": 2.3024943018219345e-06, "loss": 0.211, "step": 7029 }, { "epoch": 2.269260260704629, "grad_norm": 0.39227671162239613, "learning_rate": 2.3005700135791876e-06, "loss": 0.255, "step": 7030 }, { "epoch": 2.269583114734251, "grad_norm": 0.3272556567491197, "learning_rate": 2.298646384097059e-06, "loss": 0.222, "step": 7031 }, { "epoch": 2.2699059687638727, "grad_norm": 0.37299118583860214, "learning_rate": 2.2967234136192735e-06, "loss": 0.2411, "step": 7032 }, { "epoch": 2.2702288227934946, "grad_norm": 0.38225897556270183, "learning_rate": 2.2948011023894636e-06, "loss": 0.2485, "step": 7033 }, { "epoch": 2.2705516768231164, "grad_norm": 0.3465234687366213, "learning_rate": 2.292879450651189e-06, "loss": 0.2599, "step": 7034 }, { "epoch": 2.2708745308527383, "grad_norm": 0.3935017583459948, "learning_rate": 2.290958458647914e-06, "loss": 0.2254, "step": 7035 }, { "epoch": 2.27119738488236, "grad_norm": 0.32814493962770047, "learning_rate": 2.289038126623032e-06, "loss": 0.2622, "step": 7036 }, { "epoch": 2.271520238911982, "grad_norm": 0.3933257177867392, "learning_rate": 2.287118454819842e-06, "loss": 0.2839, "step": 7037 }, { "epoch": 2.271843092941604, "grad_norm": 0.4264847334528408, "learning_rate": 2.2851994434815665e-06, "loss": 0.2372, "step": 7038 }, { "epoch": 2.2721659469712256, "grad_norm": 0.4184523643095531, "learning_rate": 2.283281092851338e-06, "loss": 0.293, "step": 7039 }, { "epoch": 2.2724888010008475, "grad_norm": 0.43933854565758895, "learning_rate": 2.2813634031722113e-06, "loss": 0.2716, "step": 7040 }, { "epoch": 2.2728116550304693, "grad_norm": 0.4238305268858692, "learning_rate": 2.2794463746871568e-06, "loss": 0.2983, "step": 7041 }, { "epoch": 2.273134509060091, "grad_norm": 0.32629830914271696, "learning_rate": 2.2775300076390558e-06, "loss": 0.2676, "step": 7042 }, { "epoch": 2.273457363089713, "grad_norm": 0.35849386263309846, "learning_rate": 2.2756143022707133e-06, "loss": 0.2761, "step": 7043 }, { "epoch": 2.273780217119335, "grad_norm": 0.38938438859523933, "learning_rate": 2.273699258824842e-06, "loss": 0.2427, "step": 7044 }, { "epoch": 2.2741030711489567, "grad_norm": 0.3716294945726607, "learning_rate": 2.2717848775440773e-06, "loss": 0.2641, "step": 7045 }, { "epoch": 2.2744259251785786, "grad_norm": 0.40454793449508675, "learning_rate": 2.2698711586709715e-06, "loss": 0.2693, "step": 7046 }, { "epoch": 2.2747487792082004, "grad_norm": 0.3547740688266091, "learning_rate": 2.2679581024479847e-06, "loss": 0.2281, "step": 7047 }, { "epoch": 2.2750716332378222, "grad_norm": 0.3975989345985831, "learning_rate": 2.266045709117502e-06, "loss": 0.2372, "step": 7048 }, { "epoch": 2.275394487267444, "grad_norm": 0.38042594433987686, "learning_rate": 2.2641339789218224e-06, "loss": 0.2426, "step": 7049 }, { "epoch": 2.275717341297066, "grad_norm": 0.34837578561414184, "learning_rate": 2.262222912103156e-06, "loss": 0.2713, "step": 7050 }, { "epoch": 2.2760401953266878, "grad_norm": 0.3543544818199799, "learning_rate": 2.260312508903635e-06, "loss": 0.2696, "step": 7051 }, { "epoch": 2.2763630493563096, "grad_norm": 0.36068981898136343, "learning_rate": 2.2584027695653018e-06, "loss": 0.3316, "step": 7052 }, { "epoch": 2.2766859033859315, "grad_norm": 0.3700414319034842, "learning_rate": 2.256493694330122e-06, "loss": 0.298, "step": 7053 }, { "epoch": 2.2770087574155533, "grad_norm": 0.47293694620410476, "learning_rate": 2.2545852834399676e-06, "loss": 0.2594, "step": 7054 }, { "epoch": 2.2773316114451756, "grad_norm": 0.3192636218078918, "learning_rate": 2.252677537136637e-06, "loss": 0.2099, "step": 7055 }, { "epoch": 2.2776544654747974, "grad_norm": 0.3612936383216716, "learning_rate": 2.250770455661834e-06, "loss": 0.2336, "step": 7056 }, { "epoch": 2.2779773195044193, "grad_norm": 0.33424881259542694, "learning_rate": 2.2488640392571853e-06, "loss": 0.2296, "step": 7057 }, { "epoch": 2.278300173534041, "grad_norm": 0.3244604935912029, "learning_rate": 2.246958288164235e-06, "loss": 0.2361, "step": 7058 }, { "epoch": 2.278623027563663, "grad_norm": 0.3761595555958373, "learning_rate": 2.245053202624433e-06, "loss": 0.2338, "step": 7059 }, { "epoch": 2.278945881593285, "grad_norm": 0.3734135843761367, "learning_rate": 2.2431487828791557e-06, "loss": 0.2836, "step": 7060 }, { "epoch": 2.2792687356229067, "grad_norm": 0.3662606043357737, "learning_rate": 2.2412450291696865e-06, "loss": 0.2921, "step": 7061 }, { "epoch": 2.2795915896525285, "grad_norm": 0.3973402318221695, "learning_rate": 2.2393419417372303e-06, "loss": 0.2592, "step": 7062 }, { "epoch": 2.2799144436821503, "grad_norm": 0.5001436785879977, "learning_rate": 2.2374395208229054e-06, "loss": 0.3468, "step": 7063 }, { "epoch": 2.280237297711772, "grad_norm": 0.3562968393076857, "learning_rate": 2.2355377666677483e-06, "loss": 0.277, "step": 7064 }, { "epoch": 2.280560151741394, "grad_norm": 0.41632702267965, "learning_rate": 2.233636679512705e-06, "loss": 0.3197, "step": 7065 }, { "epoch": 2.280883005771016, "grad_norm": 0.33031330741010606, "learning_rate": 2.231736259598644e-06, "loss": 0.2312, "step": 7066 }, { "epoch": 2.2812058598006377, "grad_norm": 0.37281815671966645, "learning_rate": 2.2298365071663422e-06, "loss": 0.2497, "step": 7067 }, { "epoch": 2.2815287138302596, "grad_norm": 0.4256264339667865, "learning_rate": 2.227937422456499e-06, "loss": 0.2396, "step": 7068 }, { "epoch": 2.2818515678598814, "grad_norm": 0.4038325007262349, "learning_rate": 2.226039005709723e-06, "loss": 0.2414, "step": 7069 }, { "epoch": 2.2821744218895033, "grad_norm": 0.4134172003780355, "learning_rate": 2.2241412571665445e-06, "loss": 0.2074, "step": 7070 }, { "epoch": 2.282497275919125, "grad_norm": 0.389261096771528, "learning_rate": 2.2222441770674007e-06, "loss": 0.2462, "step": 7071 }, { "epoch": 2.282820129948747, "grad_norm": 0.36959017629640184, "learning_rate": 2.220347765652654e-06, "loss": 0.2187, "step": 7072 }, { "epoch": 2.283142983978369, "grad_norm": 0.3248079044189841, "learning_rate": 2.218452023162573e-06, "loss": 0.2422, "step": 7073 }, { "epoch": 2.2834658380079906, "grad_norm": 0.3877463089619499, "learning_rate": 2.2165569498373493e-06, "loss": 0.3058, "step": 7074 }, { "epoch": 2.2837886920376125, "grad_norm": 0.47020951218634305, "learning_rate": 2.2146625459170823e-06, "loss": 0.257, "step": 7075 }, { "epoch": 2.2841115460672343, "grad_norm": 0.4366498345834382, "learning_rate": 2.212768811641793e-06, "loss": 0.2556, "step": 7076 }, { "epoch": 2.284434400096856, "grad_norm": 0.38864604854858203, "learning_rate": 2.210875747251413e-06, "loss": 0.2497, "step": 7077 }, { "epoch": 2.284757254126478, "grad_norm": 0.4306379490289975, "learning_rate": 2.2089833529857953e-06, "loss": 0.2463, "step": 7078 }, { "epoch": 2.2850801081561, "grad_norm": 0.37404500166747107, "learning_rate": 2.2070916290846982e-06, "loss": 0.2329, "step": 7079 }, { "epoch": 2.2854029621857217, "grad_norm": 0.36176342809281836, "learning_rate": 2.2052005757878035e-06, "loss": 0.2259, "step": 7080 }, { "epoch": 2.2857258162153435, "grad_norm": 0.3490887335611444, "learning_rate": 2.2033101933347065e-06, "loss": 0.2485, "step": 7081 }, { "epoch": 2.2860486702449654, "grad_norm": 0.3627956577474309, "learning_rate": 2.201420481964911e-06, "loss": 0.2561, "step": 7082 }, { "epoch": 2.2863715242745872, "grad_norm": 0.37534032973346376, "learning_rate": 2.199531441917847e-06, "loss": 0.2503, "step": 7083 }, { "epoch": 2.286694378304209, "grad_norm": 0.33687401679411905, "learning_rate": 2.197643073432847e-06, "loss": 0.274, "step": 7084 }, { "epoch": 2.287017232333831, "grad_norm": 0.3955999617765481, "learning_rate": 2.1957553767491715e-06, "loss": 0.2625, "step": 7085 }, { "epoch": 2.2873400863634528, "grad_norm": 0.34920209132432745, "learning_rate": 2.1938683521059823e-06, "loss": 0.2304, "step": 7086 }, { "epoch": 2.2876629403930746, "grad_norm": 0.37694773482016497, "learning_rate": 2.191981999742367e-06, "loss": 0.2671, "step": 7087 }, { "epoch": 2.2879857944226965, "grad_norm": 0.40906651830585555, "learning_rate": 2.1900963198973224e-06, "loss": 0.2416, "step": 7088 }, { "epoch": 2.2883086484523183, "grad_norm": 0.35900599360302965, "learning_rate": 2.1882113128097635e-06, "loss": 0.2936, "step": 7089 }, { "epoch": 2.28863150248194, "grad_norm": 0.3778766032823438, "learning_rate": 2.1863269787185134e-06, "loss": 0.2328, "step": 7090 }, { "epoch": 2.288954356511562, "grad_norm": 0.3081305280060446, "learning_rate": 2.184443317862319e-06, "loss": 0.2629, "step": 7091 }, { "epoch": 2.289277210541184, "grad_norm": 0.3557202536349446, "learning_rate": 2.182560330479838e-06, "loss": 0.2402, "step": 7092 }, { "epoch": 2.289600064570806, "grad_norm": 0.38207057007739803, "learning_rate": 2.180678016809638e-06, "loss": 0.2543, "step": 7093 }, { "epoch": 2.289922918600428, "grad_norm": 0.37194368155215646, "learning_rate": 2.1787963770902083e-06, "loss": 0.279, "step": 7094 }, { "epoch": 2.29024577263005, "grad_norm": 0.33640551302137806, "learning_rate": 2.17691541155995e-06, "loss": 0.2105, "step": 7095 }, { "epoch": 2.2905686266596716, "grad_norm": 0.40375280824915993, "learning_rate": 2.1750351204571812e-06, "loss": 0.2688, "step": 7096 }, { "epoch": 2.2908914806892935, "grad_norm": 0.37483653046729126, "learning_rate": 2.173155504020127e-06, "loss": 0.2694, "step": 7097 }, { "epoch": 2.2912143347189153, "grad_norm": 0.4077617713101325, "learning_rate": 2.1712765624869374e-06, "loss": 0.3384, "step": 7098 }, { "epoch": 2.291537188748537, "grad_norm": 0.37163462667387176, "learning_rate": 2.169398296095667e-06, "loss": 0.2298, "step": 7099 }, { "epoch": 2.291860042778159, "grad_norm": 0.4242116248520392, "learning_rate": 2.1675207050842942e-06, "loss": 0.295, "step": 7100 }, { "epoch": 2.292182896807781, "grad_norm": 0.4507386459374528, "learning_rate": 2.165643789690702e-06, "loss": 0.3502, "step": 7101 }, { "epoch": 2.2925057508374027, "grad_norm": 0.3844061929538162, "learning_rate": 2.1637675501526995e-06, "loss": 0.3025, "step": 7102 }, { "epoch": 2.2928286048670246, "grad_norm": 0.38645804069531825, "learning_rate": 2.1618919867079962e-06, "loss": 0.2866, "step": 7103 }, { "epoch": 2.2931514588966464, "grad_norm": 0.40027119006132666, "learning_rate": 2.16001709959423e-06, "loss": 0.2822, "step": 7104 }, { "epoch": 2.2934743129262682, "grad_norm": 0.3443111486793933, "learning_rate": 2.158142889048941e-06, "loss": 0.2451, "step": 7105 }, { "epoch": 2.29379716695589, "grad_norm": 0.40996681861592044, "learning_rate": 2.156269355309595e-06, "loss": 0.2737, "step": 7106 }, { "epoch": 2.294120020985512, "grad_norm": 0.34519092322068257, "learning_rate": 2.1543964986135596e-06, "loss": 0.2573, "step": 7107 }, { "epoch": 2.2944428750151338, "grad_norm": 0.38876918383937015, "learning_rate": 2.1525243191981273e-06, "loss": 0.2982, "step": 7108 }, { "epoch": 2.2947657290447556, "grad_norm": 0.3665091128267546, "learning_rate": 2.1506528173004996e-06, "loss": 0.2409, "step": 7109 }, { "epoch": 2.2950885830743775, "grad_norm": 0.37882819375396015, "learning_rate": 2.1487819931577953e-06, "loss": 0.2618, "step": 7110 }, { "epoch": 2.2954114371039993, "grad_norm": 0.3354372204866365, "learning_rate": 2.1469118470070414e-06, "loss": 0.1958, "step": 7111 }, { "epoch": 2.295734291133621, "grad_norm": 0.3632492864385195, "learning_rate": 2.1450423790851873e-06, "loss": 0.2713, "step": 7112 }, { "epoch": 2.296057145163243, "grad_norm": 0.41455650247645914, "learning_rate": 2.1431735896290874e-06, "loss": 0.2975, "step": 7113 }, { "epoch": 2.296379999192865, "grad_norm": 0.47101001509622736, "learning_rate": 2.141305478875517e-06, "loss": 0.2399, "step": 7114 }, { "epoch": 2.2967028532224867, "grad_norm": 0.414091157572164, "learning_rate": 2.1394380470611645e-06, "loss": 0.2239, "step": 7115 }, { "epoch": 2.2970257072521085, "grad_norm": 0.4230294834002218, "learning_rate": 2.1375712944226274e-06, "loss": 0.285, "step": 7116 }, { "epoch": 2.2973485612817304, "grad_norm": 0.39907985954606057, "learning_rate": 2.135705221196425e-06, "loss": 0.2938, "step": 7117 }, { "epoch": 2.297671415311352, "grad_norm": 0.40140060962247465, "learning_rate": 2.133839827618983e-06, "loss": 0.2357, "step": 7118 }, { "epoch": 2.297994269340974, "grad_norm": 0.3954805303028845, "learning_rate": 2.1319751139266463e-06, "loss": 0.2804, "step": 7119 }, { "epoch": 2.298317123370596, "grad_norm": 0.40521906710480293, "learning_rate": 2.130111080355669e-06, "loss": 0.2543, "step": 7120 }, { "epoch": 2.2986399774002177, "grad_norm": 0.3774903067346121, "learning_rate": 2.1282477271422243e-06, "loss": 0.2661, "step": 7121 }, { "epoch": 2.2989628314298396, "grad_norm": 0.3405719239145565, "learning_rate": 2.126385054522394e-06, "loss": 0.2835, "step": 7122 }, { "epoch": 2.299285685459462, "grad_norm": 0.449610840483182, "learning_rate": 2.1245230627321767e-06, "loss": 0.2419, "step": 7123 }, { "epoch": 2.2996085394890837, "grad_norm": 0.45069839500885023, "learning_rate": 2.1226617520074867e-06, "loss": 0.2487, "step": 7124 }, { "epoch": 2.2999313935187056, "grad_norm": 0.41115406609909666, "learning_rate": 2.1208011225841462e-06, "loss": 0.238, "step": 7125 }, { "epoch": 2.3002542475483274, "grad_norm": 0.3785665313572255, "learning_rate": 2.1189411746978952e-06, "loss": 0.2515, "step": 7126 }, { "epoch": 2.3005771015779493, "grad_norm": 0.35017353712474847, "learning_rate": 2.11708190858439e-06, "loss": 0.2755, "step": 7127 }, { "epoch": 2.300899955607571, "grad_norm": 0.2945719865391619, "learning_rate": 2.115223324479192e-06, "loss": 0.232, "step": 7128 }, { "epoch": 2.301222809637193, "grad_norm": 0.4032239237970448, "learning_rate": 2.1133654226177865e-06, "loss": 0.2407, "step": 7129 }, { "epoch": 2.301545663666815, "grad_norm": 0.3606908960841101, "learning_rate": 2.1115082032355605e-06, "loss": 0.2766, "step": 7130 }, { "epoch": 2.3018685176964366, "grad_norm": 0.5531494469801989, "learning_rate": 2.1096516665678256e-06, "loss": 0.2342, "step": 7131 }, { "epoch": 2.3021913717260585, "grad_norm": 0.40942743530180997, "learning_rate": 2.107795812849803e-06, "loss": 0.2312, "step": 7132 }, { "epoch": 2.3025142257556803, "grad_norm": 0.3792589313849669, "learning_rate": 2.1059406423166246e-06, "loss": 0.2068, "step": 7133 }, { "epoch": 2.302837079785302, "grad_norm": 0.3664178475448566, "learning_rate": 2.1040861552033398e-06, "loss": 0.2766, "step": 7134 }, { "epoch": 2.303159933814924, "grad_norm": 0.33947739502638524, "learning_rate": 2.1022323517449067e-06, "loss": 0.2274, "step": 7135 }, { "epoch": 2.303482787844546, "grad_norm": 0.4108044414193401, "learning_rate": 2.100379232176204e-06, "loss": 0.2602, "step": 7136 }, { "epoch": 2.3038056418741677, "grad_norm": 0.33230618354129027, "learning_rate": 2.0985267967320136e-06, "loss": 0.2229, "step": 7137 }, { "epoch": 2.3041284959037895, "grad_norm": 0.3873741842834283, "learning_rate": 2.0966750456470427e-06, "loss": 0.2573, "step": 7138 }, { "epoch": 2.3044513499334114, "grad_norm": 0.3816271314411371, "learning_rate": 2.0948239791559003e-06, "loss": 0.2677, "step": 7139 }, { "epoch": 2.3047742039630332, "grad_norm": 0.46749717661600537, "learning_rate": 2.0929735974931167e-06, "loss": 0.2402, "step": 7140 }, { "epoch": 2.305097057992655, "grad_norm": 0.4369142082044152, "learning_rate": 2.0911239008931324e-06, "loss": 0.2163, "step": 7141 }, { "epoch": 2.305419912022277, "grad_norm": 0.31757865522651874, "learning_rate": 2.0892748895903027e-06, "loss": 0.2541, "step": 7142 }, { "epoch": 2.3057427660518988, "grad_norm": 0.38789049141176085, "learning_rate": 2.087426563818892e-06, "loss": 0.2206, "step": 7143 }, { "epoch": 2.3060656200815206, "grad_norm": 0.41614431718965056, "learning_rate": 2.085578923813084e-06, "loss": 0.2709, "step": 7144 }, { "epoch": 2.3063884741111425, "grad_norm": 0.45533594777968567, "learning_rate": 2.0837319698069686e-06, "loss": 0.2532, "step": 7145 }, { "epoch": 2.3067113281407643, "grad_norm": 0.3174735335964391, "learning_rate": 2.081885702034556e-06, "loss": 0.2511, "step": 7146 }, { "epoch": 2.307034182170386, "grad_norm": 0.3711103576930123, "learning_rate": 2.080040120729762e-06, "loss": 0.2619, "step": 7147 }, { "epoch": 2.307357036200008, "grad_norm": 0.3710676424381424, "learning_rate": 2.0781952261264227e-06, "loss": 0.2551, "step": 7148 }, { "epoch": 2.30767989022963, "grad_norm": 0.39257998066303906, "learning_rate": 2.0763510184582797e-06, "loss": 0.291, "step": 7149 }, { "epoch": 2.3080027442592517, "grad_norm": 0.37147801112801687, "learning_rate": 2.0745074979589935e-06, "loss": 0.2493, "step": 7150 }, { "epoch": 2.3083255982888735, "grad_norm": 0.33997456783053887, "learning_rate": 2.072664664862138e-06, "loss": 0.2399, "step": 7151 }, { "epoch": 2.3086484523184954, "grad_norm": 0.33914221500158126, "learning_rate": 2.0708225194011944e-06, "loss": 0.2293, "step": 7152 }, { "epoch": 2.308971306348117, "grad_norm": 0.6327838162840503, "learning_rate": 2.068981061809563e-06, "loss": 0.257, "step": 7153 }, { "epoch": 2.309294160377739, "grad_norm": 0.3708855913813989, "learning_rate": 2.0671402923205496e-06, "loss": 0.2645, "step": 7154 }, { "epoch": 2.309617014407361, "grad_norm": 0.3758973023896844, "learning_rate": 2.0653002111673796e-06, "loss": 0.2905, "step": 7155 }, { "epoch": 2.3099398684369827, "grad_norm": 0.3470361519471873, "learning_rate": 2.063460818583189e-06, "loss": 0.2513, "step": 7156 }, { "epoch": 2.3102627224666046, "grad_norm": 0.5160632982827273, "learning_rate": 2.061622114801028e-06, "loss": 0.23, "step": 7157 }, { "epoch": 2.3105855764962264, "grad_norm": 0.39515955048488544, "learning_rate": 2.0597841000538544e-06, "loss": 0.2332, "step": 7158 }, { "epoch": 2.3109084305258483, "grad_norm": 0.39712621562696704, "learning_rate": 2.057946774574546e-06, "loss": 0.2698, "step": 7159 }, { "epoch": 2.31123128455547, "grad_norm": 0.46854601910353905, "learning_rate": 2.0561101385958845e-06, "loss": 0.2372, "step": 7160 }, { "epoch": 2.3115541385850924, "grad_norm": 0.3709220990028771, "learning_rate": 2.0542741923505738e-06, "loss": 0.2725, "step": 7161 }, { "epoch": 2.3118769926147142, "grad_norm": 0.32726472872194495, "learning_rate": 2.052438936071222e-06, "loss": 0.2381, "step": 7162 }, { "epoch": 2.312199846644336, "grad_norm": 0.37319735586228314, "learning_rate": 2.050604369990357e-06, "loss": 0.267, "step": 7163 }, { "epoch": 2.312522700673958, "grad_norm": 0.3655784680472196, "learning_rate": 2.0487704943404125e-06, "loss": 0.2764, "step": 7164 }, { "epoch": 2.3128455547035798, "grad_norm": 0.36773229614626707, "learning_rate": 2.0469373093537413e-06, "loss": 0.3011, "step": 7165 }, { "epoch": 2.3131684087332016, "grad_norm": 0.36400947363693853, "learning_rate": 2.0451048152626013e-06, "loss": 0.2589, "step": 7166 }, { "epoch": 2.3134912627628235, "grad_norm": 0.34048183592502557, "learning_rate": 2.043273012299172e-06, "loss": 0.2646, "step": 7167 }, { "epoch": 2.3138141167924453, "grad_norm": 0.35248814245000903, "learning_rate": 2.041441900695535e-06, "loss": 0.2371, "step": 7168 }, { "epoch": 2.314136970822067, "grad_norm": 0.417963922056161, "learning_rate": 2.039611480683692e-06, "loss": 0.2268, "step": 7169 }, { "epoch": 2.314459824851689, "grad_norm": 0.38715567435709036, "learning_rate": 2.0377817524955584e-06, "loss": 0.2607, "step": 7170 }, { "epoch": 2.314782678881311, "grad_norm": 0.3324248571899364, "learning_rate": 2.0359527163629516e-06, "loss": 0.2272, "step": 7171 }, { "epoch": 2.3151055329109327, "grad_norm": 0.42582207514580006, "learning_rate": 2.0341243725176113e-06, "loss": 0.2931, "step": 7172 }, { "epoch": 2.3154283869405545, "grad_norm": 0.3751904488177221, "learning_rate": 2.032296721191187e-06, "loss": 0.2529, "step": 7173 }, { "epoch": 2.3157512409701764, "grad_norm": 0.37830082433631335, "learning_rate": 2.0304697626152403e-06, "loss": 0.2568, "step": 7174 }, { "epoch": 2.316074094999798, "grad_norm": 0.4281745485837635, "learning_rate": 2.0286434970212403e-06, "loss": 0.238, "step": 7175 }, { "epoch": 2.31639694902942, "grad_norm": 0.3922818016818767, "learning_rate": 2.0268179246405763e-06, "loss": 0.2936, "step": 7176 }, { "epoch": 2.316719803059042, "grad_norm": 0.5008442383358712, "learning_rate": 2.024993045704543e-06, "loss": 0.2366, "step": 7177 }, { "epoch": 2.3170426570886637, "grad_norm": 0.43839883433532373, "learning_rate": 2.0231688604443536e-06, "loss": 0.2214, "step": 7178 }, { "epoch": 2.3173655111182856, "grad_norm": 0.3848092027161872, "learning_rate": 2.021345369091125e-06, "loss": 0.2737, "step": 7179 }, { "epoch": 2.3176883651479074, "grad_norm": 0.7321020465054038, "learning_rate": 2.019522571875895e-06, "loss": 0.2347, "step": 7180 }, { "epoch": 2.3180112191775293, "grad_norm": 0.42929299962308326, "learning_rate": 2.017700469029607e-06, "loss": 0.2449, "step": 7181 }, { "epoch": 2.318334073207151, "grad_norm": 0.3469071642601702, "learning_rate": 2.0158790607831217e-06, "loss": 0.2814, "step": 7182 }, { "epoch": 2.318656927236773, "grad_norm": 0.34320704326269175, "learning_rate": 2.014058347367205e-06, "loss": 0.2134, "step": 7183 }, { "epoch": 2.318979781266395, "grad_norm": 0.38991613428687477, "learning_rate": 2.0122383290125433e-06, "loss": 0.2606, "step": 7184 }, { "epoch": 2.3193026352960167, "grad_norm": 0.3439138293017563, "learning_rate": 2.010419005949727e-06, "loss": 0.2559, "step": 7185 }, { "epoch": 2.3196254893256385, "grad_norm": 0.38833512598324, "learning_rate": 2.0086003784092618e-06, "loss": 0.2275, "step": 7186 }, { "epoch": 2.3199483433552603, "grad_norm": 0.34080701822676485, "learning_rate": 2.006782446621567e-06, "loss": 0.2557, "step": 7187 }, { "epoch": 2.320271197384882, "grad_norm": 0.34724829198771245, "learning_rate": 2.0049652108169714e-06, "loss": 0.3063, "step": 7188 }, { "epoch": 2.320594051414504, "grad_norm": 1.558450221079259, "learning_rate": 2.003148671225719e-06, "loss": 0.2615, "step": 7189 }, { "epoch": 2.3209169054441263, "grad_norm": 0.4321836594944734, "learning_rate": 2.0013328280779582e-06, "loss": 0.3053, "step": 7190 }, { "epoch": 2.321239759473748, "grad_norm": 0.4396190717292236, "learning_rate": 1.9995176816037576e-06, "loss": 0.2955, "step": 7191 }, { "epoch": 2.32156261350337, "grad_norm": 0.37683274300932385, "learning_rate": 1.997703232033091e-06, "loss": 0.2036, "step": 7192 }, { "epoch": 2.321885467532992, "grad_norm": 0.6488082245857898, "learning_rate": 1.995889479595849e-06, "loss": 0.2649, "step": 7193 }, { "epoch": 2.3222083215626137, "grad_norm": 0.362380903033795, "learning_rate": 1.994076424521828e-06, "loss": 0.2743, "step": 7194 }, { "epoch": 2.3225311755922355, "grad_norm": 0.35249910661745926, "learning_rate": 1.9922640670407446e-06, "loss": 0.2173, "step": 7195 }, { "epoch": 2.3228540296218574, "grad_norm": 0.4293232912060608, "learning_rate": 1.9904524073822174e-06, "loss": 0.2193, "step": 7196 }, { "epoch": 2.3231768836514792, "grad_norm": 0.5220318501512736, "learning_rate": 1.9886414457757846e-06, "loss": 0.3073, "step": 7197 }, { "epoch": 2.323499737681101, "grad_norm": 0.376917969699881, "learning_rate": 1.9868311824508896e-06, "loss": 0.2915, "step": 7198 }, { "epoch": 2.323822591710723, "grad_norm": 0.4156234731611169, "learning_rate": 1.9850216176368926e-06, "loss": 0.349, "step": 7199 }, { "epoch": 2.3241454457403448, "grad_norm": 0.3949078264959798, "learning_rate": 1.9832127515630606e-06, "loss": 0.3057, "step": 7200 }, { "epoch": 2.3244682997699666, "grad_norm": 0.4371069428973889, "learning_rate": 1.9814045844585763e-06, "loss": 0.2856, "step": 7201 }, { "epoch": 2.3247911537995885, "grad_norm": 0.38328124496729166, "learning_rate": 1.9795971165525333e-06, "loss": 0.2397, "step": 7202 }, { "epoch": 2.3251140078292103, "grad_norm": 0.41106421268822324, "learning_rate": 1.9777903480739317e-06, "loss": 0.2557, "step": 7203 }, { "epoch": 2.325436861858832, "grad_norm": 0.38350363553140726, "learning_rate": 1.9759842792516883e-06, "loss": 0.2279, "step": 7204 }, { "epoch": 2.325759715888454, "grad_norm": 0.3580081887311704, "learning_rate": 1.97417891031463e-06, "loss": 0.248, "step": 7205 }, { "epoch": 2.326082569918076, "grad_norm": 0.40188968636395833, "learning_rate": 1.972374241491496e-06, "loss": 0.2531, "step": 7206 }, { "epoch": 2.3264054239476977, "grad_norm": 0.41490767560096853, "learning_rate": 1.9705702730109317e-06, "loss": 0.2761, "step": 7207 }, { "epoch": 2.3267282779773195, "grad_norm": 0.3736963540944071, "learning_rate": 1.968767005101502e-06, "loss": 0.2339, "step": 7208 }, { "epoch": 2.3270511320069414, "grad_norm": 0.4388827779237527, "learning_rate": 1.9669644379916728e-06, "loss": 0.3022, "step": 7209 }, { "epoch": 2.327373986036563, "grad_norm": 0.35261540293087135, "learning_rate": 1.9651625719098317e-06, "loss": 0.226, "step": 7210 }, { "epoch": 2.327696840066185, "grad_norm": 0.3676799339977543, "learning_rate": 1.9633614070842696e-06, "loss": 0.3096, "step": 7211 }, { "epoch": 2.328019694095807, "grad_norm": 0.4384972264099562, "learning_rate": 1.961560943743194e-06, "loss": 0.2835, "step": 7212 }, { "epoch": 2.3283425481254287, "grad_norm": 0.3983847776784342, "learning_rate": 1.959761182114719e-06, "loss": 0.2459, "step": 7213 }, { "epoch": 2.3286654021550506, "grad_norm": 0.422357571124962, "learning_rate": 1.9579621224268747e-06, "loss": 0.2848, "step": 7214 }, { "epoch": 2.3289882561846724, "grad_norm": 0.44127306769965075, "learning_rate": 1.9561637649075952e-06, "loss": 0.2536, "step": 7215 }, { "epoch": 2.3293111102142943, "grad_norm": 0.3944558931153035, "learning_rate": 1.954366109784734e-06, "loss": 0.2657, "step": 7216 }, { "epoch": 2.329633964243916, "grad_norm": 0.4219690335494334, "learning_rate": 1.9525691572860485e-06, "loss": 0.2194, "step": 7217 }, { "epoch": 2.329956818273538, "grad_norm": 0.3387729285573652, "learning_rate": 1.9507729076392114e-06, "loss": 0.2289, "step": 7218 }, { "epoch": 2.33027967230316, "grad_norm": 0.4227725260003739, "learning_rate": 1.948977361071805e-06, "loss": 0.1959, "step": 7219 }, { "epoch": 2.3306025263327816, "grad_norm": 0.37342646510342087, "learning_rate": 1.947182517811325e-06, "loss": 0.2317, "step": 7220 }, { "epoch": 2.3309253803624035, "grad_norm": 0.3827318235987086, "learning_rate": 1.9453883780851713e-06, "loss": 0.296, "step": 7221 }, { "epoch": 2.3312482343920253, "grad_norm": 0.3358611483848386, "learning_rate": 1.943594942120662e-06, "loss": 0.3161, "step": 7222 }, { "epoch": 2.331571088421647, "grad_norm": 0.4164084012024952, "learning_rate": 1.941802210145021e-06, "loss": 0.2959, "step": 7223 }, { "epoch": 2.331893942451269, "grad_norm": 0.4280732012120147, "learning_rate": 1.940010182385385e-06, "loss": 0.2896, "step": 7224 }, { "epoch": 2.332216796480891, "grad_norm": 0.3556112932431118, "learning_rate": 1.938218859068805e-06, "loss": 0.2413, "step": 7225 }, { "epoch": 2.3325396505105127, "grad_norm": 0.4105276026858935, "learning_rate": 1.936428240422235e-06, "loss": 0.2746, "step": 7226 }, { "epoch": 2.3328625045401346, "grad_norm": 0.42489305840063557, "learning_rate": 1.9346383266725467e-06, "loss": 0.3017, "step": 7227 }, { "epoch": 2.333185358569757, "grad_norm": 0.3812828871953256, "learning_rate": 1.9328491180465165e-06, "loss": 0.3154, "step": 7228 }, { "epoch": 2.3335082125993787, "grad_norm": 0.44042838853880323, "learning_rate": 1.931060614770839e-06, "loss": 0.2522, "step": 7229 }, { "epoch": 2.3338310666290005, "grad_norm": 0.4047834272795422, "learning_rate": 1.92927281707211e-06, "loss": 0.2422, "step": 7230 }, { "epoch": 2.3341539206586224, "grad_norm": 0.3640176020037845, "learning_rate": 1.9274857251768465e-06, "loss": 0.2463, "step": 7231 }, { "epoch": 2.334476774688244, "grad_norm": 0.3730732094303598, "learning_rate": 1.9256993393114653e-06, "loss": 0.2097, "step": 7232 }, { "epoch": 2.334799628717866, "grad_norm": 0.31798096347709714, "learning_rate": 1.9239136597023017e-06, "loss": 0.1869, "step": 7233 }, { "epoch": 2.335122482747488, "grad_norm": 0.4188458396658152, "learning_rate": 1.922128686575599e-06, "loss": 0.2506, "step": 7234 }, { "epoch": 2.3354453367771097, "grad_norm": 0.45776366320216455, "learning_rate": 1.920344420157512e-06, "loss": 0.212, "step": 7235 }, { "epoch": 2.3357681908067316, "grad_norm": 0.3587259025754898, "learning_rate": 1.9185608606741015e-06, "loss": 0.2769, "step": 7236 }, { "epoch": 2.3360910448363534, "grad_norm": 0.4131682801018759, "learning_rate": 1.916778008351346e-06, "loss": 0.2588, "step": 7237 }, { "epoch": 2.3364138988659753, "grad_norm": 0.33489190506748867, "learning_rate": 1.914995863415126e-06, "loss": 0.2709, "step": 7238 }, { "epoch": 2.336736752895597, "grad_norm": 0.39204317756276036, "learning_rate": 1.913214426091242e-06, "loss": 0.292, "step": 7239 }, { "epoch": 2.337059606925219, "grad_norm": 0.3347414097113986, "learning_rate": 1.911433696605394e-06, "loss": 0.2769, "step": 7240 }, { "epoch": 2.337382460954841, "grad_norm": 0.4051276632462698, "learning_rate": 1.9096536751832025e-06, "loss": 0.2579, "step": 7241 }, { "epoch": 2.3377053149844627, "grad_norm": 0.32132770491382684, "learning_rate": 1.90787436205019e-06, "loss": 0.2564, "step": 7242 }, { "epoch": 2.3380281690140845, "grad_norm": 0.41222494104449603, "learning_rate": 1.9060957574317956e-06, "loss": 0.2424, "step": 7243 }, { "epoch": 2.3383510230437063, "grad_norm": 0.4266807087097654, "learning_rate": 1.9043178615533669e-06, "loss": 0.2399, "step": 7244 }, { "epoch": 2.338673877073328, "grad_norm": 0.3785035298612985, "learning_rate": 1.902540674640158e-06, "loss": 0.2138, "step": 7245 }, { "epoch": 2.33899673110295, "grad_norm": 0.33288602254047983, "learning_rate": 1.900764196917339e-06, "loss": 0.2744, "step": 7246 }, { "epoch": 2.339319585132572, "grad_norm": 0.3653206930528455, "learning_rate": 1.8989884286099837e-06, "loss": 0.2106, "step": 7247 }, { "epoch": 2.3396424391621937, "grad_norm": 0.368906993024496, "learning_rate": 1.897213369943084e-06, "loss": 0.2512, "step": 7248 }, { "epoch": 2.3399652931918156, "grad_norm": 0.3505397429895147, "learning_rate": 1.8954390211415319e-06, "loss": 0.2753, "step": 7249 }, { "epoch": 2.3402881472214374, "grad_norm": 0.37609257968910514, "learning_rate": 1.8936653824301385e-06, "loss": 0.2312, "step": 7250 }, { "epoch": 2.3406110012510593, "grad_norm": 0.3953433008232391, "learning_rate": 1.8918924540336208e-06, "loss": 0.2608, "step": 7251 }, { "epoch": 2.340933855280681, "grad_norm": 0.4213379297944791, "learning_rate": 1.8901202361766077e-06, "loss": 0.296, "step": 7252 }, { "epoch": 2.341256709310303, "grad_norm": 0.3360835270812653, "learning_rate": 1.8883487290836337e-06, "loss": 0.2403, "step": 7253 }, { "epoch": 2.341579563339925, "grad_norm": 0.38881586704325044, "learning_rate": 1.8865779329791505e-06, "loss": 0.2472, "step": 7254 }, { "epoch": 2.3419024173695466, "grad_norm": 0.3948686987165178, "learning_rate": 1.8848078480875108e-06, "loss": 0.2106, "step": 7255 }, { "epoch": 2.3422252713991685, "grad_norm": 0.3526305911128923, "learning_rate": 1.8830384746329863e-06, "loss": 0.242, "step": 7256 }, { "epoch": 2.3425481254287903, "grad_norm": 0.46140610925712094, "learning_rate": 1.8812698128397498e-06, "loss": 0.2466, "step": 7257 }, { "epoch": 2.3428709794584126, "grad_norm": 0.3053800573626244, "learning_rate": 1.8795018629318925e-06, "loss": 0.2712, "step": 7258 }, { "epoch": 2.3431938334880344, "grad_norm": 0.35693079023944885, "learning_rate": 1.8777346251334074e-06, "loss": 0.2673, "step": 7259 }, { "epoch": 2.3435166875176563, "grad_norm": 0.36080373925100484, "learning_rate": 1.875968099668203e-06, "loss": 0.2421, "step": 7260 }, { "epoch": 2.343839541547278, "grad_norm": 0.3131351875924989, "learning_rate": 1.8742022867600977e-06, "loss": 0.2648, "step": 7261 }, { "epoch": 2.3441623955769, "grad_norm": 0.40736187972706034, "learning_rate": 1.8724371866328135e-06, "loss": 0.3001, "step": 7262 }, { "epoch": 2.344485249606522, "grad_norm": 0.36944835364296863, "learning_rate": 1.87067279950999e-06, "loss": 0.2815, "step": 7263 }, { "epoch": 2.3448081036361437, "grad_norm": 0.3718270451720857, "learning_rate": 1.8689091256151687e-06, "loss": 0.2273, "step": 7264 }, { "epoch": 2.3451309576657655, "grad_norm": 0.3528072242244851, "learning_rate": 1.8671461651718068e-06, "loss": 0.2504, "step": 7265 }, { "epoch": 2.3454538116953874, "grad_norm": 0.3687369511154515, "learning_rate": 1.8653839184032686e-06, "loss": 0.2313, "step": 7266 }, { "epoch": 2.345776665725009, "grad_norm": 0.37288016122099693, "learning_rate": 1.863622385532831e-06, "loss": 0.2465, "step": 7267 }, { "epoch": 2.346099519754631, "grad_norm": 0.469487863963992, "learning_rate": 1.861861566783673e-06, "loss": 0.2482, "step": 7268 }, { "epoch": 2.346422373784253, "grad_norm": 0.37808490648567145, "learning_rate": 1.8601014623788929e-06, "loss": 0.2104, "step": 7269 }, { "epoch": 2.3467452278138747, "grad_norm": 0.4326662976525239, "learning_rate": 1.8583420725414887e-06, "loss": 0.2715, "step": 7270 }, { "epoch": 2.3470680818434966, "grad_norm": 0.48825121753704875, "learning_rate": 1.8565833974943767e-06, "loss": 0.2571, "step": 7271 }, { "epoch": 2.3473909358731184, "grad_norm": 0.44100860768256306, "learning_rate": 1.854825437460375e-06, "loss": 0.2499, "step": 7272 }, { "epoch": 2.3477137899027403, "grad_norm": 0.5551301886207132, "learning_rate": 1.8530681926622179e-06, "loss": 0.2624, "step": 7273 }, { "epoch": 2.348036643932362, "grad_norm": 0.3853742220516908, "learning_rate": 1.8513116633225432e-06, "loss": 0.2748, "step": 7274 }, { "epoch": 2.348359497961984, "grad_norm": 0.4418679730275688, "learning_rate": 1.8495558496639039e-06, "loss": 0.2345, "step": 7275 }, { "epoch": 2.348682351991606, "grad_norm": 0.3716132521202852, "learning_rate": 1.8478007519087554e-06, "loss": 0.3004, "step": 7276 }, { "epoch": 2.3490052060212276, "grad_norm": 0.34750725953212624, "learning_rate": 1.8460463702794695e-06, "loss": 0.2293, "step": 7277 }, { "epoch": 2.3493280600508495, "grad_norm": 0.3248393074067032, "learning_rate": 1.844292704998321e-06, "loss": 0.2411, "step": 7278 }, { "epoch": 2.3496509140804713, "grad_norm": 0.4043564115944545, "learning_rate": 1.8425397562874982e-06, "loss": 0.2575, "step": 7279 }, { "epoch": 2.349973768110093, "grad_norm": 0.4132567517277977, "learning_rate": 1.8407875243690974e-06, "loss": 0.284, "step": 7280 }, { "epoch": 2.350296622139715, "grad_norm": 0.5499112032967415, "learning_rate": 1.8390360094651271e-06, "loss": 0.2628, "step": 7281 }, { "epoch": 2.350619476169337, "grad_norm": 0.39448366390135353, "learning_rate": 1.8372852117974953e-06, "loss": 0.3077, "step": 7282 }, { "epoch": 2.3509423301989587, "grad_norm": 0.38224058955851353, "learning_rate": 1.8355351315880299e-06, "loss": 0.2983, "step": 7283 }, { "epoch": 2.3512651842285806, "grad_norm": 0.3724571659505902, "learning_rate": 1.8337857690584648e-06, "loss": 0.272, "step": 7284 }, { "epoch": 2.3515880382582024, "grad_norm": 0.4096071006555764, "learning_rate": 1.832037124430438e-06, "loss": 0.2434, "step": 7285 }, { "epoch": 2.3519108922878242, "grad_norm": 0.4223927554172502, "learning_rate": 1.8302891979255046e-06, "loss": 0.3395, "step": 7286 }, { "epoch": 2.352233746317446, "grad_norm": 0.3907587207754847, "learning_rate": 1.8285419897651193e-06, "loss": 0.2949, "step": 7287 }, { "epoch": 2.352556600347068, "grad_norm": 0.3895673359150032, "learning_rate": 1.8267955001706566e-06, "loss": 0.2104, "step": 7288 }, { "epoch": 2.3528794543766898, "grad_norm": 0.37526052774642027, "learning_rate": 1.8250497293633894e-06, "loss": 0.331, "step": 7289 }, { "epoch": 2.3532023084063116, "grad_norm": 0.41529701828044957, "learning_rate": 1.8233046775645084e-06, "loss": 0.2449, "step": 7290 }, { "epoch": 2.3535251624359335, "grad_norm": 0.5132537619979701, "learning_rate": 1.821560344995106e-06, "loss": 0.2858, "step": 7291 }, { "epoch": 2.3538480164655553, "grad_norm": 0.4913897301129678, "learning_rate": 1.8198167318761893e-06, "loss": 0.269, "step": 7292 }, { "epoch": 2.354170870495177, "grad_norm": 0.33627061820740306, "learning_rate": 1.8180738384286694e-06, "loss": 0.2532, "step": 7293 }, { "epoch": 2.354493724524799, "grad_norm": 0.3640125182725707, "learning_rate": 1.816331664873371e-06, "loss": 0.2522, "step": 7294 }, { "epoch": 2.354816578554421, "grad_norm": 0.38732288699433287, "learning_rate": 1.8145902114310228e-06, "loss": 0.2701, "step": 7295 }, { "epoch": 2.355139432584043, "grad_norm": 0.3670412779913131, "learning_rate": 1.8128494783222649e-06, "loss": 0.2292, "step": 7296 }, { "epoch": 2.355462286613665, "grad_norm": 0.37454852333823774, "learning_rate": 1.8111094657676459e-06, "loss": 0.2177, "step": 7297 }, { "epoch": 2.355785140643287, "grad_norm": 0.4039951929463375, "learning_rate": 1.8093701739876242e-06, "loss": 0.278, "step": 7298 }, { "epoch": 2.3561079946729087, "grad_norm": 0.3731127180622434, "learning_rate": 1.8076316032025671e-06, "loss": 0.2691, "step": 7299 }, { "epoch": 2.3564308487025305, "grad_norm": 0.38130952907285864, "learning_rate": 1.8058937536327445e-06, "loss": 0.2812, "step": 7300 }, { "epoch": 2.3567537027321523, "grad_norm": 0.44733655968823755, "learning_rate": 1.8041566254983445e-06, "loss": 0.2595, "step": 7301 }, { "epoch": 2.357076556761774, "grad_norm": 0.4230704799920278, "learning_rate": 1.8024202190194537e-06, "loss": 0.2683, "step": 7302 }, { "epoch": 2.357399410791396, "grad_norm": 0.36764691064905985, "learning_rate": 1.8006845344160778e-06, "loss": 0.2931, "step": 7303 }, { "epoch": 2.357722264821018, "grad_norm": 0.3520687984925023, "learning_rate": 1.7989495719081212e-06, "loss": 0.225, "step": 7304 }, { "epoch": 2.3580451188506397, "grad_norm": 0.4058664446040839, "learning_rate": 1.7972153317154046e-06, "loss": 0.28, "step": 7305 }, { "epoch": 2.3583679728802616, "grad_norm": 0.45057738883619697, "learning_rate": 1.79548181405765e-06, "loss": 0.2739, "step": 7306 }, { "epoch": 2.3586908269098834, "grad_norm": 0.3552966653045276, "learning_rate": 1.7937490191544966e-06, "loss": 0.3071, "step": 7307 }, { "epoch": 2.3590136809395053, "grad_norm": 0.3974337524927838, "learning_rate": 1.7920169472254822e-06, "loss": 0.2681, "step": 7308 }, { "epoch": 2.359336534969127, "grad_norm": 0.3637763718678521, "learning_rate": 1.7902855984900618e-06, "loss": 0.29, "step": 7309 }, { "epoch": 2.359659388998749, "grad_norm": 0.3929750390919796, "learning_rate": 1.7885549731675918e-06, "loss": 0.2741, "step": 7310 }, { "epoch": 2.359982243028371, "grad_norm": 0.3549733199431505, "learning_rate": 1.7868250714773418e-06, "loss": 0.2761, "step": 7311 }, { "epoch": 2.3603050970579926, "grad_norm": 0.4678362599079407, "learning_rate": 1.7850958936384874e-06, "loss": 0.2927, "step": 7312 }, { "epoch": 2.3606279510876145, "grad_norm": 0.35381875274903457, "learning_rate": 1.7833674398701155e-06, "loss": 0.2306, "step": 7313 }, { "epoch": 2.3609508051172363, "grad_norm": 0.43100222267903565, "learning_rate": 1.7816397103912147e-06, "loss": 0.2519, "step": 7314 }, { "epoch": 2.361273659146858, "grad_norm": 0.37388035950937065, "learning_rate": 1.7799127054206896e-06, "loss": 0.2767, "step": 7315 }, { "epoch": 2.36159651317648, "grad_norm": 0.38339749446432947, "learning_rate": 1.778186425177346e-06, "loss": 0.2871, "step": 7316 }, { "epoch": 2.361919367206102, "grad_norm": 0.41304344137991866, "learning_rate": 1.7764608698799026e-06, "loss": 0.2796, "step": 7317 }, { "epoch": 2.3622422212357237, "grad_norm": 0.35129476211601807, "learning_rate": 1.7747360397469875e-06, "loss": 0.2767, "step": 7318 }, { "epoch": 2.3625650752653455, "grad_norm": 0.35861598557674784, "learning_rate": 1.7730119349971302e-06, "loss": 0.2342, "step": 7319 }, { "epoch": 2.3628879292949674, "grad_norm": 0.35701582323248315, "learning_rate": 1.771288555848776e-06, "loss": 0.3138, "step": 7320 }, { "epoch": 2.3632107833245892, "grad_norm": 0.4203653096510524, "learning_rate": 1.7695659025202709e-06, "loss": 0.3057, "step": 7321 }, { "epoch": 2.363533637354211, "grad_norm": 0.37652926776611156, "learning_rate": 1.7678439752298765e-06, "loss": 0.2848, "step": 7322 }, { "epoch": 2.363856491383833, "grad_norm": 0.38163677618359293, "learning_rate": 1.7661227741957545e-06, "loss": 0.259, "step": 7323 }, { "epoch": 2.3641793454134548, "grad_norm": 0.4261204909256844, "learning_rate": 1.764402299635984e-06, "loss": 0.2066, "step": 7324 }, { "epoch": 2.364502199443077, "grad_norm": 0.31130139814491214, "learning_rate": 1.7626825517685407e-06, "loss": 0.2408, "step": 7325 }, { "epoch": 2.364825053472699, "grad_norm": 0.36042403661106764, "learning_rate": 1.7609635308113204e-06, "loss": 0.2326, "step": 7326 }, { "epoch": 2.3651479075023207, "grad_norm": 0.4157014989004567, "learning_rate": 1.7592452369821144e-06, "loss": 0.2433, "step": 7327 }, { "epoch": 2.3654707615319426, "grad_norm": 0.3894074387524195, "learning_rate": 1.757527670498632e-06, "loss": 0.2748, "step": 7328 }, { "epoch": 2.3657936155615644, "grad_norm": 0.7434505599550009, "learning_rate": 1.755810831578486e-06, "loss": 0.2475, "step": 7329 }, { "epoch": 2.3661164695911863, "grad_norm": 0.37866783652838065, "learning_rate": 1.7540947204391997e-06, "loss": 0.2384, "step": 7330 }, { "epoch": 2.366439323620808, "grad_norm": 0.3271273545411594, "learning_rate": 1.7523793372981967e-06, "loss": 0.2513, "step": 7331 }, { "epoch": 2.36676217765043, "grad_norm": 0.3809641438885556, "learning_rate": 1.7506646823728199e-06, "loss": 0.2414, "step": 7332 }, { "epoch": 2.367085031680052, "grad_norm": 0.37484591805574086, "learning_rate": 1.7489507558803079e-06, "loss": 0.2713, "step": 7333 }, { "epoch": 2.3674078857096736, "grad_norm": 0.4488675642795669, "learning_rate": 1.7472375580378152e-06, "loss": 0.2239, "step": 7334 }, { "epoch": 2.3677307397392955, "grad_norm": 0.39978148126052204, "learning_rate": 1.7455250890624044e-06, "loss": 0.227, "step": 7335 }, { "epoch": 2.3680535937689173, "grad_norm": 0.42211279043859246, "learning_rate": 1.7438133491710384e-06, "loss": 0.2471, "step": 7336 }, { "epoch": 2.368376447798539, "grad_norm": 0.44829544189741144, "learning_rate": 1.7421023385805962e-06, "loss": 0.284, "step": 7337 }, { "epoch": 2.368699301828161, "grad_norm": 0.4314273339664403, "learning_rate": 1.7403920575078562e-06, "loss": 0.248, "step": 7338 }, { "epoch": 2.369022155857783, "grad_norm": 0.3636630011771146, "learning_rate": 1.7386825061695138e-06, "loss": 0.2532, "step": 7339 }, { "epoch": 2.3693450098874047, "grad_norm": 0.4487352522377407, "learning_rate": 1.7369736847821613e-06, "loss": 0.233, "step": 7340 }, { "epoch": 2.3696678639170266, "grad_norm": 0.4237744706681811, "learning_rate": 1.7352655935623093e-06, "loss": 0.2983, "step": 7341 }, { "epoch": 2.3699907179466484, "grad_norm": 0.35340272686281604, "learning_rate": 1.7335582327263658e-06, "loss": 0.2307, "step": 7342 }, { "epoch": 2.3703135719762702, "grad_norm": 0.3386458148622444, "learning_rate": 1.7318516024906532e-06, "loss": 0.276, "step": 7343 }, { "epoch": 2.370636426005892, "grad_norm": 0.38929489483339735, "learning_rate": 1.7301457030713998e-06, "loss": 0.1922, "step": 7344 }, { "epoch": 2.370959280035514, "grad_norm": 0.3968983003008476, "learning_rate": 1.7284405346847426e-06, "loss": 0.2586, "step": 7345 }, { "epoch": 2.3712821340651358, "grad_norm": 0.3631473198067394, "learning_rate": 1.7267360975467194e-06, "loss": 0.2757, "step": 7346 }, { "epoch": 2.3716049880947576, "grad_norm": 0.3988839651783654, "learning_rate": 1.7250323918732843e-06, "loss": 0.2659, "step": 7347 }, { "epoch": 2.3719278421243795, "grad_norm": 0.3432861426954921, "learning_rate": 1.7233294178802906e-06, "loss": 0.215, "step": 7348 }, { "epoch": 2.3722506961540013, "grad_norm": 0.37596652348660875, "learning_rate": 1.721627175783506e-06, "loss": 0.3045, "step": 7349 }, { "epoch": 2.372573550183623, "grad_norm": 0.36000993389812097, "learning_rate": 1.7199256657985995e-06, "loss": 0.2637, "step": 7350 }, { "epoch": 2.372896404213245, "grad_norm": 0.49649372881256554, "learning_rate": 1.7182248881411542e-06, "loss": 0.2791, "step": 7351 }, { "epoch": 2.373219258242867, "grad_norm": 0.395417939215377, "learning_rate": 1.7165248430266508e-06, "loss": 0.3046, "step": 7352 }, { "epoch": 2.3735421122724887, "grad_norm": 0.40629176207822765, "learning_rate": 1.7148255306704858e-06, "loss": 0.2478, "step": 7353 }, { "epoch": 2.3738649663021105, "grad_norm": 0.32545576096476425, "learning_rate": 1.7131269512879607e-06, "loss": 0.2052, "step": 7354 }, { "epoch": 2.3741878203317324, "grad_norm": 0.40384976040008785, "learning_rate": 1.7114291050942803e-06, "loss": 0.295, "step": 7355 }, { "epoch": 2.374510674361354, "grad_norm": 0.36214931291460184, "learning_rate": 1.7097319923045635e-06, "loss": 0.2251, "step": 7356 }, { "epoch": 2.374833528390976, "grad_norm": 0.4322143953190718, "learning_rate": 1.708035613133827e-06, "loss": 0.2415, "step": 7357 }, { "epoch": 2.375156382420598, "grad_norm": 0.37342663025837164, "learning_rate": 1.7063399677970026e-06, "loss": 0.2157, "step": 7358 }, { "epoch": 2.3754792364502197, "grad_norm": 0.44402206766621305, "learning_rate": 1.7046450565089276e-06, "loss": 0.2803, "step": 7359 }, { "epoch": 2.3758020904798416, "grad_norm": 0.4028863736911547, "learning_rate": 1.7029508794843415e-06, "loss": 0.2537, "step": 7360 }, { "epoch": 2.3761249445094634, "grad_norm": 0.3847345971383495, "learning_rate": 1.7012574369378955e-06, "loss": 0.2958, "step": 7361 }, { "epoch": 2.3764477985390853, "grad_norm": 0.40909330113041364, "learning_rate": 1.6995647290841496e-06, "loss": 0.2772, "step": 7362 }, { "epoch": 2.3767706525687076, "grad_norm": 0.3903957191258611, "learning_rate": 1.6978727561375621e-06, "loss": 0.2319, "step": 7363 }, { "epoch": 2.3770935065983294, "grad_norm": 0.38042877115056345, "learning_rate": 1.696181518312509e-06, "loss": 0.2487, "step": 7364 }, { "epoch": 2.3774163606279513, "grad_norm": 0.3560424306952159, "learning_rate": 1.6944910158232632e-06, "loss": 0.2368, "step": 7365 }, { "epoch": 2.377739214657573, "grad_norm": 0.41535778575658744, "learning_rate": 1.6928012488840125e-06, "loss": 0.2402, "step": 7366 }, { "epoch": 2.378062068687195, "grad_norm": 0.38077130100318635, "learning_rate": 1.6911122177088456e-06, "loss": 0.2687, "step": 7367 }, { "epoch": 2.378384922716817, "grad_norm": 0.3816043650252351, "learning_rate": 1.689423922511762e-06, "loss": 0.2359, "step": 7368 }, { "epoch": 2.3787077767464386, "grad_norm": 0.3741919146961018, "learning_rate": 1.6877363635066653e-06, "loss": 0.2949, "step": 7369 }, { "epoch": 2.3790306307760605, "grad_norm": 0.39240977516627223, "learning_rate": 1.6860495409073681e-06, "loss": 0.2807, "step": 7370 }, { "epoch": 2.3793534848056823, "grad_norm": 0.3901454890089896, "learning_rate": 1.6843634549275863e-06, "loss": 0.21, "step": 7371 }, { "epoch": 2.379676338835304, "grad_norm": 0.3678159493889844, "learning_rate": 1.6826781057809465e-06, "loss": 0.3152, "step": 7372 }, { "epoch": 2.379999192864926, "grad_norm": 0.40592268945017895, "learning_rate": 1.6809934936809811e-06, "loss": 0.2418, "step": 7373 }, { "epoch": 2.380322046894548, "grad_norm": 0.4319677183968547, "learning_rate": 1.6793096188411256e-06, "loss": 0.2588, "step": 7374 }, { "epoch": 2.3806449009241697, "grad_norm": 0.4007150290075606, "learning_rate": 1.6776264814747248e-06, "loss": 0.2994, "step": 7375 }, { "epoch": 2.3809677549537915, "grad_norm": 0.41595715234004565, "learning_rate": 1.6759440817950313e-06, "loss": 0.2294, "step": 7376 }, { "epoch": 2.3812906089834134, "grad_norm": 0.35666347150567024, "learning_rate": 1.6742624200152051e-06, "loss": 0.2549, "step": 7377 }, { "epoch": 2.3816134630130352, "grad_norm": 0.38581554347013397, "learning_rate": 1.6725814963483047e-06, "loss": 0.2835, "step": 7378 }, { "epoch": 2.381936317042657, "grad_norm": 0.38702909426780446, "learning_rate": 1.6709013110073062e-06, "loss": 0.2588, "step": 7379 }, { "epoch": 2.382259171072279, "grad_norm": 0.4090849018006985, "learning_rate": 1.669221864205083e-06, "loss": 0.2593, "step": 7380 }, { "epoch": 2.3825820251019008, "grad_norm": 0.40231785394022174, "learning_rate": 1.6675431561544221e-06, "loss": 0.2686, "step": 7381 }, { "epoch": 2.3829048791315226, "grad_norm": 0.4447236756464392, "learning_rate": 1.6658651870680085e-06, "loss": 0.2046, "step": 7382 }, { "epoch": 2.3832277331611444, "grad_norm": 0.4140364835806922, "learning_rate": 1.6641879571584448e-06, "loss": 0.2519, "step": 7383 }, { "epoch": 2.3835505871907663, "grad_norm": 0.3873969741436372, "learning_rate": 1.6625114666382282e-06, "loss": 0.2377, "step": 7384 }, { "epoch": 2.383873441220388, "grad_norm": 0.37081514153907713, "learning_rate": 1.6608357157197723e-06, "loss": 0.2801, "step": 7385 }, { "epoch": 2.38419629525001, "grad_norm": 0.36119049079931764, "learning_rate": 1.6591607046153873e-06, "loss": 0.2918, "step": 7386 }, { "epoch": 2.384519149279632, "grad_norm": 0.378953968223751, "learning_rate": 1.6574864335373008e-06, "loss": 0.2199, "step": 7387 }, { "epoch": 2.3848420033092537, "grad_norm": 0.35146168899375413, "learning_rate": 1.6558129026976349e-06, "loss": 0.2558, "step": 7388 }, { "epoch": 2.3851648573388755, "grad_norm": 0.3319577629764559, "learning_rate": 1.6541401123084265e-06, "loss": 0.23, "step": 7389 }, { "epoch": 2.3854877113684974, "grad_norm": 0.35133145877262956, "learning_rate": 1.6524680625816154e-06, "loss": 0.2648, "step": 7390 }, { "epoch": 2.385810565398119, "grad_norm": 0.35941955585976676, "learning_rate": 1.6507967537290505e-06, "loss": 0.2434, "step": 7391 }, { "epoch": 2.386133419427741, "grad_norm": 0.4003699199847275, "learning_rate": 1.6491261859624798e-06, "loss": 0.2314, "step": 7392 }, { "epoch": 2.3864562734573633, "grad_norm": 0.3555271783984353, "learning_rate": 1.647456359493565e-06, "loss": 0.2658, "step": 7393 }, { "epoch": 2.386779127486985, "grad_norm": 0.3703563424308721, "learning_rate": 1.6457872745338707e-06, "loss": 0.2842, "step": 7394 }, { "epoch": 2.387101981516607, "grad_norm": 0.3902279313645756, "learning_rate": 1.6441189312948662e-06, "loss": 0.2493, "step": 7395 }, { "epoch": 2.387424835546229, "grad_norm": 0.34648621230063176, "learning_rate": 1.6424513299879302e-06, "loss": 0.2544, "step": 7396 }, { "epoch": 2.3877476895758507, "grad_norm": 0.3595724182498537, "learning_rate": 1.6407844708243433e-06, "loss": 0.2275, "step": 7397 }, { "epoch": 2.3880705436054726, "grad_norm": 0.36479557399613644, "learning_rate": 1.6391183540152968e-06, "loss": 0.2913, "step": 7398 }, { "epoch": 2.3883933976350944, "grad_norm": 0.3415807535719083, "learning_rate": 1.6374529797718818e-06, "loss": 0.2826, "step": 7399 }, { "epoch": 2.3887162516647162, "grad_norm": 0.38096315696179617, "learning_rate": 1.635788348305103e-06, "loss": 0.2374, "step": 7400 }, { "epoch": 2.389039105694338, "grad_norm": 0.3521168292736416, "learning_rate": 1.6341244598258629e-06, "loss": 0.2586, "step": 7401 }, { "epoch": 2.38936195972396, "grad_norm": 0.45084568993853086, "learning_rate": 1.6324613145449781e-06, "loss": 0.2392, "step": 7402 }, { "epoch": 2.3896848137535818, "grad_norm": 0.4202956148798465, "learning_rate": 1.630798912673163e-06, "loss": 0.2794, "step": 7403 }, { "epoch": 2.3900076677832036, "grad_norm": 0.32219123970530883, "learning_rate": 1.6291372544210434e-06, "loss": 0.2112, "step": 7404 }, { "epoch": 2.3903305218128255, "grad_norm": 0.31464226545720986, "learning_rate": 1.6274763399991504e-06, "loss": 0.2514, "step": 7405 }, { "epoch": 2.3906533758424473, "grad_norm": 0.482113532816287, "learning_rate": 1.6258161696179165e-06, "loss": 0.2936, "step": 7406 }, { "epoch": 2.390976229872069, "grad_norm": 0.3559902787421325, "learning_rate": 1.6241567434876845e-06, "loss": 0.2244, "step": 7407 }, { "epoch": 2.391299083901691, "grad_norm": 0.4456920337721557, "learning_rate": 1.6224980618187016e-06, "loss": 0.283, "step": 7408 }, { "epoch": 2.391621937931313, "grad_norm": 0.3910234266909346, "learning_rate": 1.6208401248211225e-06, "loss": 0.2988, "step": 7409 }, { "epoch": 2.3919447919609347, "grad_norm": 0.35299727889912047, "learning_rate": 1.6191829327050014e-06, "loss": 0.2187, "step": 7410 }, { "epoch": 2.3922676459905565, "grad_norm": 0.4201236998742466, "learning_rate": 1.6175264856803066e-06, "loss": 0.248, "step": 7411 }, { "epoch": 2.3925905000201784, "grad_norm": 0.3902846074481439, "learning_rate": 1.6158707839569031e-06, "loss": 0.2404, "step": 7412 }, { "epoch": 2.3929133540498, "grad_norm": 0.43259972059831353, "learning_rate": 1.614215827744571e-06, "loss": 0.246, "step": 7413 }, { "epoch": 2.393236208079422, "grad_norm": 0.4001023731188022, "learning_rate": 1.6125616172529855e-06, "loss": 0.2302, "step": 7414 }, { "epoch": 2.393559062109044, "grad_norm": 0.4324664024126479, "learning_rate": 1.6109081526917386e-06, "loss": 0.2927, "step": 7415 }, { "epoch": 2.3938819161386657, "grad_norm": 0.4461933379791665, "learning_rate": 1.6092554342703165e-06, "loss": 0.2791, "step": 7416 }, { "epoch": 2.3942047701682876, "grad_norm": 0.42387761182437106, "learning_rate": 1.60760346219812e-06, "loss": 0.2453, "step": 7417 }, { "epoch": 2.3945276241979094, "grad_norm": 0.3703134419459878, "learning_rate": 1.6059522366844495e-06, "loss": 0.2781, "step": 7418 }, { "epoch": 2.3948504782275313, "grad_norm": 0.40163306618231376, "learning_rate": 1.6043017579385158e-06, "loss": 0.2303, "step": 7419 }, { "epoch": 2.395173332257153, "grad_norm": 0.3516394756096079, "learning_rate": 1.6026520261694284e-06, "loss": 0.29, "step": 7420 }, { "epoch": 2.395496186286775, "grad_norm": 0.43454698798339036, "learning_rate": 1.6010030415862081e-06, "loss": 0.2389, "step": 7421 }, { "epoch": 2.395819040316397, "grad_norm": 0.374000277552128, "learning_rate": 1.599354804397779e-06, "loss": 0.2942, "step": 7422 }, { "epoch": 2.3961418943460187, "grad_norm": 0.38138629534567003, "learning_rate": 1.5977073148129731e-06, "loss": 0.2046, "step": 7423 }, { "epoch": 2.3964647483756405, "grad_norm": 0.3816149067713875, "learning_rate": 1.59606057304052e-06, "loss": 0.2632, "step": 7424 }, { "epoch": 2.3967876024052623, "grad_norm": 0.3692639299288063, "learning_rate": 1.5944145792890637e-06, "loss": 0.2432, "step": 7425 }, { "epoch": 2.397110456434884, "grad_norm": 0.3782407161695202, "learning_rate": 1.5927693337671468e-06, "loss": 0.2895, "step": 7426 }, { "epoch": 2.397433310464506, "grad_norm": 0.4036189661676799, "learning_rate": 1.5911248366832195e-06, "loss": 0.1885, "step": 7427 }, { "epoch": 2.397756164494128, "grad_norm": 0.42712778566687437, "learning_rate": 1.5894810882456415e-06, "loss": 0.2263, "step": 7428 }, { "epoch": 2.3980790185237497, "grad_norm": 0.3798133558563599, "learning_rate": 1.5878380886626675e-06, "loss": 0.249, "step": 7429 }, { "epoch": 2.3984018725533716, "grad_norm": 0.40551803245928053, "learning_rate": 1.5861958381424679e-06, "loss": 0.2372, "step": 7430 }, { "epoch": 2.398724726582994, "grad_norm": 0.463518765651306, "learning_rate": 1.5845543368931103e-06, "loss": 0.2847, "step": 7431 }, { "epoch": 2.3990475806126157, "grad_norm": 0.36509337791119834, "learning_rate": 1.5829135851225743e-06, "loss": 0.2872, "step": 7432 }, { "epoch": 2.3993704346422375, "grad_norm": 0.35780010211261004, "learning_rate": 1.5812735830387367e-06, "loss": 0.2169, "step": 7433 }, { "epoch": 2.3996932886718594, "grad_norm": 0.37341194909057646, "learning_rate": 1.5796343308493874e-06, "loss": 0.236, "step": 7434 }, { "epoch": 2.4000161427014812, "grad_norm": 0.3704331235409446, "learning_rate": 1.5779958287622135e-06, "loss": 0.2608, "step": 7435 }, { "epoch": 2.400338996731103, "grad_norm": 0.4016917767498653, "learning_rate": 1.5763580769848128e-06, "loss": 0.3186, "step": 7436 }, { "epoch": 2.400661850760725, "grad_norm": 0.4045855312554399, "learning_rate": 1.5747210757246882e-06, "loss": 0.2215, "step": 7437 }, { "epoch": 2.4009847047903468, "grad_norm": 0.32875596077576197, "learning_rate": 1.5730848251892428e-06, "loss": 0.2749, "step": 7438 }, { "epoch": 2.4013075588199686, "grad_norm": 0.33504473843737836, "learning_rate": 1.571449325585787e-06, "loss": 0.2572, "step": 7439 }, { "epoch": 2.4016304128495904, "grad_norm": 0.44856144191286723, "learning_rate": 1.5698145771215406e-06, "loss": 0.2715, "step": 7440 }, { "epoch": 2.4019532668792123, "grad_norm": 0.4078959220814524, "learning_rate": 1.5681805800036185e-06, "loss": 0.3014, "step": 7441 }, { "epoch": 2.402276120908834, "grad_norm": 0.45466723169576984, "learning_rate": 1.56654733443905e-06, "loss": 0.25, "step": 7442 }, { "epoch": 2.402598974938456, "grad_norm": 0.4687503909705532, "learning_rate": 1.5649148406347613e-06, "loss": 0.2404, "step": 7443 }, { "epoch": 2.402921828968078, "grad_norm": 0.38490402540493174, "learning_rate": 1.563283098797592e-06, "loss": 0.263, "step": 7444 }, { "epoch": 2.4032446829976997, "grad_norm": 0.4076719430388774, "learning_rate": 1.5616521091342761e-06, "loss": 0.2572, "step": 7445 }, { "epoch": 2.4035675370273215, "grad_norm": 0.35289942075668757, "learning_rate": 1.5600218718514605e-06, "loss": 0.2399, "step": 7446 }, { "epoch": 2.4038903910569434, "grad_norm": 0.3675661000138694, "learning_rate": 1.558392387155696e-06, "loss": 0.207, "step": 7447 }, { "epoch": 2.404213245086565, "grad_norm": 0.38319126661347586, "learning_rate": 1.5567636552534323e-06, "loss": 0.2521, "step": 7448 }, { "epoch": 2.404536099116187, "grad_norm": 0.3796490141813449, "learning_rate": 1.5551356763510301e-06, "loss": 0.2841, "step": 7449 }, { "epoch": 2.404858953145809, "grad_norm": 0.40266272901360367, "learning_rate": 1.5535084506547497e-06, "loss": 0.2317, "step": 7450 }, { "epoch": 2.4051818071754307, "grad_norm": 0.373742638712196, "learning_rate": 1.551881978370762e-06, "loss": 0.3052, "step": 7451 }, { "epoch": 2.4055046612050526, "grad_norm": 0.3766718603620831, "learning_rate": 1.550256259705135e-06, "loss": 0.266, "step": 7452 }, { "epoch": 2.4058275152346744, "grad_norm": 0.4142828525121827, "learning_rate": 1.5486312948638462e-06, "loss": 0.3084, "step": 7453 }, { "epoch": 2.4061503692642963, "grad_norm": 0.3640897575958807, "learning_rate": 1.5470070840527763e-06, "loss": 0.2457, "step": 7454 }, { "epoch": 2.406473223293918, "grad_norm": 0.38631031373962005, "learning_rate": 1.5453836274777135e-06, "loss": 0.2838, "step": 7455 }, { "epoch": 2.40679607732354, "grad_norm": 0.3965073065138416, "learning_rate": 1.5437609253443432e-06, "loss": 0.2714, "step": 7456 }, { "epoch": 2.407118931353162, "grad_norm": 0.3396771582045619, "learning_rate": 1.5421389778582634e-06, "loss": 0.2186, "step": 7457 }, { "epoch": 2.4074417853827836, "grad_norm": 0.35609255403321943, "learning_rate": 1.540517785224969e-06, "loss": 0.2023, "step": 7458 }, { "epoch": 2.4077646394124055, "grad_norm": 0.3354061502378979, "learning_rate": 1.5388973476498653e-06, "loss": 0.2129, "step": 7459 }, { "epoch": 2.4080874934420278, "grad_norm": 0.4121157599230649, "learning_rate": 1.5372776653382576e-06, "loss": 0.3158, "step": 7460 }, { "epoch": 2.4084103474716496, "grad_norm": 0.41047606528118313, "learning_rate": 1.5356587384953608e-06, "loss": 0.2873, "step": 7461 }, { "epoch": 2.4087332015012715, "grad_norm": 0.45514425253866814, "learning_rate": 1.534040567326286e-06, "loss": 0.2374, "step": 7462 }, { "epoch": 2.4090560555308933, "grad_norm": 0.3262738408509872, "learning_rate": 1.5324231520360578e-06, "loss": 0.2201, "step": 7463 }, { "epoch": 2.409378909560515, "grad_norm": 0.4239205723310455, "learning_rate": 1.5308064928295964e-06, "loss": 0.2384, "step": 7464 }, { "epoch": 2.409701763590137, "grad_norm": 0.37197023994817496, "learning_rate": 1.5291905899117325e-06, "loss": 0.2855, "step": 7465 }, { "epoch": 2.410024617619759, "grad_norm": 0.4050406804337511, "learning_rate": 1.5275754434872e-06, "loss": 0.2114, "step": 7466 }, { "epoch": 2.4103474716493807, "grad_norm": 0.34179392958630517, "learning_rate": 1.5259610537606326e-06, "loss": 0.3263, "step": 7467 }, { "epoch": 2.4106703256790025, "grad_norm": 0.3809307551250663, "learning_rate": 1.5243474209365732e-06, "loss": 0.257, "step": 7468 }, { "epoch": 2.4109931797086244, "grad_norm": 0.4007731141573615, "learning_rate": 1.5227345452194685e-06, "loss": 0.3048, "step": 7469 }, { "epoch": 2.411316033738246, "grad_norm": 0.3365125604108551, "learning_rate": 1.5211224268136634e-06, "loss": 0.2391, "step": 7470 }, { "epoch": 2.411638887767868, "grad_norm": 0.3357875371975666, "learning_rate": 1.5195110659234135e-06, "loss": 0.2461, "step": 7471 }, { "epoch": 2.41196174179749, "grad_norm": 0.3959212385609822, "learning_rate": 1.5179004627528781e-06, "loss": 0.2017, "step": 7472 }, { "epoch": 2.4122845958271117, "grad_norm": 0.3826647575994011, "learning_rate": 1.5162906175061148e-06, "loss": 0.2665, "step": 7473 }, { "epoch": 2.4126074498567336, "grad_norm": 0.39740244654236273, "learning_rate": 1.514681530387091e-06, "loss": 0.22, "step": 7474 }, { "epoch": 2.4129303038863554, "grad_norm": 0.3499987596851546, "learning_rate": 1.5130732015996742e-06, "loss": 0.2507, "step": 7475 }, { "epoch": 2.4132531579159773, "grad_norm": 0.3600730082012549, "learning_rate": 1.5114656313476395e-06, "loss": 0.2714, "step": 7476 }, { "epoch": 2.413576011945599, "grad_norm": 0.37195702339597975, "learning_rate": 1.5098588198346611e-06, "loss": 0.2703, "step": 7477 }, { "epoch": 2.413898865975221, "grad_norm": 0.34765859859030707, "learning_rate": 1.5082527672643222e-06, "loss": 0.288, "step": 7478 }, { "epoch": 2.414221720004843, "grad_norm": 0.38304207254163447, "learning_rate": 1.5066474738401062e-06, "loss": 0.2476, "step": 7479 }, { "epoch": 2.4145445740344647, "grad_norm": 0.4407547999990755, "learning_rate": 1.5050429397654025e-06, "loss": 0.294, "step": 7480 }, { "epoch": 2.4148674280640865, "grad_norm": 0.3555222551495172, "learning_rate": 1.503439165243502e-06, "loss": 0.2344, "step": 7481 }, { "epoch": 2.4151902820937083, "grad_norm": 0.36909839936964495, "learning_rate": 1.5018361504776004e-06, "loss": 0.253, "step": 7482 }, { "epoch": 2.41551313612333, "grad_norm": 0.37473252865591494, "learning_rate": 1.5002338956708013e-06, "loss": 0.277, "step": 7483 }, { "epoch": 2.415835990152952, "grad_norm": 0.3780742243060214, "learning_rate": 1.4986324010261035e-06, "loss": 0.2957, "step": 7484 }, { "epoch": 2.416158844182574, "grad_norm": 0.37660045969636413, "learning_rate": 1.497031666746415e-06, "loss": 0.3079, "step": 7485 }, { "epoch": 2.4164816982121957, "grad_norm": 0.3839237892233625, "learning_rate": 1.4954316930345488e-06, "loss": 0.2921, "step": 7486 }, { "epoch": 2.4168045522418176, "grad_norm": 0.3606502476239153, "learning_rate": 1.4938324800932207e-06, "loss": 0.2556, "step": 7487 }, { "epoch": 2.4171274062714394, "grad_norm": 0.43409787178682324, "learning_rate": 1.4922340281250437e-06, "loss": 0.2423, "step": 7488 }, { "epoch": 2.4174502603010612, "grad_norm": 0.412360620361354, "learning_rate": 1.4906363373325442e-06, "loss": 0.2584, "step": 7489 }, { "epoch": 2.417773114330683, "grad_norm": 0.36803371783990707, "learning_rate": 1.489039407918143e-06, "loss": 0.3248, "step": 7490 }, { "epoch": 2.418095968360305, "grad_norm": 0.32272878292321505, "learning_rate": 1.4874432400841736e-06, "loss": 0.2494, "step": 7491 }, { "epoch": 2.418418822389927, "grad_norm": 0.3786675641119068, "learning_rate": 1.4858478340328637e-06, "loss": 0.2349, "step": 7492 }, { "epoch": 2.4187416764195486, "grad_norm": 0.5014399900836334, "learning_rate": 1.484253189966354e-06, "loss": 0.2931, "step": 7493 }, { "epoch": 2.4190645304491705, "grad_norm": 0.4036089938522227, "learning_rate": 1.4826593080866781e-06, "loss": 0.279, "step": 7494 }, { "epoch": 2.4193873844787923, "grad_norm": 0.4375479091140786, "learning_rate": 1.4810661885957844e-06, "loss": 0.229, "step": 7495 }, { "epoch": 2.419710238508414, "grad_norm": 0.3482883574797761, "learning_rate": 1.4794738316955128e-06, "loss": 0.2364, "step": 7496 }, { "epoch": 2.420033092538036, "grad_norm": 0.4817927504899608, "learning_rate": 1.4778822375876188e-06, "loss": 0.2574, "step": 7497 }, { "epoch": 2.4203559465676583, "grad_norm": 0.34161377837193413, "learning_rate": 1.4762914064737503e-06, "loss": 0.2651, "step": 7498 }, { "epoch": 2.42067880059728, "grad_norm": 0.40030859416996933, "learning_rate": 1.4747013385554654e-06, "loss": 0.2745, "step": 7499 }, { "epoch": 2.421001654626902, "grad_norm": 0.4204803024568245, "learning_rate": 1.4731120340342233e-06, "loss": 0.3006, "step": 7500 }, { "epoch": 2.421324508656524, "grad_norm": 0.37121345023241903, "learning_rate": 1.4715234931113874e-06, "loss": 0.2896, "step": 7501 }, { "epoch": 2.4216473626861457, "grad_norm": 0.3908012987999985, "learning_rate": 1.469935715988225e-06, "loss": 0.2294, "step": 7502 }, { "epoch": 2.4219702167157675, "grad_norm": 0.37294533515851624, "learning_rate": 1.4683487028659013e-06, "loss": 0.2374, "step": 7503 }, { "epoch": 2.4222930707453894, "grad_norm": 0.40304416497478346, "learning_rate": 1.4667624539454933e-06, "loss": 0.2668, "step": 7504 }, { "epoch": 2.422615924775011, "grad_norm": 0.4623058230620425, "learning_rate": 1.465176969427972e-06, "loss": 0.2542, "step": 7505 }, { "epoch": 2.422938778804633, "grad_norm": 0.38033132143078574, "learning_rate": 1.4635922495142195e-06, "loss": 0.2415, "step": 7506 }, { "epoch": 2.423261632834255, "grad_norm": 0.42023109429953176, "learning_rate": 1.4620082944050151e-06, "loss": 0.3105, "step": 7507 }, { "epoch": 2.4235844868638767, "grad_norm": 0.33588621025492693, "learning_rate": 1.4604251043010472e-06, "loss": 0.2464, "step": 7508 }, { "epoch": 2.4239073408934986, "grad_norm": 0.4223440098801463, "learning_rate": 1.4588426794028989e-06, "loss": 0.2798, "step": 7509 }, { "epoch": 2.4242301949231204, "grad_norm": 0.43025076556103486, "learning_rate": 1.4572610199110665e-06, "loss": 0.249, "step": 7510 }, { "epoch": 2.4245530489527423, "grad_norm": 0.35798342948579936, "learning_rate": 1.4556801260259395e-06, "loss": 0.2258, "step": 7511 }, { "epoch": 2.424875902982364, "grad_norm": 0.37932195180408257, "learning_rate": 1.4540999979478187e-06, "loss": 0.2391, "step": 7512 }, { "epoch": 2.425198757011986, "grad_norm": 0.36595842509759446, "learning_rate": 1.4525206358769005e-06, "loss": 0.2615, "step": 7513 }, { "epoch": 2.425521611041608, "grad_norm": 0.36421707893113237, "learning_rate": 1.45094204001329e-06, "loss": 0.2086, "step": 7514 }, { "epoch": 2.4258444650712296, "grad_norm": 0.3300349720965926, "learning_rate": 1.4493642105569945e-06, "loss": 0.2727, "step": 7515 }, { "epoch": 2.4261673191008515, "grad_norm": 0.39529636577628735, "learning_rate": 1.4477871477079203e-06, "loss": 0.2871, "step": 7516 }, { "epoch": 2.4264901731304733, "grad_norm": 0.41339110786589817, "learning_rate": 1.4462108516658796e-06, "loss": 0.2844, "step": 7517 }, { "epoch": 2.426813027160095, "grad_norm": 0.4400239776078566, "learning_rate": 1.4446353226305895e-06, "loss": 0.2609, "step": 7518 }, { "epoch": 2.427135881189717, "grad_norm": 0.39390705776048657, "learning_rate": 1.4430605608016628e-06, "loss": 0.2967, "step": 7519 }, { "epoch": 2.427458735219339, "grad_norm": 0.4548294147577945, "learning_rate": 1.4414865663786225e-06, "loss": 0.2944, "step": 7520 }, { "epoch": 2.4277815892489607, "grad_norm": 0.35357958853883364, "learning_rate": 1.4399133395608931e-06, "loss": 0.2411, "step": 7521 }, { "epoch": 2.4281044432785825, "grad_norm": 0.3530736427089821, "learning_rate": 1.4383408805477962e-06, "loss": 0.2113, "step": 7522 }, { "epoch": 2.4284272973082044, "grad_norm": 0.3826670819394469, "learning_rate": 1.4367691895385648e-06, "loss": 0.2929, "step": 7523 }, { "epoch": 2.4287501513378262, "grad_norm": 0.38343865356874685, "learning_rate": 1.4351982667323255e-06, "loss": 0.2422, "step": 7524 }, { "epoch": 2.429073005367448, "grad_norm": 0.4496914350785894, "learning_rate": 1.4336281123281156e-06, "loss": 0.2593, "step": 7525 }, { "epoch": 2.42939585939707, "grad_norm": 0.42099752063940504, "learning_rate": 1.4320587265248678e-06, "loss": 0.2418, "step": 7526 }, { "epoch": 2.4297187134266918, "grad_norm": 0.386516422322551, "learning_rate": 1.4304901095214262e-06, "loss": 0.2504, "step": 7527 }, { "epoch": 2.430041567456314, "grad_norm": 0.3853675369140167, "learning_rate": 1.428922261516527e-06, "loss": 0.2113, "step": 7528 }, { "epoch": 2.430364421485936, "grad_norm": 0.38409954002900004, "learning_rate": 1.427355182708819e-06, "loss": 0.2401, "step": 7529 }, { "epoch": 2.4306872755155577, "grad_norm": 0.40709762803116406, "learning_rate": 1.4257888732968443e-06, "loss": 0.3043, "step": 7530 }, { "epoch": 2.4310101295451796, "grad_norm": 0.3571600453652071, "learning_rate": 1.424223333479055e-06, "loss": 0.238, "step": 7531 }, { "epoch": 2.4313329835748014, "grad_norm": 0.38123274520874495, "learning_rate": 1.4226585634538029e-06, "loss": 0.2528, "step": 7532 }, { "epoch": 2.4316558376044233, "grad_norm": 0.39067272557341853, "learning_rate": 1.4210945634193435e-06, "loss": 0.2494, "step": 7533 }, { "epoch": 2.431978691634045, "grad_norm": 0.3646869885680612, "learning_rate": 1.41953133357383e-06, "loss": 0.2657, "step": 7534 }, { "epoch": 2.432301545663667, "grad_norm": 0.31774650618912037, "learning_rate": 1.4179688741153248e-06, "loss": 0.2261, "step": 7535 }, { "epoch": 2.432624399693289, "grad_norm": 0.4403305575502273, "learning_rate": 1.4164071852417856e-06, "loss": 0.3133, "step": 7536 }, { "epoch": 2.4329472537229107, "grad_norm": 0.36254703085532797, "learning_rate": 1.414846267151079e-06, "loss": 0.2416, "step": 7537 }, { "epoch": 2.4332701077525325, "grad_norm": 0.37756665833301767, "learning_rate": 1.4132861200409724e-06, "loss": 0.2575, "step": 7538 }, { "epoch": 2.4335929617821543, "grad_norm": 0.34182991732051843, "learning_rate": 1.4117267441091305e-06, "loss": 0.2479, "step": 7539 }, { "epoch": 2.433915815811776, "grad_norm": 0.35762358916771375, "learning_rate": 1.4101681395531286e-06, "loss": 0.2508, "step": 7540 }, { "epoch": 2.434238669841398, "grad_norm": 0.4049908534502677, "learning_rate": 1.4086103065704342e-06, "loss": 0.2056, "step": 7541 }, { "epoch": 2.43456152387102, "grad_norm": 0.37931843733593845, "learning_rate": 1.407053245358428e-06, "loss": 0.2406, "step": 7542 }, { "epoch": 2.4348843779006417, "grad_norm": 0.3906297338704909, "learning_rate": 1.4054969561143832e-06, "loss": 0.2456, "step": 7543 }, { "epoch": 2.4352072319302636, "grad_norm": 0.3608001851151376, "learning_rate": 1.4039414390354836e-06, "loss": 0.2596, "step": 7544 }, { "epoch": 2.4355300859598854, "grad_norm": 0.39541011338478216, "learning_rate": 1.402386694318807e-06, "loss": 0.2612, "step": 7545 }, { "epoch": 2.4358529399895072, "grad_norm": 0.36793531847147665, "learning_rate": 1.400832722161339e-06, "loss": 0.2333, "step": 7546 }, { "epoch": 2.436175794019129, "grad_norm": 0.4322471801388868, "learning_rate": 1.399279522759966e-06, "loss": 0.2824, "step": 7547 }, { "epoch": 2.436498648048751, "grad_norm": 0.38120874876955374, "learning_rate": 1.3977270963114786e-06, "loss": 0.2259, "step": 7548 }, { "epoch": 2.436821502078373, "grad_norm": 0.3968498913974847, "learning_rate": 1.3961754430125624e-06, "loss": 0.2552, "step": 7549 }, { "epoch": 2.4371443561079946, "grad_norm": 0.3551675166376999, "learning_rate": 1.3946245630598139e-06, "loss": 0.2714, "step": 7550 }, { "epoch": 2.4374672101376165, "grad_norm": 0.467786853046526, "learning_rate": 1.3930744566497248e-06, "loss": 0.267, "step": 7551 }, { "epoch": 2.4377900641672383, "grad_norm": 0.3178369661500688, "learning_rate": 1.3915251239786933e-06, "loss": 0.2357, "step": 7552 }, { "epoch": 2.43811291819686, "grad_norm": 0.44971602915913356, "learning_rate": 1.3899765652430153e-06, "loss": 0.2648, "step": 7553 }, { "epoch": 2.438435772226482, "grad_norm": 0.36884807843463624, "learning_rate": 1.3884287806388934e-06, "loss": 0.2374, "step": 7554 }, { "epoch": 2.438758626256104, "grad_norm": 0.3913118242617717, "learning_rate": 1.3868817703624278e-06, "loss": 0.2909, "step": 7555 }, { "epoch": 2.4390814802857257, "grad_norm": 0.3471110226914096, "learning_rate": 1.3853355346096232e-06, "loss": 0.2366, "step": 7556 }, { "epoch": 2.4394043343153475, "grad_norm": 0.3648411540136182, "learning_rate": 1.383790073576388e-06, "loss": 0.2483, "step": 7557 }, { "epoch": 2.4397271883449694, "grad_norm": 0.4389657571215856, "learning_rate": 1.3822453874585263e-06, "loss": 0.2361, "step": 7558 }, { "epoch": 2.440050042374591, "grad_norm": 0.413433931308575, "learning_rate": 1.3807014764517517e-06, "loss": 0.2207, "step": 7559 }, { "epoch": 2.440372896404213, "grad_norm": 0.4020913484179433, "learning_rate": 1.379158340751671e-06, "loss": 0.2319, "step": 7560 }, { "epoch": 2.440695750433835, "grad_norm": 0.39894057855732606, "learning_rate": 1.3776159805538008e-06, "loss": 0.256, "step": 7561 }, { "epoch": 2.4410186044634568, "grad_norm": 0.3876331541300902, "learning_rate": 1.3760743960535536e-06, "loss": 0.2599, "step": 7562 }, { "epoch": 2.4413414584930786, "grad_norm": 0.42767016493092974, "learning_rate": 1.374533587446247e-06, "loss": 0.2592, "step": 7563 }, { "epoch": 2.4416643125227004, "grad_norm": 0.4292827390107482, "learning_rate": 1.3729935549271e-06, "loss": 0.265, "step": 7564 }, { "epoch": 2.4419871665523223, "grad_norm": 0.3805124644956036, "learning_rate": 1.371454298691234e-06, "loss": 0.2256, "step": 7565 }, { "epoch": 2.4423100205819446, "grad_norm": 0.4037199548430224, "learning_rate": 1.3699158189336664e-06, "loss": 0.2227, "step": 7566 }, { "epoch": 2.4426328746115664, "grad_norm": 0.36634755036406735, "learning_rate": 1.3683781158493252e-06, "loss": 0.2374, "step": 7567 }, { "epoch": 2.4429557286411883, "grad_norm": 0.46447174906884575, "learning_rate": 1.3668411896330316e-06, "loss": 0.2655, "step": 7568 }, { "epoch": 2.44327858267081, "grad_norm": 0.4052783391250448, "learning_rate": 1.3653050404795144e-06, "loss": 0.2295, "step": 7569 }, { "epoch": 2.443601436700432, "grad_norm": 0.3699270981722583, "learning_rate": 1.3637696685833993e-06, "loss": 0.2423, "step": 7570 }, { "epoch": 2.443924290730054, "grad_norm": 0.32324830754723893, "learning_rate": 1.3622350741392184e-06, "loss": 0.2279, "step": 7571 }, { "epoch": 2.4442471447596756, "grad_norm": 0.3807902269529668, "learning_rate": 1.3607012573414e-06, "loss": 0.2586, "step": 7572 }, { "epoch": 2.4445699987892975, "grad_norm": 0.35809246006865886, "learning_rate": 1.35916821838428e-06, "loss": 0.2221, "step": 7573 }, { "epoch": 2.4448928528189193, "grad_norm": 0.38173749970192666, "learning_rate": 1.357635957462088e-06, "loss": 0.2897, "step": 7574 }, { "epoch": 2.445215706848541, "grad_norm": 0.4088682026581575, "learning_rate": 1.3561044747689617e-06, "loss": 0.2547, "step": 7575 }, { "epoch": 2.445538560878163, "grad_norm": 0.38039046119468833, "learning_rate": 1.3545737704989404e-06, "loss": 0.2771, "step": 7576 }, { "epoch": 2.445861414907785, "grad_norm": 0.35920779995214813, "learning_rate": 1.3530438448459564e-06, "loss": 0.2356, "step": 7577 }, { "epoch": 2.4461842689374067, "grad_norm": 0.3657900985861722, "learning_rate": 1.3515146980038534e-06, "loss": 0.2478, "step": 7578 }, { "epoch": 2.4465071229670285, "grad_norm": 0.3897405661247495, "learning_rate": 1.3499863301663712e-06, "loss": 0.268, "step": 7579 }, { "epoch": 2.4468299769966504, "grad_norm": 0.3717187567605843, "learning_rate": 1.348458741527153e-06, "loss": 0.2852, "step": 7580 }, { "epoch": 2.4471528310262722, "grad_norm": 0.41774384902241324, "learning_rate": 1.3469319322797396e-06, "loss": 0.3102, "step": 7581 }, { "epoch": 2.447475685055894, "grad_norm": 0.35572588232668456, "learning_rate": 1.3454059026175792e-06, "loss": 0.2223, "step": 7582 }, { "epoch": 2.447798539085516, "grad_norm": 0.39161469583071784, "learning_rate": 1.3438806527340136e-06, "loss": 0.2876, "step": 7583 }, { "epoch": 2.4481213931151378, "grad_norm": 0.37510786977272975, "learning_rate": 1.3423561828222936e-06, "loss": 0.2441, "step": 7584 }, { "epoch": 2.4484442471447596, "grad_norm": 0.3726243596469846, "learning_rate": 1.3408324930755636e-06, "loss": 0.2661, "step": 7585 }, { "epoch": 2.4487671011743815, "grad_norm": 0.3524222327268181, "learning_rate": 1.3393095836868776e-06, "loss": 0.269, "step": 7586 }, { "epoch": 2.4490899552040033, "grad_norm": 0.4038499963724196, "learning_rate": 1.3377874548491814e-06, "loss": 0.2993, "step": 7587 }, { "epoch": 2.449412809233625, "grad_norm": 0.3561896955113587, "learning_rate": 1.3362661067553305e-06, "loss": 0.2445, "step": 7588 }, { "epoch": 2.449735663263247, "grad_norm": 0.3817837757018997, "learning_rate": 1.3347455395980743e-06, "loss": 0.243, "step": 7589 }, { "epoch": 2.450058517292869, "grad_norm": 0.3490148523048441, "learning_rate": 1.3332257535700692e-06, "loss": 0.2677, "step": 7590 }, { "epoch": 2.4503813713224907, "grad_norm": 0.27076473119193833, "learning_rate": 1.3317067488638682e-06, "loss": 0.1977, "step": 7591 }, { "epoch": 2.4507042253521125, "grad_norm": 0.37508068016422935, "learning_rate": 1.330188525671927e-06, "loss": 0.2229, "step": 7592 }, { "epoch": 2.4510270793817344, "grad_norm": 0.376597715985309, "learning_rate": 1.3286710841866054e-06, "loss": 0.2482, "step": 7593 }, { "epoch": 2.451349933411356, "grad_norm": 0.2976757498454606, "learning_rate": 1.3271544246001576e-06, "loss": 0.2129, "step": 7594 }, { "epoch": 2.4516727874409785, "grad_norm": 0.3712514928325076, "learning_rate": 1.3256385471047429e-06, "loss": 0.2724, "step": 7595 }, { "epoch": 2.4519956414706003, "grad_norm": 0.42392253083626935, "learning_rate": 1.324123451892422e-06, "loss": 0.272, "step": 7596 }, { "epoch": 2.452318495500222, "grad_norm": 0.3792330836682424, "learning_rate": 1.3226091391551577e-06, "loss": 0.2417, "step": 7597 }, { "epoch": 2.452641349529844, "grad_norm": 0.3603545774854096, "learning_rate": 1.3210956090848062e-06, "loss": 0.2489, "step": 7598 }, { "epoch": 2.452964203559466, "grad_norm": 0.3992894643360454, "learning_rate": 1.3195828618731334e-06, "loss": 0.2518, "step": 7599 }, { "epoch": 2.4532870575890877, "grad_norm": 0.3913031157371264, "learning_rate": 1.3180708977118002e-06, "loss": 0.2565, "step": 7600 }, { "epoch": 2.4536099116187096, "grad_norm": 0.4184048682724693, "learning_rate": 1.3165597167923726e-06, "loss": 0.2532, "step": 7601 }, { "epoch": 2.4539327656483314, "grad_norm": 0.3876184531374757, "learning_rate": 1.3150493193063114e-06, "loss": 0.2748, "step": 7602 }, { "epoch": 2.4542556196779532, "grad_norm": 0.35859037740666255, "learning_rate": 1.3135397054449872e-06, "loss": 0.2391, "step": 7603 }, { "epoch": 2.454578473707575, "grad_norm": 0.43389877911024155, "learning_rate": 1.3120308753996606e-06, "loss": 0.2532, "step": 7604 }, { "epoch": 2.454901327737197, "grad_norm": 0.3964790353900285, "learning_rate": 1.3105228293615025e-06, "loss": 0.2512, "step": 7605 }, { "epoch": 2.455224181766819, "grad_norm": 0.3991906181621775, "learning_rate": 1.3090155675215776e-06, "loss": 0.2601, "step": 7606 }, { "epoch": 2.4555470357964406, "grad_norm": 0.43636694162121537, "learning_rate": 1.3075090900708553e-06, "loss": 0.2411, "step": 7607 }, { "epoch": 2.4558698898260625, "grad_norm": 0.3611727233976087, "learning_rate": 1.306003397200203e-06, "loss": 0.259, "step": 7608 }, { "epoch": 2.4561927438556843, "grad_norm": 0.3838188035909737, "learning_rate": 1.30449848910039e-06, "loss": 0.2486, "step": 7609 }, { "epoch": 2.456515597885306, "grad_norm": 0.37822523315467393, "learning_rate": 1.302994365962087e-06, "loss": 0.267, "step": 7610 }, { "epoch": 2.456838451914928, "grad_norm": 0.39634062543322374, "learning_rate": 1.3014910279758638e-06, "loss": 0.2914, "step": 7611 }, { "epoch": 2.45716130594455, "grad_norm": 0.4267759837479577, "learning_rate": 1.2999884753321933e-06, "loss": 0.2804, "step": 7612 }, { "epoch": 2.4574841599741717, "grad_norm": 0.4158892338161007, "learning_rate": 1.2984867082214432e-06, "loss": 0.2631, "step": 7613 }, { "epoch": 2.4578070140037935, "grad_norm": 0.3349984761470034, "learning_rate": 1.2969857268338888e-06, "loss": 0.2405, "step": 7614 }, { "epoch": 2.4581298680334154, "grad_norm": 0.4439946863271988, "learning_rate": 1.2954855313596998e-06, "loss": 0.2503, "step": 7615 }, { "epoch": 2.458452722063037, "grad_norm": 0.4133101772052333, "learning_rate": 1.2939861219889512e-06, "loss": 0.3049, "step": 7616 }, { "epoch": 2.458775576092659, "grad_norm": 0.327231619056208, "learning_rate": 1.2924874989116132e-06, "loss": 0.2381, "step": 7617 }, { "epoch": 2.459098430122281, "grad_norm": 0.4068615063732263, "learning_rate": 1.2909896623175617e-06, "loss": 0.2322, "step": 7618 }, { "epoch": 2.4594212841519028, "grad_norm": 0.4257148417284979, "learning_rate": 1.289492612396569e-06, "loss": 0.2786, "step": 7619 }, { "epoch": 2.4597441381815246, "grad_norm": 0.4260689387829159, "learning_rate": 1.2879963493383109e-06, "loss": 0.2556, "step": 7620 }, { "epoch": 2.4600669922111464, "grad_norm": 0.3524530647709192, "learning_rate": 1.2865008733323596e-06, "loss": 0.2504, "step": 7621 }, { "epoch": 2.4603898462407683, "grad_norm": 0.40520134534904306, "learning_rate": 1.2850061845681926e-06, "loss": 0.2178, "step": 7622 }, { "epoch": 2.46071270027039, "grad_norm": 0.369739370365446, "learning_rate": 1.2835122832351817e-06, "loss": 0.2852, "step": 7623 }, { "epoch": 2.461035554300012, "grad_norm": 0.36576854186209284, "learning_rate": 1.2820191695226038e-06, "loss": 0.2183, "step": 7624 }, { "epoch": 2.461358408329634, "grad_norm": 0.32648924416724345, "learning_rate": 1.2805268436196352e-06, "loss": 0.2484, "step": 7625 }, { "epoch": 2.4616812623592557, "grad_norm": 0.36688274985477365, "learning_rate": 1.279035305715352e-06, "loss": 0.3084, "step": 7626 }, { "epoch": 2.4620041163888775, "grad_norm": 0.3767913842345262, "learning_rate": 1.2775445559987272e-06, "loss": 0.249, "step": 7627 }, { "epoch": 2.4623269704184994, "grad_norm": 0.4004311485229556, "learning_rate": 1.2760545946586399e-06, "loss": 0.2587, "step": 7628 }, { "epoch": 2.462649824448121, "grad_norm": 0.41504302415446137, "learning_rate": 1.274565421883863e-06, "loss": 0.2759, "step": 7629 }, { "epoch": 2.462972678477743, "grad_norm": 0.4227527233223398, "learning_rate": 1.2730770378630736e-06, "loss": 0.3307, "step": 7630 }, { "epoch": 2.463295532507365, "grad_norm": 0.36438328848375245, "learning_rate": 1.271589442784851e-06, "loss": 0.2457, "step": 7631 }, { "epoch": 2.4636183865369867, "grad_norm": 0.38389743095494705, "learning_rate": 1.2701026368376673e-06, "loss": 0.2593, "step": 7632 }, { "epoch": 2.463941240566609, "grad_norm": 0.4051084939527261, "learning_rate": 1.2686166202099017e-06, "loss": 0.2236, "step": 7633 }, { "epoch": 2.464264094596231, "grad_norm": 0.3346380284697067, "learning_rate": 1.2671313930898278e-06, "loss": 0.2273, "step": 7634 }, { "epoch": 2.4645869486258527, "grad_norm": 0.3827999362975292, "learning_rate": 1.2656469556656245e-06, "loss": 0.2976, "step": 7635 }, { "epoch": 2.4649098026554745, "grad_norm": 0.4334456979125938, "learning_rate": 1.264163308125365e-06, "loss": 0.2631, "step": 7636 }, { "epoch": 2.4652326566850964, "grad_norm": 0.4128132686154325, "learning_rate": 1.2626804506570295e-06, "loss": 0.2645, "step": 7637 }, { "epoch": 2.4655555107147182, "grad_norm": 0.36163788209491643, "learning_rate": 1.2611983834484896e-06, "loss": 0.2182, "step": 7638 }, { "epoch": 2.46587836474434, "grad_norm": 0.41825406003030985, "learning_rate": 1.2597171066875248e-06, "loss": 0.2372, "step": 7639 }, { "epoch": 2.466201218773962, "grad_norm": 0.3588206077628511, "learning_rate": 1.2582366205618072e-06, "loss": 0.2409, "step": 7640 }, { "epoch": 2.4665240728035838, "grad_norm": 0.36753391051226963, "learning_rate": 1.2567569252589146e-06, "loss": 0.2533, "step": 7641 }, { "epoch": 2.4668469268332056, "grad_norm": 0.3618983123733062, "learning_rate": 1.2552780209663223e-06, "loss": 0.2342, "step": 7642 }, { "epoch": 2.4671697808628275, "grad_norm": 0.3823606086288263, "learning_rate": 1.2537999078714065e-06, "loss": 0.2481, "step": 7643 }, { "epoch": 2.4674926348924493, "grad_norm": 0.38917641381614787, "learning_rate": 1.2523225861614392e-06, "loss": 0.2316, "step": 7644 }, { "epoch": 2.467815488922071, "grad_norm": 0.45318964785670285, "learning_rate": 1.2508460560235991e-06, "loss": 0.2859, "step": 7645 }, { "epoch": 2.468138342951693, "grad_norm": 0.4057962282526761, "learning_rate": 1.2493703176449553e-06, "loss": 0.2802, "step": 7646 }, { "epoch": 2.468461196981315, "grad_norm": 0.343895120498862, "learning_rate": 1.2478953712124869e-06, "loss": 0.2776, "step": 7647 }, { "epoch": 2.4687840510109367, "grad_norm": 0.37871297156662376, "learning_rate": 1.2464212169130635e-06, "loss": 0.217, "step": 7648 }, { "epoch": 2.4691069050405585, "grad_norm": 0.38634028317483926, "learning_rate": 1.2449478549334607e-06, "loss": 0.2971, "step": 7649 }, { "epoch": 2.4694297590701804, "grad_norm": 0.34573797318948907, "learning_rate": 1.2434752854603523e-06, "loss": 0.2126, "step": 7650 }, { "epoch": 2.469752613099802, "grad_norm": 0.3949670773231018, "learning_rate": 1.2420035086803078e-06, "loss": 0.2567, "step": 7651 }, { "epoch": 2.470075467129424, "grad_norm": 0.3544013138909921, "learning_rate": 1.2405325247798027e-06, "loss": 0.2143, "step": 7652 }, { "epoch": 2.470398321159046, "grad_norm": 0.382239038555892, "learning_rate": 1.2390623339452049e-06, "loss": 0.2461, "step": 7653 }, { "epoch": 2.4707211751886677, "grad_norm": 0.43974086791035005, "learning_rate": 1.23759293636279e-06, "loss": 0.2731, "step": 7654 }, { "epoch": 2.4710440292182896, "grad_norm": 0.4667922961309211, "learning_rate": 1.2361243322187235e-06, "loss": 0.2526, "step": 7655 }, { "epoch": 2.4713668832479114, "grad_norm": 0.41799860981260556, "learning_rate": 1.2346565216990791e-06, "loss": 0.258, "step": 7656 }, { "epoch": 2.4716897372775333, "grad_norm": 0.3600473935247985, "learning_rate": 1.2331895049898244e-06, "loss": 0.2681, "step": 7657 }, { "epoch": 2.472012591307155, "grad_norm": 0.3390843951898393, "learning_rate": 1.2317232822768321e-06, "loss": 0.2715, "step": 7658 }, { "epoch": 2.472335445336777, "grad_norm": 0.3854615346020901, "learning_rate": 1.2302578537458665e-06, "loss": 0.2552, "step": 7659 }, { "epoch": 2.472658299366399, "grad_norm": 0.38905596537430764, "learning_rate": 1.2287932195825982e-06, "loss": 0.251, "step": 7660 }, { "epoch": 2.4729811533960206, "grad_norm": 0.41257892465736906, "learning_rate": 1.2273293799725914e-06, "loss": 0.2666, "step": 7661 }, { "epoch": 2.4733040074256425, "grad_norm": 0.45275490250379996, "learning_rate": 1.2258663351013163e-06, "loss": 0.2613, "step": 7662 }, { "epoch": 2.473626861455265, "grad_norm": 0.40028279323821225, "learning_rate": 1.2244040851541343e-06, "loss": 0.3215, "step": 7663 }, { "epoch": 2.4739497154848866, "grad_norm": 0.3658729265659084, "learning_rate": 1.2229426303163141e-06, "loss": 0.2579, "step": 7664 }, { "epoch": 2.4742725695145085, "grad_norm": 0.36829838021386435, "learning_rate": 1.2214819707730174e-06, "loss": 0.332, "step": 7665 }, { "epoch": 2.4745954235441303, "grad_norm": 0.3991730957327342, "learning_rate": 1.2200221067093104e-06, "loss": 0.31, "step": 7666 }, { "epoch": 2.474918277573752, "grad_norm": 0.3873554372563281, "learning_rate": 1.2185630383101534e-06, "loss": 0.287, "step": 7667 }, { "epoch": 2.475241131603374, "grad_norm": 0.3892548368659974, "learning_rate": 1.2171047657604088e-06, "loss": 0.2452, "step": 7668 }, { "epoch": 2.475563985632996, "grad_norm": 0.32146331259008093, "learning_rate": 1.2156472892448401e-06, "loss": 0.2134, "step": 7669 }, { "epoch": 2.4758868396626177, "grad_norm": 0.32256791611048874, "learning_rate": 1.2141906089481039e-06, "loss": 0.2769, "step": 7670 }, { "epoch": 2.4762096936922395, "grad_norm": 0.32196217410941497, "learning_rate": 1.2127347250547619e-06, "loss": 0.3058, "step": 7671 }, { "epoch": 2.4765325477218614, "grad_norm": 0.3797354048484236, "learning_rate": 1.2112796377492738e-06, "loss": 0.2203, "step": 7672 }, { "epoch": 2.476855401751483, "grad_norm": 0.427201483912427, "learning_rate": 1.2098253472159932e-06, "loss": 0.2466, "step": 7673 }, { "epoch": 2.477178255781105, "grad_norm": 0.4263063925239267, "learning_rate": 1.20837185363918e-06, "loss": 0.3175, "step": 7674 }, { "epoch": 2.477501109810727, "grad_norm": 0.3609905065618649, "learning_rate": 1.2069191572029895e-06, "loss": 0.3272, "step": 7675 }, { "epoch": 2.4778239638403488, "grad_norm": 0.437780010735256, "learning_rate": 1.2054672580914744e-06, "loss": 0.2926, "step": 7676 }, { "epoch": 2.4781468178699706, "grad_norm": 0.4453893378038791, "learning_rate": 1.2040161564885919e-06, "loss": 0.303, "step": 7677 }, { "epoch": 2.4784696718995924, "grad_norm": 0.3573201806262209, "learning_rate": 1.2025658525781897e-06, "loss": 0.2565, "step": 7678 }, { "epoch": 2.4787925259292143, "grad_norm": 0.47995968084032875, "learning_rate": 1.201116346544024e-06, "loss": 0.2852, "step": 7679 }, { "epoch": 2.479115379958836, "grad_norm": 0.36415622343679843, "learning_rate": 1.199667638569742e-06, "loss": 0.3178, "step": 7680 }, { "epoch": 2.479438233988458, "grad_norm": 0.4998941520228553, "learning_rate": 1.198219728838895e-06, "loss": 0.2637, "step": 7681 }, { "epoch": 2.47976108801808, "grad_norm": 0.35325582212291295, "learning_rate": 1.1967726175349288e-06, "loss": 0.2356, "step": 7682 }, { "epoch": 2.4800839420477017, "grad_norm": 0.3766406114141265, "learning_rate": 1.1953263048411939e-06, "loss": 0.297, "step": 7683 }, { "epoch": 2.4804067960773235, "grad_norm": 0.4566776649350787, "learning_rate": 1.193880790940932e-06, "loss": 0.2521, "step": 7684 }, { "epoch": 2.4807296501069453, "grad_norm": 0.3985210032469688, "learning_rate": 1.1924360760172899e-06, "loss": 0.2471, "step": 7685 }, { "epoch": 2.481052504136567, "grad_norm": 0.3854696020006282, "learning_rate": 1.1909921602533127e-06, "loss": 0.2931, "step": 7686 }, { "epoch": 2.481375358166189, "grad_norm": 0.36888317371618173, "learning_rate": 1.1895490438319396e-06, "loss": 0.2874, "step": 7687 }, { "epoch": 2.481698212195811, "grad_norm": 0.41326951719881316, "learning_rate": 1.1881067269360122e-06, "loss": 0.2539, "step": 7688 }, { "epoch": 2.4820210662254327, "grad_norm": 0.39829149259726715, "learning_rate": 1.1866652097482703e-06, "loss": 0.2434, "step": 7689 }, { "epoch": 2.4823439202550546, "grad_norm": 0.43297190656145434, "learning_rate": 1.1852244924513536e-06, "loss": 0.2513, "step": 7690 }, { "epoch": 2.4826667742846764, "grad_norm": 0.3691696098128744, "learning_rate": 1.183784575227797e-06, "loss": 0.253, "step": 7691 }, { "epoch": 2.4829896283142983, "grad_norm": 0.4005749043124558, "learning_rate": 1.1823454582600379e-06, "loss": 0.2795, "step": 7692 }, { "epoch": 2.48331248234392, "grad_norm": 0.33737179031143705, "learning_rate": 1.180907141730408e-06, "loss": 0.268, "step": 7693 }, { "epoch": 2.483635336373542, "grad_norm": 0.557365571960817, "learning_rate": 1.1794696258211429e-06, "loss": 0.2574, "step": 7694 }, { "epoch": 2.483958190403164, "grad_norm": 0.6910556013371357, "learning_rate": 1.1780329107143706e-06, "loss": 0.2743, "step": 7695 }, { "epoch": 2.4842810444327856, "grad_norm": 0.3638633393339038, "learning_rate": 1.1765969965921245e-06, "loss": 0.2783, "step": 7696 }, { "epoch": 2.4846038984624075, "grad_norm": 0.3931074354533679, "learning_rate": 1.175161883636329e-06, "loss": 0.2124, "step": 7697 }, { "epoch": 2.4849267524920293, "grad_norm": 0.36972085177735126, "learning_rate": 1.1737275720288154e-06, "loss": 0.2419, "step": 7698 }, { "epoch": 2.485249606521651, "grad_norm": 0.4850397076972308, "learning_rate": 1.1722940619513043e-06, "loss": 0.1893, "step": 7699 }, { "epoch": 2.485572460551273, "grad_norm": 0.3712210894850069, "learning_rate": 1.1708613535854234e-06, "loss": 0.2682, "step": 7700 }, { "epoch": 2.4858953145808953, "grad_norm": 0.4192991761455853, "learning_rate": 1.169429447112692e-06, "loss": 0.2286, "step": 7701 }, { "epoch": 2.486218168610517, "grad_norm": 0.40848623013605073, "learning_rate": 1.1679983427145316e-06, "loss": 0.2626, "step": 7702 }, { "epoch": 2.486541022640139, "grad_norm": 0.41287016890350986, "learning_rate": 1.1665680405722613e-06, "loss": 0.2666, "step": 7703 }, { "epoch": 2.486863876669761, "grad_norm": 0.3272309629004224, "learning_rate": 1.1651385408671008e-06, "loss": 0.2515, "step": 7704 }, { "epoch": 2.4871867306993827, "grad_norm": 0.38238106579376535, "learning_rate": 1.1637098437801616e-06, "loss": 0.2694, "step": 7705 }, { "epoch": 2.4875095847290045, "grad_norm": 0.3738811101698207, "learning_rate": 1.1622819494924594e-06, "loss": 0.2234, "step": 7706 }, { "epoch": 2.4878324387586264, "grad_norm": 0.3885446462802243, "learning_rate": 1.1608548581849087e-06, "loss": 0.2259, "step": 7707 }, { "epoch": 2.488155292788248, "grad_norm": 0.3640869030560212, "learning_rate": 1.1594285700383168e-06, "loss": 0.2362, "step": 7708 }, { "epoch": 2.48847814681787, "grad_norm": 0.39238054764933133, "learning_rate": 1.1580030852333945e-06, "loss": 0.2456, "step": 7709 }, { "epoch": 2.488801000847492, "grad_norm": 0.42116999002937866, "learning_rate": 1.156578403950746e-06, "loss": 0.2619, "step": 7710 }, { "epoch": 2.4891238548771137, "grad_norm": 0.3491773088838707, "learning_rate": 1.1551545263708801e-06, "loss": 0.2654, "step": 7711 }, { "epoch": 2.4894467089067356, "grad_norm": 0.3784070622196331, "learning_rate": 1.1537314526741976e-06, "loss": 0.305, "step": 7712 }, { "epoch": 2.4897695629363574, "grad_norm": 0.3678194033697834, "learning_rate": 1.1523091830410018e-06, "loss": 0.2708, "step": 7713 }, { "epoch": 2.4900924169659793, "grad_norm": 0.379572898704419, "learning_rate": 1.1508877176514893e-06, "loss": 0.2958, "step": 7714 }, { "epoch": 2.490415270995601, "grad_norm": 0.39114500792469103, "learning_rate": 1.149467056685761e-06, "loss": 0.2771, "step": 7715 }, { "epoch": 2.490738125025223, "grad_norm": 0.4156090721820816, "learning_rate": 1.1480472003238105e-06, "loss": 0.2301, "step": 7716 }, { "epoch": 2.491060979054845, "grad_norm": 0.33376318769131397, "learning_rate": 1.1466281487455338e-06, "loss": 0.2364, "step": 7717 }, { "epoch": 2.4913838330844666, "grad_norm": 0.3598554742147162, "learning_rate": 1.1452099021307205e-06, "loss": 0.2436, "step": 7718 }, { "epoch": 2.4917066871140885, "grad_norm": 0.3437612148036076, "learning_rate": 1.1437924606590602e-06, "loss": 0.2535, "step": 7719 }, { "epoch": 2.4920295411437103, "grad_norm": 0.45276869899128164, "learning_rate": 1.1423758245101434e-06, "loss": 0.2222, "step": 7720 }, { "epoch": 2.492352395173332, "grad_norm": 0.3459122148865155, "learning_rate": 1.1409599938634556e-06, "loss": 0.2702, "step": 7721 }, { "epoch": 2.492675249202954, "grad_norm": 0.3804784128502778, "learning_rate": 1.1395449688983785e-06, "loss": 0.2421, "step": 7722 }, { "epoch": 2.492998103232576, "grad_norm": 0.6111256388798738, "learning_rate": 1.1381307497941946e-06, "loss": 0.3071, "step": 7723 }, { "epoch": 2.4933209572621977, "grad_norm": 0.3881437668658401, "learning_rate": 1.136717336730086e-06, "loss": 0.2058, "step": 7724 }, { "epoch": 2.4936438112918196, "grad_norm": 0.38815087389736824, "learning_rate": 1.1353047298851262e-06, "loss": 0.2179, "step": 7725 }, { "epoch": 2.4939666653214414, "grad_norm": 0.34964139786250303, "learning_rate": 1.1338929294382942e-06, "loss": 0.2636, "step": 7726 }, { "epoch": 2.4942895193510632, "grad_norm": 0.398929921412557, "learning_rate": 1.1324819355684597e-06, "loss": 0.2383, "step": 7727 }, { "epoch": 2.494612373380685, "grad_norm": 0.3638711964242119, "learning_rate": 1.1310717484543964e-06, "loss": 0.2731, "step": 7728 }, { "epoch": 2.494935227410307, "grad_norm": 0.4308652140908285, "learning_rate": 1.129662368274771e-06, "loss": 0.2769, "step": 7729 }, { "epoch": 2.495258081439929, "grad_norm": 0.35703439714425045, "learning_rate": 1.1282537952081525e-06, "loss": 0.2962, "step": 7730 }, { "epoch": 2.495580935469551, "grad_norm": 0.38452144150121664, "learning_rate": 1.1268460294330021e-06, "loss": 0.2296, "step": 7731 }, { "epoch": 2.495903789499173, "grad_norm": 0.42511611464607946, "learning_rate": 1.1254390711276852e-06, "loss": 0.2561, "step": 7732 }, { "epoch": 2.4962266435287948, "grad_norm": 0.33032828829234784, "learning_rate": 1.1240329204704583e-06, "loss": 0.2378, "step": 7733 }, { "epoch": 2.4965494975584166, "grad_norm": 0.393176002457963, "learning_rate": 1.1226275776394798e-06, "loss": 0.2533, "step": 7734 }, { "epoch": 2.4968723515880384, "grad_norm": 0.46227830413815346, "learning_rate": 1.1212230428128049e-06, "loss": 0.2204, "step": 7735 }, { "epoch": 2.4971952056176603, "grad_norm": 0.3236575869815875, "learning_rate": 1.1198193161683876e-06, "loss": 0.2411, "step": 7736 }, { "epoch": 2.497518059647282, "grad_norm": 0.4021138154782673, "learning_rate": 1.1184163978840759e-06, "loss": 0.2959, "step": 7737 }, { "epoch": 2.497840913676904, "grad_norm": 0.3327783506636752, "learning_rate": 1.1170142881376208e-06, "loss": 0.2356, "step": 7738 }, { "epoch": 2.498163767706526, "grad_norm": 0.36374846126675336, "learning_rate": 1.1156129871066628e-06, "loss": 0.2653, "step": 7739 }, { "epoch": 2.4984866217361477, "grad_norm": 0.46364165555534165, "learning_rate": 1.1142124949687483e-06, "loss": 0.2578, "step": 7740 }, { "epoch": 2.4988094757657695, "grad_norm": 0.38384766073739285, "learning_rate": 1.1128128119013184e-06, "loss": 0.2872, "step": 7741 }, { "epoch": 2.4991323297953913, "grad_norm": 0.3783381396810294, "learning_rate": 1.1114139380817074e-06, "loss": 0.2051, "step": 7742 }, { "epoch": 2.499455183825013, "grad_norm": 0.37379000793075656, "learning_rate": 1.1100158736871549e-06, "loss": 0.2629, "step": 7743 }, { "epoch": 2.499778037854635, "grad_norm": 0.38983266879672296, "learning_rate": 1.1086186188947902e-06, "loss": 0.2663, "step": 7744 }, { "epoch": 2.500100891884257, "grad_norm": 0.36220494298320743, "learning_rate": 1.107222173881646e-06, "loss": 0.2948, "step": 7745 }, { "epoch": 2.5004237459138787, "grad_norm": 0.3359190026716148, "learning_rate": 1.105826538824648e-06, "loss": 0.2457, "step": 7746 }, { "epoch": 2.5007465999435006, "grad_norm": 0.3486114299237988, "learning_rate": 1.1044317139006232e-06, "loss": 0.2984, "step": 7747 }, { "epoch": 2.5010694539731224, "grad_norm": 0.318527351776449, "learning_rate": 1.103037699286292e-06, "loss": 0.24, "step": 7748 }, { "epoch": 2.5013923080027443, "grad_norm": 0.37054367011596023, "learning_rate": 1.1016444951582754e-06, "loss": 0.2663, "step": 7749 }, { "epoch": 2.501715162032366, "grad_norm": 2.46920058461666, "learning_rate": 1.1002521016930916e-06, "loss": 0.2553, "step": 7750 }, { "epoch": 2.502038016061988, "grad_norm": 0.3630393774988838, "learning_rate": 1.098860519067151e-06, "loss": 0.2616, "step": 7751 }, { "epoch": 2.50236087009161, "grad_norm": 0.33385607289709773, "learning_rate": 1.0974697474567686e-06, "loss": 0.2571, "step": 7752 }, { "epoch": 2.5026837241212316, "grad_norm": 0.4888903126698457, "learning_rate": 1.096079787038154e-06, "loss": 0.2654, "step": 7753 }, { "epoch": 2.5030065781508535, "grad_norm": 0.44784965799988713, "learning_rate": 1.0946906379874097e-06, "loss": 0.2649, "step": 7754 }, { "epoch": 2.5033294321804753, "grad_norm": 0.3658299951400164, "learning_rate": 1.0933023004805435e-06, "loss": 0.2954, "step": 7755 }, { "epoch": 2.503652286210097, "grad_norm": 0.4286607502132568, "learning_rate": 1.091914774693451e-06, "loss": 0.2737, "step": 7756 }, { "epoch": 2.503975140239719, "grad_norm": 0.3532025012395527, "learning_rate": 1.0905280608019336e-06, "loss": 0.2362, "step": 7757 }, { "epoch": 2.504297994269341, "grad_norm": 0.3515448611512862, "learning_rate": 1.089142158981683e-06, "loss": 0.2784, "step": 7758 }, { "epoch": 2.5046208482989627, "grad_norm": 0.39709130484914773, "learning_rate": 1.0877570694082923e-06, "loss": 0.2636, "step": 7759 }, { "epoch": 2.5049437023285845, "grad_norm": 0.3637427904595242, "learning_rate": 1.0863727922572525e-06, "loss": 0.211, "step": 7760 }, { "epoch": 2.5052665563582064, "grad_norm": 0.39184159667864865, "learning_rate": 1.0849893277039461e-06, "loss": 0.2768, "step": 7761 }, { "epoch": 2.5055894103878282, "grad_norm": 0.3830234246768077, "learning_rate": 1.08360667592366e-06, "loss": 0.2816, "step": 7762 }, { "epoch": 2.50591226441745, "grad_norm": 0.4281466962614907, "learning_rate": 1.0822248370915706e-06, "loss": 0.2279, "step": 7763 }, { "epoch": 2.506235118447072, "grad_norm": 0.3618655908054842, "learning_rate": 1.0808438113827571e-06, "loss": 0.2779, "step": 7764 }, { "epoch": 2.5065579724766938, "grad_norm": 0.35754828470581435, "learning_rate": 1.0794635989721923e-06, "loss": 0.2484, "step": 7765 }, { "epoch": 2.5068808265063156, "grad_norm": 0.38934713665399584, "learning_rate": 1.0780842000347474e-06, "loss": 0.2895, "step": 7766 }, { "epoch": 2.5072036805359375, "grad_norm": 0.42414761847640337, "learning_rate": 1.0767056147451915e-06, "loss": 0.2632, "step": 7767 }, { "epoch": 2.5075265345655593, "grad_norm": 0.4110384084403437, "learning_rate": 1.07532784327819e-06, "loss": 0.2454, "step": 7768 }, { "epoch": 2.507849388595181, "grad_norm": 0.4056159630219674, "learning_rate": 1.0739508858083013e-06, "loss": 0.2818, "step": 7769 }, { "epoch": 2.508172242624803, "grad_norm": 0.41093329394089356, "learning_rate": 1.0725747425099888e-06, "loss": 0.2609, "step": 7770 }, { "epoch": 2.5084950966544253, "grad_norm": 0.4184331898069117, "learning_rate": 1.0711994135576036e-06, "loss": 0.1991, "step": 7771 }, { "epoch": 2.508817950684047, "grad_norm": 0.4959643419183895, "learning_rate": 1.0698248991254006e-06, "loss": 0.2272, "step": 7772 }, { "epoch": 2.509140804713669, "grad_norm": 0.4116490772128671, "learning_rate": 1.068451199387527e-06, "loss": 0.2606, "step": 7773 }, { "epoch": 2.509463658743291, "grad_norm": 0.3722174275779911, "learning_rate": 1.0670783145180312e-06, "loss": 0.2622, "step": 7774 }, { "epoch": 2.5097865127729126, "grad_norm": 0.3923078982266304, "learning_rate": 1.0657062446908526e-06, "loss": 0.2709, "step": 7775 }, { "epoch": 2.5101093668025345, "grad_norm": 0.4052088323217145, "learning_rate": 1.0643349900798341e-06, "loss": 0.2824, "step": 7776 }, { "epoch": 2.5104322208321563, "grad_norm": 0.5652920564640646, "learning_rate": 1.062964550858708e-06, "loss": 0.2394, "step": 7777 }, { "epoch": 2.510755074861778, "grad_norm": 0.42584121198560554, "learning_rate": 1.0615949272011085e-06, "loss": 0.2015, "step": 7778 }, { "epoch": 2.5110779288914, "grad_norm": 0.34423487873223907, "learning_rate": 1.0602261192805678e-06, "loss": 0.2298, "step": 7779 }, { "epoch": 2.511400782921022, "grad_norm": 0.3486889179178638, "learning_rate": 1.058858127270508e-06, "loss": 0.2354, "step": 7780 }, { "epoch": 2.5117236369506437, "grad_norm": 0.31511476448934594, "learning_rate": 1.057490951344253e-06, "loss": 0.2256, "step": 7781 }, { "epoch": 2.5120464909802656, "grad_norm": 0.41428589877206934, "learning_rate": 1.0561245916750248e-06, "loss": 0.2347, "step": 7782 }, { "epoch": 2.5123693450098874, "grad_norm": 0.37136400214698884, "learning_rate": 1.0547590484359352e-06, "loss": 0.2707, "step": 7783 }, { "epoch": 2.5126921990395092, "grad_norm": 0.43084292837156746, "learning_rate": 1.0533943217999989e-06, "loss": 0.2158, "step": 7784 }, { "epoch": 2.513015053069131, "grad_norm": 0.37854724361240855, "learning_rate": 1.052030411940126e-06, "loss": 0.2234, "step": 7785 }, { "epoch": 2.513337907098753, "grad_norm": 0.40813712386631124, "learning_rate": 1.0506673190291185e-06, "loss": 0.3111, "step": 7786 }, { "epoch": 2.5136607611283748, "grad_norm": 0.3620787148957638, "learning_rate": 1.0493050432396822e-06, "loss": 0.2421, "step": 7787 }, { "epoch": 2.5139836151579966, "grad_norm": 0.4178622765919232, "learning_rate": 1.0479435847444124e-06, "loss": 0.2914, "step": 7788 }, { "epoch": 2.5143064691876185, "grad_norm": 0.38844421157271997, "learning_rate": 1.046582943715807e-06, "loss": 0.3205, "step": 7789 }, { "epoch": 2.5146293232172403, "grad_norm": 0.4002328625209632, "learning_rate": 1.0452231203262538e-06, "loss": 0.2765, "step": 7790 }, { "epoch": 2.514952177246862, "grad_norm": 0.38940692685360634, "learning_rate": 1.0438641147480444e-06, "loss": 0.3101, "step": 7791 }, { "epoch": 2.515275031276484, "grad_norm": 0.37027939835883467, "learning_rate": 1.04250592715336e-06, "loss": 0.241, "step": 7792 }, { "epoch": 2.515597885306106, "grad_norm": 0.42448567552979666, "learning_rate": 1.041148557714283e-06, "loss": 0.2796, "step": 7793 }, { "epoch": 2.5159207393357277, "grad_norm": 0.37840065617044494, "learning_rate": 1.0397920066027874e-06, "loss": 0.241, "step": 7794 }, { "epoch": 2.5162435933653495, "grad_norm": 0.38151627200275645, "learning_rate": 1.0384362739907493e-06, "loss": 0.2731, "step": 7795 }, { "epoch": 2.516566447394972, "grad_norm": 0.36239013572615003, "learning_rate": 1.037081360049938e-06, "loss": 0.2523, "step": 7796 }, { "epoch": 2.5168893014245937, "grad_norm": 0.37634553839469637, "learning_rate": 1.0357272649520175e-06, "loss": 0.2839, "step": 7797 }, { "epoch": 2.5172121554542155, "grad_norm": 0.37149099029310095, "learning_rate": 1.03437398886855e-06, "loss": 0.2201, "step": 7798 }, { "epoch": 2.5175350094838373, "grad_norm": 0.40468905017912826, "learning_rate": 1.033021531970995e-06, "loss": 0.2406, "step": 7799 }, { "epoch": 2.517857863513459, "grad_norm": 0.3530776690558432, "learning_rate": 1.0316698944307076e-06, "loss": 0.2369, "step": 7800 }, { "epoch": 2.518180717543081, "grad_norm": 0.5649339491677771, "learning_rate": 1.0303190764189354e-06, "loss": 0.2718, "step": 7801 }, { "epoch": 2.518503571572703, "grad_norm": 0.3849109349891044, "learning_rate": 1.028969078106829e-06, "loss": 0.2781, "step": 7802 }, { "epoch": 2.5188264256023247, "grad_norm": 0.4016838898352562, "learning_rate": 1.0276198996654272e-06, "loss": 0.2685, "step": 7803 }, { "epoch": 2.5191492796319466, "grad_norm": 0.3150158947978242, "learning_rate": 1.0262715412656732e-06, "loss": 0.2536, "step": 7804 }, { "epoch": 2.5194721336615684, "grad_norm": 0.4052736932417823, "learning_rate": 1.024924003078398e-06, "loss": 0.2663, "step": 7805 }, { "epoch": 2.5197949876911903, "grad_norm": 0.4447936822242498, "learning_rate": 1.0235772852743367e-06, "loss": 0.2349, "step": 7806 }, { "epoch": 2.520117841720812, "grad_norm": 0.36687145354626155, "learning_rate": 1.0222313880241128e-06, "loss": 0.2933, "step": 7807 }, { "epoch": 2.520440695750434, "grad_norm": 0.3683184595919791, "learning_rate": 1.0208863114982538e-06, "loss": 0.2645, "step": 7808 }, { "epoch": 2.520763549780056, "grad_norm": 0.42610093592594434, "learning_rate": 1.0195420558671747e-06, "loss": 0.2674, "step": 7809 }, { "epoch": 2.5210864038096776, "grad_norm": 0.4333093415609584, "learning_rate": 1.0181986213011948e-06, "loss": 0.2397, "step": 7810 }, { "epoch": 2.5214092578392995, "grad_norm": 0.3547231534664446, "learning_rate": 1.0168560079705222e-06, "loss": 0.2527, "step": 7811 }, { "epoch": 2.5217321118689213, "grad_norm": 0.3659070678637137, "learning_rate": 1.0155142160452658e-06, "loss": 0.3044, "step": 7812 }, { "epoch": 2.522054965898543, "grad_norm": 0.40506722910589327, "learning_rate": 1.0141732456954287e-06, "loss": 0.2218, "step": 7813 }, { "epoch": 2.522377819928165, "grad_norm": 0.4122804706796278, "learning_rate": 1.0128330970909092e-06, "loss": 0.2944, "step": 7814 }, { "epoch": 2.522700673957787, "grad_norm": 0.3906375689594309, "learning_rate": 1.0114937704015049e-06, "loss": 0.2308, "step": 7815 }, { "epoch": 2.5230235279874087, "grad_norm": 0.36126966457025816, "learning_rate": 1.0101552657969023e-06, "loss": 0.2861, "step": 7816 }, { "epoch": 2.5233463820170305, "grad_norm": 0.4337679451107268, "learning_rate": 1.008817583446693e-06, "loss": 0.2002, "step": 7817 }, { "epoch": 2.5236692360466524, "grad_norm": 0.41844595669065343, "learning_rate": 1.0074807235203556e-06, "loss": 0.2863, "step": 7818 }, { "epoch": 2.5239920900762742, "grad_norm": 0.35124356477741636, "learning_rate": 1.0061446861872704e-06, "loss": 0.2807, "step": 7819 }, { "epoch": 2.524314944105896, "grad_norm": 0.4214668944632894, "learning_rate": 1.0048094716167097e-06, "loss": 0.2748, "step": 7820 }, { "epoch": 2.524637798135518, "grad_norm": 0.38687733233883653, "learning_rate": 1.003475079977847e-06, "loss": 0.2453, "step": 7821 }, { "epoch": 2.5249606521651398, "grad_norm": 0.35279631710680864, "learning_rate": 1.0021415114397425e-06, "loss": 0.3148, "step": 7822 }, { "epoch": 2.5252835061947616, "grad_norm": 0.3320051227390806, "learning_rate": 1.000808766171363e-06, "loss": 0.2349, "step": 7823 }, { "epoch": 2.5256063602243835, "grad_norm": 0.4385090218162292, "learning_rate": 9.99476844341561e-07, "loss": 0.2768, "step": 7824 }, { "epoch": 2.5259292142540053, "grad_norm": 0.41322538553373234, "learning_rate": 9.981457461190917e-07, "loss": 0.3184, "step": 7825 }, { "epoch": 2.526252068283627, "grad_norm": 0.42580460593087777, "learning_rate": 9.968154716726026e-07, "loss": 0.256, "step": 7826 }, { "epoch": 2.526574922313249, "grad_norm": 0.3749606930859922, "learning_rate": 9.954860211706372e-07, "loss": 0.272, "step": 7827 }, { "epoch": 2.526897776342871, "grad_norm": 0.3322039237333728, "learning_rate": 9.941573947816377e-07, "loss": 0.2172, "step": 7828 }, { "epoch": 2.5272206303724927, "grad_norm": 0.37004722930146317, "learning_rate": 9.928295926739348e-07, "loss": 0.2917, "step": 7829 }, { "epoch": 2.5275434844021145, "grad_norm": 0.4242394688919767, "learning_rate": 9.915026150157622e-07, "loss": 0.2147, "step": 7830 }, { "epoch": 2.5278663384317364, "grad_norm": 0.34152096818591526, "learning_rate": 9.901764619752472e-07, "loss": 0.2636, "step": 7831 }, { "epoch": 2.528189192461358, "grad_norm": 0.38624755400679595, "learning_rate": 9.888511337204075e-07, "loss": 0.2422, "step": 7832 }, { "epoch": 2.52851204649098, "grad_norm": 0.4053176975034907, "learning_rate": 9.875266304191635e-07, "loss": 0.2788, "step": 7833 }, { "epoch": 2.528834900520602, "grad_norm": 0.4073732012437412, "learning_rate": 9.86202952239329e-07, "loss": 0.2569, "step": 7834 }, { "epoch": 2.5291577545502237, "grad_norm": 0.41174736405806406, "learning_rate": 9.848800993486083e-07, "loss": 0.2673, "step": 7835 }, { "epoch": 2.5294806085798456, "grad_norm": 0.4101123104452711, "learning_rate": 9.835580719146086e-07, "loss": 0.2977, "step": 7836 }, { "epoch": 2.5298034626094674, "grad_norm": 0.4007175475238962, "learning_rate": 9.822368701048265e-07, "loss": 0.2918, "step": 7837 }, { "epoch": 2.5301263166390897, "grad_norm": 0.64975290888761, "learning_rate": 9.809164940866586e-07, "loss": 0.2139, "step": 7838 }, { "epoch": 2.5304491706687116, "grad_norm": 0.3637335287448008, "learning_rate": 9.795969440273927e-07, "loss": 0.2874, "step": 7839 }, { "epoch": 2.5307720246983334, "grad_norm": 0.3457681062200886, "learning_rate": 9.782782200942161e-07, "loss": 0.1917, "step": 7840 }, { "epoch": 2.5310948787279552, "grad_norm": 0.3482132514375453, "learning_rate": 9.76960322454206e-07, "loss": 0.3153, "step": 7841 }, { "epoch": 2.531417732757577, "grad_norm": 0.3186164332106436, "learning_rate": 9.756432512743418e-07, "loss": 0.2153, "step": 7842 }, { "epoch": 2.531740586787199, "grad_norm": 0.37869119777278637, "learning_rate": 9.743270067214921e-07, "loss": 0.2455, "step": 7843 }, { "epoch": 2.5320634408168208, "grad_norm": 0.38830155112779663, "learning_rate": 9.730115889624242e-07, "loss": 0.3215, "step": 7844 }, { "epoch": 2.5323862948464426, "grad_norm": 0.4123715289927678, "learning_rate": 9.716969981637997e-07, "loss": 0.2638, "step": 7845 }, { "epoch": 2.5327091488760645, "grad_norm": 0.38712704734555703, "learning_rate": 9.70383234492177e-07, "loss": 0.2372, "step": 7846 }, { "epoch": 2.5330320029056863, "grad_norm": 0.511853669197285, "learning_rate": 9.690702981140056e-07, "loss": 0.2817, "step": 7847 }, { "epoch": 2.533354856935308, "grad_norm": 0.4406254458139783, "learning_rate": 9.677581891956349e-07, "loss": 0.2402, "step": 7848 }, { "epoch": 2.53367771096493, "grad_norm": 0.38681808670165946, "learning_rate": 9.664469079033045e-07, "loss": 0.2435, "step": 7849 }, { "epoch": 2.534000564994552, "grad_norm": 0.4517516280739236, "learning_rate": 9.651364544031555e-07, "loss": 0.2515, "step": 7850 }, { "epoch": 2.5343234190241737, "grad_norm": 0.4079985254101907, "learning_rate": 9.638268288612168e-07, "loss": 0.2232, "step": 7851 }, { "epoch": 2.5346462730537955, "grad_norm": 0.3911702937468274, "learning_rate": 9.625180314434181e-07, "loss": 0.2262, "step": 7852 }, { "epoch": 2.5349691270834174, "grad_norm": 0.36191405140989746, "learning_rate": 9.612100623155836e-07, "loss": 0.2502, "step": 7853 }, { "epoch": 2.535291981113039, "grad_norm": 0.3798050100506279, "learning_rate": 9.599029216434291e-07, "loss": 0.2661, "step": 7854 }, { "epoch": 2.535614835142661, "grad_norm": 0.37102414114311977, "learning_rate": 9.585966095925685e-07, "loss": 0.2373, "step": 7855 }, { "epoch": 2.535937689172283, "grad_norm": 0.41468429321640815, "learning_rate": 9.57291126328509e-07, "loss": 0.2264, "step": 7856 }, { "epoch": 2.5362605432019047, "grad_norm": 0.4443725393473458, "learning_rate": 9.559864720166545e-07, "loss": 0.3048, "step": 7857 }, { "epoch": 2.5365833972315266, "grad_norm": 0.4134520102291026, "learning_rate": 9.546826468223014e-07, "loss": 0.2612, "step": 7858 }, { "epoch": 2.5369062512611484, "grad_norm": 0.4060169641017828, "learning_rate": 9.533796509106432e-07, "loss": 0.21, "step": 7859 }, { "epoch": 2.5372291052907703, "grad_norm": 0.40280817962121235, "learning_rate": 9.520774844467681e-07, "loss": 0.2602, "step": 7860 }, { "epoch": 2.537551959320392, "grad_norm": 0.3542273638159377, "learning_rate": 9.507761475956605e-07, "loss": 0.2725, "step": 7861 }, { "epoch": 2.537874813350014, "grad_norm": 0.43081366686029426, "learning_rate": 9.494756405221938e-07, "loss": 0.2957, "step": 7862 }, { "epoch": 2.5381976673796363, "grad_norm": 0.3815032236129567, "learning_rate": 9.481759633911452e-07, "loss": 0.2415, "step": 7863 }, { "epoch": 2.538520521409258, "grad_norm": 0.4199172739843309, "learning_rate": 9.46877116367178e-07, "loss": 0.2941, "step": 7864 }, { "epoch": 2.53884337543888, "grad_norm": 0.3898140542419139, "learning_rate": 9.455790996148572e-07, "loss": 0.2038, "step": 7865 }, { "epoch": 2.539166229468502, "grad_norm": 0.36645953115441443, "learning_rate": 9.442819132986372e-07, "loss": 0.2737, "step": 7866 }, { "epoch": 2.5394890834981236, "grad_norm": 0.35688069707468395, "learning_rate": 9.42985557582873e-07, "loss": 0.2942, "step": 7867 }, { "epoch": 2.5398119375277455, "grad_norm": 0.4063156630581738, "learning_rate": 9.416900326318067e-07, "loss": 0.2419, "step": 7868 }, { "epoch": 2.5401347915573673, "grad_norm": 0.3092936609861036, "learning_rate": 9.403953386095846e-07, "loss": 0.238, "step": 7869 }, { "epoch": 2.540457645586989, "grad_norm": 0.34847606085273186, "learning_rate": 9.391014756802369e-07, "loss": 0.2533, "step": 7870 }, { "epoch": 2.540780499616611, "grad_norm": 0.34357991199633603, "learning_rate": 9.378084440076983e-07, "loss": 0.2425, "step": 7871 }, { "epoch": 2.541103353646233, "grad_norm": 0.4013421355971165, "learning_rate": 9.365162437557948e-07, "loss": 0.2801, "step": 7872 }, { "epoch": 2.5414262076758547, "grad_norm": 0.35288254283372866, "learning_rate": 9.352248750882425e-07, "loss": 0.2501, "step": 7873 }, { "epoch": 2.5417490617054765, "grad_norm": 0.39465797036970257, "learning_rate": 9.339343381686608e-07, "loss": 0.292, "step": 7874 }, { "epoch": 2.5420719157350984, "grad_norm": 0.3442382562129394, "learning_rate": 9.326446331605546e-07, "loss": 0.2684, "step": 7875 }, { "epoch": 2.5423947697647202, "grad_norm": 0.35640662052533223, "learning_rate": 9.31355760227329e-07, "loss": 0.2653, "step": 7876 }, { "epoch": 2.542717623794342, "grad_norm": 0.3186675505872343, "learning_rate": 9.300677195322834e-07, "loss": 0.2578, "step": 7877 }, { "epoch": 2.543040477823964, "grad_norm": 0.3294881467906031, "learning_rate": 9.287805112386125e-07, "loss": 0.2173, "step": 7878 }, { "epoch": 2.5433633318535858, "grad_norm": 0.3315931918826931, "learning_rate": 9.274941355093989e-07, "loss": 0.2628, "step": 7879 }, { "epoch": 2.5436861858832076, "grad_norm": 0.4032045815826182, "learning_rate": 9.262085925076305e-07, "loss": 0.2715, "step": 7880 }, { "epoch": 2.5440090399128295, "grad_norm": 0.4138723677470853, "learning_rate": 9.249238823961781e-07, "loss": 0.2691, "step": 7881 }, { "epoch": 2.5443318939424513, "grad_norm": 0.35687926715705653, "learning_rate": 9.236400053378172e-07, "loss": 0.2496, "step": 7882 }, { "epoch": 2.544654747972073, "grad_norm": 0.4164827441703173, "learning_rate": 9.223569614952102e-07, "loss": 0.2611, "step": 7883 }, { "epoch": 2.544977602001695, "grad_norm": 0.37208433836135707, "learning_rate": 9.210747510309195e-07, "loss": 0.2853, "step": 7884 }, { "epoch": 2.545300456031317, "grad_norm": 0.3843770762313891, "learning_rate": 9.197933741073969e-07, "loss": 0.2338, "step": 7885 }, { "epoch": 2.5456233100609387, "grad_norm": 0.39297709550196197, "learning_rate": 9.185128308869933e-07, "loss": 0.2177, "step": 7886 }, { "epoch": 2.5459461640905605, "grad_norm": 0.3685155970500108, "learning_rate": 9.172331215319499e-07, "loss": 0.2492, "step": 7887 }, { "epoch": 2.5462690181201824, "grad_norm": 0.3315827848998688, "learning_rate": 9.159542462044043e-07, "loss": 0.2791, "step": 7888 }, { "epoch": 2.546591872149804, "grad_norm": 0.44853012008892806, "learning_rate": 9.146762050663904e-07, "loss": 0.288, "step": 7889 }, { "epoch": 2.546914726179426, "grad_norm": 0.4533916864792484, "learning_rate": 9.133989982798318e-07, "loss": 0.2298, "step": 7890 }, { "epoch": 2.547237580209048, "grad_norm": 0.4281195361618767, "learning_rate": 9.121226260065501e-07, "loss": 0.2837, "step": 7891 }, { "epoch": 2.5475604342386697, "grad_norm": 0.38448594238376343, "learning_rate": 9.108470884082592e-07, "loss": 0.2787, "step": 7892 }, { "epoch": 2.5478832882682916, "grad_norm": 0.37598805622980846, "learning_rate": 9.095723856465698e-07, "loss": 0.2249, "step": 7893 }, { "epoch": 2.5482061422979134, "grad_norm": 0.404306639843327, "learning_rate": 9.082985178829819e-07, "loss": 0.2725, "step": 7894 }, { "epoch": 2.5485289963275353, "grad_norm": 0.33695464746068704, "learning_rate": 9.070254852788965e-07, "loss": 0.2889, "step": 7895 }, { "epoch": 2.548851850357157, "grad_norm": 0.45051890995299104, "learning_rate": 9.05753287995601e-07, "loss": 0.294, "step": 7896 }, { "epoch": 2.549174704386779, "grad_norm": 0.33538906677082975, "learning_rate": 9.04481926194285e-07, "loss": 0.2662, "step": 7897 }, { "epoch": 2.549497558416401, "grad_norm": 0.3856362199810011, "learning_rate": 9.032114000360253e-07, "loss": 0.2304, "step": 7898 }, { "epoch": 2.5498204124460226, "grad_norm": 0.38655720908386043, "learning_rate": 9.019417096817975e-07, "loss": 0.247, "step": 7899 }, { "epoch": 2.5501432664756445, "grad_norm": 0.41706738646227487, "learning_rate": 9.006728552924679e-07, "loss": 0.1732, "step": 7900 }, { "epoch": 2.5504661205052663, "grad_norm": 0.4798756639966013, "learning_rate": 8.99404837028801e-07, "loss": 0.2404, "step": 7901 }, { "epoch": 2.550788974534888, "grad_norm": 0.38060634972653545, "learning_rate": 8.981376550514511e-07, "loss": 0.2643, "step": 7902 }, { "epoch": 2.55111182856451, "grad_norm": 0.4083973035116797, "learning_rate": 8.968713095209697e-07, "loss": 0.2481, "step": 7903 }, { "epoch": 2.551434682594132, "grad_norm": 0.4199396484767685, "learning_rate": 8.95605800597799e-07, "loss": 0.2273, "step": 7904 }, { "epoch": 2.5517575366237537, "grad_norm": 0.4209576439736934, "learning_rate": 8.943411284422778e-07, "loss": 0.3074, "step": 7905 }, { "epoch": 2.552080390653376, "grad_norm": 0.3777372907970976, "learning_rate": 8.930772932146419e-07, "loss": 0.2177, "step": 7906 }, { "epoch": 2.552403244682998, "grad_norm": 0.3346180085976405, "learning_rate": 8.918142950750124e-07, "loss": 0.2346, "step": 7907 }, { "epoch": 2.5527260987126197, "grad_norm": 0.4089473638312951, "learning_rate": 8.905521341834122e-07, "loss": 0.2326, "step": 7908 }, { "epoch": 2.5530489527422415, "grad_norm": 0.33815086601376576, "learning_rate": 8.89290810699754e-07, "loss": 0.302, "step": 7909 }, { "epoch": 2.5533718067718634, "grad_norm": 0.36562412576850056, "learning_rate": 8.880303247838486e-07, "loss": 0.281, "step": 7910 }, { "epoch": 2.553694660801485, "grad_norm": 0.3640190896674103, "learning_rate": 8.867706765953948e-07, "loss": 0.2839, "step": 7911 }, { "epoch": 2.554017514831107, "grad_norm": 0.366704584759916, "learning_rate": 8.855118662939899e-07, "loss": 0.271, "step": 7912 }, { "epoch": 2.554340368860729, "grad_norm": 0.38415384054049184, "learning_rate": 8.84253894039122e-07, "loss": 0.266, "step": 7913 }, { "epoch": 2.5546632228903507, "grad_norm": 0.380408157421671, "learning_rate": 8.829967599901764e-07, "loss": 0.2727, "step": 7914 }, { "epoch": 2.5549860769199726, "grad_norm": 0.40765546954718196, "learning_rate": 8.81740464306428e-07, "loss": 0.272, "step": 7915 }, { "epoch": 2.5553089309495944, "grad_norm": 0.5456194728278464, "learning_rate": 8.804850071470505e-07, "loss": 0.2633, "step": 7916 }, { "epoch": 2.5556317849792163, "grad_norm": 0.3386848868191626, "learning_rate": 8.792303886711056e-07, "loss": 0.2305, "step": 7917 }, { "epoch": 2.555954639008838, "grad_norm": 0.4348497330165023, "learning_rate": 8.779766090375538e-07, "loss": 0.2659, "step": 7918 }, { "epoch": 2.55627749303846, "grad_norm": 0.3806042241724317, "learning_rate": 8.767236684052454e-07, "loss": 0.2514, "step": 7919 }, { "epoch": 2.556600347068082, "grad_norm": 0.36336734621621014, "learning_rate": 8.754715669329294e-07, "loss": 0.2474, "step": 7920 }, { "epoch": 2.5569232010977037, "grad_norm": 0.41203174252685576, "learning_rate": 8.742203047792413e-07, "loss": 0.247, "step": 7921 }, { "epoch": 2.5572460551273255, "grad_norm": 0.4386427164584035, "learning_rate": 8.729698821027168e-07, "loss": 0.3014, "step": 7922 }, { "epoch": 2.5575689091569473, "grad_norm": 0.34614936617676495, "learning_rate": 8.717202990617817e-07, "loss": 0.2475, "step": 7923 }, { "epoch": 2.557891763186569, "grad_norm": 0.34464142187209135, "learning_rate": 8.704715558147586e-07, "loss": 0.2593, "step": 7924 }, { "epoch": 2.558214617216191, "grad_norm": 0.3734461469378693, "learning_rate": 8.692236525198591e-07, "loss": 0.2676, "step": 7925 }, { "epoch": 2.558537471245813, "grad_norm": 0.5928823107217883, "learning_rate": 8.679765893351909e-07, "loss": 0.2287, "step": 7926 }, { "epoch": 2.5588603252754347, "grad_norm": 0.36234313245410904, "learning_rate": 8.667303664187579e-07, "loss": 0.2179, "step": 7927 }, { "epoch": 2.5591831793050566, "grad_norm": 0.35060929188369594, "learning_rate": 8.654849839284517e-07, "loss": 0.2235, "step": 7928 }, { "epoch": 2.5595060333346784, "grad_norm": 0.3957031017994903, "learning_rate": 8.642404420220632e-07, "loss": 0.2703, "step": 7929 }, { "epoch": 2.5598288873643003, "grad_norm": 0.42612808645053374, "learning_rate": 8.629967408572709e-07, "loss": 0.2904, "step": 7930 }, { "epoch": 2.5601517413939225, "grad_norm": 0.3747823002276042, "learning_rate": 8.617538805916531e-07, "loss": 0.2635, "step": 7931 }, { "epoch": 2.5604745954235444, "grad_norm": 0.38342822048548486, "learning_rate": 8.605118613826762e-07, "loss": 0.2562, "step": 7932 }, { "epoch": 2.5607974494531662, "grad_norm": 0.37961758194708706, "learning_rate": 8.592706833877046e-07, "loss": 0.301, "step": 7933 }, { "epoch": 2.561120303482788, "grad_norm": 0.43804717047333874, "learning_rate": 8.580303467639914e-07, "loss": 0.2228, "step": 7934 }, { "epoch": 2.56144315751241, "grad_norm": 0.40076059925623125, "learning_rate": 8.567908516686876e-07, "loss": 0.2392, "step": 7935 }, { "epoch": 2.5617660115420318, "grad_norm": 0.3550752806860436, "learning_rate": 8.555521982588335e-07, "loss": 0.2172, "step": 7936 }, { "epoch": 2.5620888655716536, "grad_norm": 0.3285701958141742, "learning_rate": 8.543143866913661e-07, "loss": 0.25, "step": 7937 }, { "epoch": 2.5624117196012755, "grad_norm": 0.38269005680840024, "learning_rate": 8.53077417123114e-07, "loss": 0.2437, "step": 7938 }, { "epoch": 2.5627345736308973, "grad_norm": 0.4367381979992385, "learning_rate": 8.518412897108019e-07, "loss": 0.2426, "step": 7939 }, { "epoch": 2.563057427660519, "grad_norm": 0.3722485966649466, "learning_rate": 8.506060046110411e-07, "loss": 0.2597, "step": 7940 }, { "epoch": 2.563380281690141, "grad_norm": 0.34649047838206365, "learning_rate": 8.49371561980345e-07, "loss": 0.2315, "step": 7941 }, { "epoch": 2.563703135719763, "grad_norm": 0.47572495700756195, "learning_rate": 8.481379619751115e-07, "loss": 0.2412, "step": 7942 }, { "epoch": 2.5640259897493847, "grad_norm": 0.3728754064789954, "learning_rate": 8.469052047516391e-07, "loss": 0.2662, "step": 7943 }, { "epoch": 2.5643488437790065, "grad_norm": 0.3454263316872869, "learning_rate": 8.456732904661144e-07, "loss": 0.1907, "step": 7944 }, { "epoch": 2.5646716978086284, "grad_norm": 0.4736308694156165, "learning_rate": 8.444422192746201e-07, "loss": 0.2701, "step": 7945 }, { "epoch": 2.56499455183825, "grad_norm": 0.3875897076242761, "learning_rate": 8.432119913331332e-07, "loss": 0.2446, "step": 7946 }, { "epoch": 2.565317405867872, "grad_norm": 0.4712869074359741, "learning_rate": 8.419826067975181e-07, "loss": 0.2639, "step": 7947 }, { "epoch": 2.565640259897494, "grad_norm": 0.34782404911731135, "learning_rate": 8.407540658235402e-07, "loss": 0.2507, "step": 7948 }, { "epoch": 2.5659631139271157, "grad_norm": 0.3555214390452739, "learning_rate": 8.3952636856685e-07, "loss": 0.2735, "step": 7949 }, { "epoch": 2.5662859679567376, "grad_norm": 0.32261188222273696, "learning_rate": 8.382995151829981e-07, "loss": 0.2584, "step": 7950 }, { "epoch": 2.5666088219863594, "grad_norm": 0.35440060073709917, "learning_rate": 8.370735058274234e-07, "loss": 0.2818, "step": 7951 }, { "epoch": 2.5669316760159813, "grad_norm": 0.3157303987495713, "learning_rate": 8.358483406554609e-07, "loss": 0.2499, "step": 7952 }, { "epoch": 2.567254530045603, "grad_norm": 0.3740905051330326, "learning_rate": 8.346240198223349e-07, "loss": 0.2536, "step": 7953 }, { "epoch": 2.567577384075225, "grad_norm": 0.36041715347603553, "learning_rate": 8.334005434831668e-07, "loss": 0.2853, "step": 7954 }, { "epoch": 2.567900238104847, "grad_norm": 0.37938489243602513, "learning_rate": 8.321779117929696e-07, "loss": 0.2631, "step": 7955 }, { "epoch": 2.5682230921344686, "grad_norm": 0.5037610121425764, "learning_rate": 8.309561249066499e-07, "loss": 0.2335, "step": 7956 }, { "epoch": 2.5685459461640905, "grad_norm": 0.363484225848648, "learning_rate": 8.297351829790043e-07, "loss": 0.2322, "step": 7957 }, { "epoch": 2.5688688001937123, "grad_norm": 0.3692801199688773, "learning_rate": 8.28515086164727e-07, "loss": 0.2349, "step": 7958 }, { "epoch": 2.569191654223334, "grad_norm": 0.41521565184608544, "learning_rate": 8.272958346183995e-07, "loss": 0.2581, "step": 7959 }, { "epoch": 2.569514508252956, "grad_norm": 0.4048280958883021, "learning_rate": 8.260774284945025e-07, "loss": 0.2511, "step": 7960 }, { "epoch": 2.569837362282578, "grad_norm": 0.4198726338207331, "learning_rate": 8.24859867947404e-07, "loss": 0.247, "step": 7961 }, { "epoch": 2.5701602163121997, "grad_norm": 0.3984966563277021, "learning_rate": 8.236431531313665e-07, "loss": 0.2691, "step": 7962 }, { "epoch": 2.5704830703418216, "grad_norm": 0.34236920019213946, "learning_rate": 8.224272842005501e-07, "loss": 0.2637, "step": 7963 }, { "epoch": 2.5708059243714434, "grad_norm": 0.4296547808862839, "learning_rate": 8.212122613089998e-07, "loss": 0.2105, "step": 7964 }, { "epoch": 2.5711287784010652, "grad_norm": 0.35187144873420906, "learning_rate": 8.1999808461066e-07, "loss": 0.2318, "step": 7965 }, { "epoch": 2.571451632430687, "grad_norm": 0.38757993937583085, "learning_rate": 8.187847542593619e-07, "loss": 0.2244, "step": 7966 }, { "epoch": 2.571774486460309, "grad_norm": 0.3587773187823166, "learning_rate": 8.175722704088373e-07, "loss": 0.2898, "step": 7967 }, { "epoch": 2.5720973404899308, "grad_norm": 0.33391496880885546, "learning_rate": 8.163606332127019e-07, "loss": 0.206, "step": 7968 }, { "epoch": 2.5724201945195526, "grad_norm": 0.39439272613051946, "learning_rate": 8.151498428244702e-07, "loss": 0.2364, "step": 7969 }, { "epoch": 2.5727430485491745, "grad_norm": 0.33425823728919446, "learning_rate": 8.139398993975481e-07, "loss": 0.238, "step": 7970 }, { "epoch": 2.5730659025787963, "grad_norm": 0.34985651027638026, "learning_rate": 8.127308030852343e-07, "loss": 0.2228, "step": 7971 }, { "epoch": 2.573388756608418, "grad_norm": 0.353778102628498, "learning_rate": 8.115225540407181e-07, "loss": 0.2799, "step": 7972 }, { "epoch": 2.5737116106380404, "grad_norm": 0.4430810365250751, "learning_rate": 8.103151524170846e-07, "loss": 0.2777, "step": 7973 }, { "epoch": 2.5740344646676623, "grad_norm": 0.36043583107713933, "learning_rate": 8.09108598367308e-07, "loss": 0.3085, "step": 7974 }, { "epoch": 2.574357318697284, "grad_norm": 0.38413683477778066, "learning_rate": 8.079028920442594e-07, "loss": 0.2394, "step": 7975 }, { "epoch": 2.574680172726906, "grad_norm": 0.42987853836459, "learning_rate": 8.066980336006973e-07, "loss": 0.1795, "step": 7976 }, { "epoch": 2.575003026756528, "grad_norm": 0.4630903675029377, "learning_rate": 8.05494023189279e-07, "loss": 0.2551, "step": 7977 }, { "epoch": 2.5753258807861497, "grad_norm": 0.40240591826040073, "learning_rate": 8.042908609625474e-07, "loss": 0.2869, "step": 7978 }, { "epoch": 2.5756487348157715, "grad_norm": 0.3445033916140805, "learning_rate": 8.030885470729446e-07, "loss": 0.209, "step": 7979 }, { "epoch": 2.5759715888453933, "grad_norm": 0.34003654935531386, "learning_rate": 8.018870816727997e-07, "loss": 0.2563, "step": 7980 }, { "epoch": 2.576294442875015, "grad_norm": 0.35632535136118165, "learning_rate": 8.006864649143366e-07, "loss": 0.2657, "step": 7981 }, { "epoch": 2.576617296904637, "grad_norm": 0.38334709245140314, "learning_rate": 7.994866969496748e-07, "loss": 0.3028, "step": 7982 }, { "epoch": 2.576940150934259, "grad_norm": 0.37051648830327766, "learning_rate": 7.982877779308207e-07, "loss": 0.2092, "step": 7983 }, { "epoch": 2.5772630049638807, "grad_norm": 0.39304631194173967, "learning_rate": 7.970897080096753e-07, "loss": 0.233, "step": 7984 }, { "epoch": 2.5775858589935026, "grad_norm": 0.34380158562232926, "learning_rate": 7.958924873380349e-07, "loss": 0.2498, "step": 7985 }, { "epoch": 2.5779087130231244, "grad_norm": 0.3501431660617096, "learning_rate": 7.94696116067583e-07, "loss": 0.2579, "step": 7986 }, { "epoch": 2.5782315670527463, "grad_norm": 0.4030813263065681, "learning_rate": 7.935005943499001e-07, "loss": 0.2401, "step": 7987 }, { "epoch": 2.578554421082368, "grad_norm": 0.33675892039060745, "learning_rate": 7.923059223364576e-07, "loss": 0.2401, "step": 7988 }, { "epoch": 2.57887727511199, "grad_norm": 0.376836325609142, "learning_rate": 7.911121001786159e-07, "loss": 0.286, "step": 7989 }, { "epoch": 2.579200129141612, "grad_norm": 0.41643729170165833, "learning_rate": 7.899191280276341e-07, "loss": 0.2236, "step": 7990 }, { "epoch": 2.5795229831712336, "grad_norm": 0.36801516813384894, "learning_rate": 7.887270060346572e-07, "loss": 0.263, "step": 7991 }, { "epoch": 2.5798458372008555, "grad_norm": 0.4127078148588038, "learning_rate": 7.875357343507276e-07, "loss": 0.2561, "step": 7992 }, { "epoch": 2.5801686912304773, "grad_norm": 0.3211639704868279, "learning_rate": 7.863453131267754e-07, "loss": 0.2587, "step": 7993 }, { "epoch": 2.580491545260099, "grad_norm": 0.3895889914071279, "learning_rate": 7.851557425136283e-07, "loss": 0.2797, "step": 7994 }, { "epoch": 2.580814399289721, "grad_norm": 0.3886064108496671, "learning_rate": 7.839670226620008e-07, "loss": 0.2023, "step": 7995 }, { "epoch": 2.581137253319343, "grad_norm": 0.4088741316466049, "learning_rate": 7.82779153722504e-07, "loss": 0.2699, "step": 7996 }, { "epoch": 2.5814601073489647, "grad_norm": 0.34670090304552104, "learning_rate": 7.815921358456368e-07, "loss": 0.2752, "step": 7997 }, { "epoch": 2.581782961378587, "grad_norm": 0.35184964498998156, "learning_rate": 7.804059691817955e-07, "loss": 0.2471, "step": 7998 }, { "epoch": 2.582105815408209, "grad_norm": 0.4104100195763944, "learning_rate": 7.79220653881264e-07, "loss": 0.2996, "step": 7999 }, { "epoch": 2.5824286694378307, "grad_norm": 0.3901125820702896, "learning_rate": 7.780361900942198e-07, "loss": 0.2873, "step": 8000 }, { "epoch": 2.5827515234674525, "grad_norm": 0.328609555604106, "learning_rate": 7.768525779707336e-07, "loss": 0.2625, "step": 8001 }, { "epoch": 2.5830743774970744, "grad_norm": 0.3147405541356099, "learning_rate": 7.756698176607672e-07, "loss": 0.2205, "step": 8002 }, { "epoch": 2.583397231526696, "grad_norm": 0.30227376976422526, "learning_rate": 7.744879093141772e-07, "loss": 0.2468, "step": 8003 }, { "epoch": 2.583720085556318, "grad_norm": 0.399451668022018, "learning_rate": 7.733068530807056e-07, "loss": 0.22, "step": 8004 }, { "epoch": 2.58404293958594, "grad_norm": 0.3184352806046786, "learning_rate": 7.721266491099935e-07, "loss": 0.2363, "step": 8005 }, { "epoch": 2.5843657936155617, "grad_norm": 0.3687237460863607, "learning_rate": 7.709472975515688e-07, "loss": 0.3156, "step": 8006 }, { "epoch": 2.5846886476451836, "grad_norm": 0.3618318510366374, "learning_rate": 7.697687985548571e-07, "loss": 0.2366, "step": 8007 }, { "epoch": 2.5850115016748054, "grad_norm": 0.40650206758698726, "learning_rate": 7.685911522691688e-07, "loss": 0.2618, "step": 8008 }, { "epoch": 2.5853343557044273, "grad_norm": 0.42936145106299106, "learning_rate": 7.674143588437129e-07, "loss": 0.2228, "step": 8009 }, { "epoch": 2.585657209734049, "grad_norm": 0.36815792861777663, "learning_rate": 7.662384184275861e-07, "loss": 0.216, "step": 8010 }, { "epoch": 2.585980063763671, "grad_norm": 0.44627443782539944, "learning_rate": 7.650633311697791e-07, "loss": 0.258, "step": 8011 }, { "epoch": 2.586302917793293, "grad_norm": 0.34717776440074977, "learning_rate": 7.638890972191719e-07, "loss": 0.2666, "step": 8012 }, { "epoch": 2.5866257718229146, "grad_norm": 0.3345066219107964, "learning_rate": 7.627157167245424e-07, "loss": 0.2535, "step": 8013 }, { "epoch": 2.5869486258525365, "grad_norm": 0.3150490414838378, "learning_rate": 7.615431898345515e-07, "loss": 0.2618, "step": 8014 }, { "epoch": 2.5872714798821583, "grad_norm": 0.3932902731989576, "learning_rate": 7.603715166977596e-07, "loss": 0.327, "step": 8015 }, { "epoch": 2.58759433391178, "grad_norm": 0.373879366904817, "learning_rate": 7.592006974626145e-07, "loss": 0.29, "step": 8016 }, { "epoch": 2.587917187941402, "grad_norm": 0.3393002321484704, "learning_rate": 7.58030732277461e-07, "loss": 0.2606, "step": 8017 }, { "epoch": 2.588240041971024, "grad_norm": 0.4191819863637784, "learning_rate": 7.568616212905271e-07, "loss": 0.2056, "step": 8018 }, { "epoch": 2.5885628960006457, "grad_norm": 0.3536738297380442, "learning_rate": 7.556933646499412e-07, "loss": 0.2623, "step": 8019 }, { "epoch": 2.5888857500302676, "grad_norm": 0.4343336663958829, "learning_rate": 7.545259625037187e-07, "loss": 0.2459, "step": 8020 }, { "epoch": 2.5892086040598894, "grad_norm": 0.4261806959368437, "learning_rate": 7.533594149997667e-07, "loss": 0.2558, "step": 8021 }, { "epoch": 2.5895314580895112, "grad_norm": 0.3821185480548583, "learning_rate": 7.521937222858873e-07, "loss": 0.2673, "step": 8022 }, { "epoch": 2.589854312119133, "grad_norm": 0.36716347411123795, "learning_rate": 7.510288845097704e-07, "loss": 0.2602, "step": 8023 }, { "epoch": 2.590177166148755, "grad_norm": 0.39039069414055455, "learning_rate": 7.49864901819001e-07, "loss": 0.2244, "step": 8024 }, { "epoch": 2.5905000201783768, "grad_norm": 0.34261628534222577, "learning_rate": 7.487017743610511e-07, "loss": 0.2537, "step": 8025 }, { "epoch": 2.5908228742079986, "grad_norm": 0.37292095589228247, "learning_rate": 7.475395022832912e-07, "loss": 0.2867, "step": 8026 }, { "epoch": 2.5911457282376205, "grad_norm": 0.3465120488457769, "learning_rate": 7.46378085732976e-07, "loss": 0.2394, "step": 8027 }, { "epoch": 2.5914685822672423, "grad_norm": 0.3875829168014664, "learning_rate": 7.452175248572588e-07, "loss": 0.2729, "step": 8028 }, { "epoch": 2.591791436296864, "grad_norm": 0.3995718881795687, "learning_rate": 7.440578198031775e-07, "loss": 0.2428, "step": 8029 }, { "epoch": 2.592114290326486, "grad_norm": 0.4871917782147583, "learning_rate": 7.42898970717669e-07, "loss": 0.2103, "step": 8030 }, { "epoch": 2.592437144356108, "grad_norm": 0.4316560633062225, "learning_rate": 7.417409777475539e-07, "loss": 0.2613, "step": 8031 }, { "epoch": 2.5927599983857297, "grad_norm": 0.48585875991929167, "learning_rate": 7.405838410395502e-07, "loss": 0.2311, "step": 8032 }, { "epoch": 2.5930828524153515, "grad_norm": 0.42433752320164614, "learning_rate": 7.394275607402654e-07, "loss": 0.2967, "step": 8033 }, { "epoch": 2.5934057064449734, "grad_norm": 0.3358892170588928, "learning_rate": 7.382721369962006e-07, "loss": 0.2561, "step": 8034 }, { "epoch": 2.593728560474595, "grad_norm": 0.3807977912116129, "learning_rate": 7.37117569953743e-07, "loss": 0.2535, "step": 8035 }, { "epoch": 2.594051414504217, "grad_norm": 0.3999117929729431, "learning_rate": 7.359638597591764e-07, "loss": 0.3023, "step": 8036 }, { "epoch": 2.594374268533839, "grad_norm": 0.32265298204243553, "learning_rate": 7.348110065586755e-07, "loss": 0.2192, "step": 8037 }, { "epoch": 2.5946971225634607, "grad_norm": 0.3688244632672477, "learning_rate": 7.336590104983018e-07, "loss": 0.2661, "step": 8038 }, { "epoch": 2.5950199765930826, "grad_norm": 0.37371767566609654, "learning_rate": 7.325078717240161e-07, "loss": 0.2677, "step": 8039 }, { "epoch": 2.5953428306227044, "grad_norm": 0.393774348952902, "learning_rate": 7.313575903816616e-07, "loss": 0.2388, "step": 8040 }, { "epoch": 2.5956656846523267, "grad_norm": 0.4599768211622512, "learning_rate": 7.30208166616981e-07, "loss": 0.2181, "step": 8041 }, { "epoch": 2.5959885386819486, "grad_norm": 0.3500461655819021, "learning_rate": 7.29059600575602e-07, "loss": 0.2578, "step": 8042 }, { "epoch": 2.5963113927115704, "grad_norm": 0.4396228884162167, "learning_rate": 7.27911892403049e-07, "loss": 0.2754, "step": 8043 }, { "epoch": 2.5966342467411923, "grad_norm": 0.39805913738798915, "learning_rate": 7.267650422447314e-07, "loss": 0.2929, "step": 8044 }, { "epoch": 2.596957100770814, "grad_norm": 0.38427816920000507, "learning_rate": 7.256190502459572e-07, "loss": 0.2726, "step": 8045 }, { "epoch": 2.597279954800436, "grad_norm": 0.40119416075186215, "learning_rate": 7.244739165519196e-07, "loss": 0.2783, "step": 8046 }, { "epoch": 2.597602808830058, "grad_norm": 0.373933829153597, "learning_rate": 7.233296413077057e-07, "loss": 0.2582, "step": 8047 }, { "epoch": 2.5979256628596796, "grad_norm": 0.43387112242254877, "learning_rate": 7.221862246582945e-07, "loss": 0.293, "step": 8048 }, { "epoch": 2.5982485168893015, "grad_norm": 0.36279902994179053, "learning_rate": 7.210436667485568e-07, "loss": 0.2232, "step": 8049 }, { "epoch": 2.5985713709189233, "grad_norm": 0.425043904484997, "learning_rate": 7.199019677232494e-07, "loss": 0.2025, "step": 8050 }, { "epoch": 2.598894224948545, "grad_norm": 0.3507071709205676, "learning_rate": 7.187611277270272e-07, "loss": 0.2657, "step": 8051 }, { "epoch": 2.599217078978167, "grad_norm": 0.3342200266129634, "learning_rate": 7.176211469044308e-07, "loss": 0.2449, "step": 8052 }, { "epoch": 2.599539933007789, "grad_norm": 0.4357244980187175, "learning_rate": 7.164820253998969e-07, "loss": 0.2259, "step": 8053 }, { "epoch": 2.5998627870374107, "grad_norm": 0.4052755562752081, "learning_rate": 7.153437633577467e-07, "loss": 0.2342, "step": 8054 }, { "epoch": 2.6001856410670325, "grad_norm": 0.49110496456095293, "learning_rate": 7.142063609221991e-07, "loss": 0.2583, "step": 8055 }, { "epoch": 2.6005084950966544, "grad_norm": 0.3894666996139149, "learning_rate": 7.130698182373621e-07, "loss": 0.2532, "step": 8056 }, { "epoch": 2.6008313491262762, "grad_norm": 0.38331583080213955, "learning_rate": 7.119341354472322e-07, "loss": 0.268, "step": 8057 }, { "epoch": 2.601154203155898, "grad_norm": 0.37531645951962544, "learning_rate": 7.107993126957002e-07, "loss": 0.2262, "step": 8058 }, { "epoch": 2.60147705718552, "grad_norm": 0.39576201025004004, "learning_rate": 7.096653501265452e-07, "loss": 0.281, "step": 8059 }, { "epoch": 2.6017999112151418, "grad_norm": 0.3327769125359889, "learning_rate": 7.085322478834405e-07, "loss": 0.2875, "step": 8060 }, { "epoch": 2.6021227652447636, "grad_norm": 0.41050076843999495, "learning_rate": 7.074000061099465e-07, "loss": 0.2765, "step": 8061 }, { "epoch": 2.6024456192743854, "grad_norm": 0.37623377594653196, "learning_rate": 7.062686249495181e-07, "loss": 0.2307, "step": 8062 }, { "epoch": 2.6027684733040073, "grad_norm": 0.36565700926882816, "learning_rate": 7.051381045455002e-07, "loss": 0.2292, "step": 8063 }, { "epoch": 2.603091327333629, "grad_norm": 0.35122308758260923, "learning_rate": 7.040084450411254e-07, "loss": 0.2674, "step": 8064 }, { "epoch": 2.603414181363251, "grad_norm": 0.4301979403080589, "learning_rate": 7.028796465795228e-07, "loss": 0.3015, "step": 8065 }, { "epoch": 2.6037370353928733, "grad_norm": 0.3674196106257125, "learning_rate": 7.017517093037104e-07, "loss": 0.234, "step": 8066 }, { "epoch": 2.604059889422495, "grad_norm": 0.39970545397136215, "learning_rate": 7.00624633356593e-07, "loss": 0.2709, "step": 8067 }, { "epoch": 2.604382743452117, "grad_norm": 0.3471545610270284, "learning_rate": 6.99498418880973e-07, "loss": 0.2416, "step": 8068 }, { "epoch": 2.604705597481739, "grad_norm": 0.3690684067710651, "learning_rate": 6.983730660195365e-07, "loss": 0.2096, "step": 8069 }, { "epoch": 2.6050284515113606, "grad_norm": 0.360874005529757, "learning_rate": 6.972485749148675e-07, "loss": 0.2879, "step": 8070 }, { "epoch": 2.6053513055409825, "grad_norm": 0.3607958495957252, "learning_rate": 6.961249457094354e-07, "loss": 0.2109, "step": 8071 }, { "epoch": 2.6056741595706043, "grad_norm": 0.39373425454463107, "learning_rate": 6.95002178545604e-07, "loss": 0.2568, "step": 8072 }, { "epoch": 2.605997013600226, "grad_norm": 0.3810141510547404, "learning_rate": 6.938802735656244e-07, "loss": 0.2657, "step": 8073 }, { "epoch": 2.606319867629848, "grad_norm": 0.483566032208028, "learning_rate": 6.927592309116412e-07, "loss": 0.2488, "step": 8074 }, { "epoch": 2.60664272165947, "grad_norm": 0.40595209221186296, "learning_rate": 6.916390507256909e-07, "loss": 0.2494, "step": 8075 }, { "epoch": 2.6069655756890917, "grad_norm": 0.4287466663502798, "learning_rate": 6.905197331496957e-07, "loss": 0.2472, "step": 8076 }, { "epoch": 2.6072884297187136, "grad_norm": 0.4100566809908829, "learning_rate": 6.894012783254746e-07, "loss": 0.2263, "step": 8077 }, { "epoch": 2.6076112837483354, "grad_norm": 0.3450970168248917, "learning_rate": 6.882836863947312e-07, "loss": 0.2781, "step": 8078 }, { "epoch": 2.6079341377779572, "grad_norm": 0.36711373530355695, "learning_rate": 6.871669574990652e-07, "loss": 0.2892, "step": 8079 }, { "epoch": 2.608256991807579, "grad_norm": 0.34994688951579034, "learning_rate": 6.860510917799629e-07, "loss": 0.2958, "step": 8080 }, { "epoch": 2.608579845837201, "grad_norm": 0.40048819464138646, "learning_rate": 6.849360893788059e-07, "loss": 0.2641, "step": 8081 }, { "epoch": 2.6089026998668228, "grad_norm": 0.42127195811015117, "learning_rate": 6.838219504368598e-07, "loss": 0.2247, "step": 8082 }, { "epoch": 2.6092255538964446, "grad_norm": 0.3637565380409361, "learning_rate": 6.827086750952874e-07, "loss": 0.2585, "step": 8083 }, { "epoch": 2.6095484079260665, "grad_norm": 0.3739853635948652, "learning_rate": 6.815962634951367e-07, "loss": 0.259, "step": 8084 }, { "epoch": 2.6098712619556883, "grad_norm": 0.35204623459103423, "learning_rate": 6.804847157773517e-07, "loss": 0.262, "step": 8085 }, { "epoch": 2.61019411598531, "grad_norm": 0.3574442289172755, "learning_rate": 6.793740320827605e-07, "loss": 0.2963, "step": 8086 }, { "epoch": 2.610516970014932, "grad_norm": 0.4180183675781611, "learning_rate": 6.782642125520879e-07, "loss": 0.2263, "step": 8087 }, { "epoch": 2.610839824044554, "grad_norm": 0.3818145083951875, "learning_rate": 6.771552573259443e-07, "loss": 0.2788, "step": 8088 }, { "epoch": 2.6111626780741757, "grad_norm": 0.3696298490257885, "learning_rate": 6.760471665448354e-07, "loss": 0.2403, "step": 8089 }, { "epoch": 2.6114855321037975, "grad_norm": 0.34833582629635307, "learning_rate": 6.749399403491521e-07, "loss": 0.3563, "step": 8090 }, { "epoch": 2.6118083861334194, "grad_norm": 0.4108809724596198, "learning_rate": 6.738335788791799e-07, "loss": 0.2795, "step": 8091 }, { "epoch": 2.612131240163041, "grad_norm": 0.3555955216034746, "learning_rate": 6.727280822750942e-07, "loss": 0.2556, "step": 8092 }, { "epoch": 2.612454094192663, "grad_norm": 0.3672407957678763, "learning_rate": 6.71623450676957e-07, "loss": 0.2433, "step": 8093 }, { "epoch": 2.612776948222285, "grad_norm": 0.4714679715398263, "learning_rate": 6.705196842247263e-07, "loss": 0.2376, "step": 8094 }, { "epoch": 2.6130998022519067, "grad_norm": 0.41066525074171817, "learning_rate": 6.694167830582471e-07, "loss": 0.2819, "step": 8095 }, { "epoch": 2.6134226562815286, "grad_norm": 0.3697325536818692, "learning_rate": 6.683147473172549e-07, "loss": 0.2457, "step": 8096 }, { "epoch": 2.6137455103111504, "grad_norm": 0.39613346760905893, "learning_rate": 6.672135771413757e-07, "loss": 0.2653, "step": 8097 }, { "epoch": 2.6140683643407723, "grad_norm": 0.40354368466593965, "learning_rate": 6.661132726701286e-07, "loss": 0.289, "step": 8098 }, { "epoch": 2.614391218370394, "grad_norm": 0.3823435982218428, "learning_rate": 6.650138340429171e-07, "loss": 0.2795, "step": 8099 }, { "epoch": 2.614714072400016, "grad_norm": 0.4341768909551324, "learning_rate": 6.639152613990424e-07, "loss": 0.2233, "step": 8100 }, { "epoch": 2.615036926429638, "grad_norm": 0.3659979363844553, "learning_rate": 6.628175548776888e-07, "loss": 0.2409, "step": 8101 }, { "epoch": 2.6153597804592597, "grad_norm": 0.35561839292795294, "learning_rate": 6.617207146179377e-07, "loss": 0.212, "step": 8102 }, { "epoch": 2.6156826344888815, "grad_norm": 0.42799006783038385, "learning_rate": 6.606247407587529e-07, "loss": 0.2845, "step": 8103 }, { "epoch": 2.6160054885185033, "grad_norm": 0.3765952367009152, "learning_rate": 6.595296334389967e-07, "loss": 0.3054, "step": 8104 }, { "epoch": 2.616328342548125, "grad_norm": 0.36940421734949114, "learning_rate": 6.584353927974146e-07, "loss": 0.2779, "step": 8105 }, { "epoch": 2.616651196577747, "grad_norm": 0.3644295392504743, "learning_rate": 6.573420189726484e-07, "loss": 0.3072, "step": 8106 }, { "epoch": 2.616974050607369, "grad_norm": 0.42425957472084724, "learning_rate": 6.562495121032239e-07, "loss": 0.2871, "step": 8107 }, { "epoch": 2.617296904636991, "grad_norm": 0.38878315663887536, "learning_rate": 6.55157872327562e-07, "loss": 0.2757, "step": 8108 }, { "epoch": 2.617619758666613, "grad_norm": 0.3405666462891957, "learning_rate": 6.540670997839729e-07, "loss": 0.2503, "step": 8109 }, { "epoch": 2.617942612696235, "grad_norm": 0.3759063116676872, "learning_rate": 6.529771946106536e-07, "loss": 0.2307, "step": 8110 }, { "epoch": 2.6182654667258567, "grad_norm": 0.42565498591912, "learning_rate": 6.518881569456945e-07, "loss": 0.2752, "step": 8111 }, { "epoch": 2.6185883207554785, "grad_norm": 0.39371163970486317, "learning_rate": 6.507999869270759e-07, "loss": 0.2173, "step": 8112 }, { "epoch": 2.6189111747851004, "grad_norm": 0.38852396951954415, "learning_rate": 6.497126846926685e-07, "loss": 0.2716, "step": 8113 }, { "epoch": 2.6192340288147222, "grad_norm": 0.5053839214445683, "learning_rate": 6.486262503802296e-07, "loss": 0.3368, "step": 8114 }, { "epoch": 2.619556882844344, "grad_norm": 0.3446550405376468, "learning_rate": 6.475406841274106e-07, "loss": 0.2506, "step": 8115 }, { "epoch": 2.619879736873966, "grad_norm": 0.44667158019272335, "learning_rate": 6.464559860717492e-07, "loss": 0.2772, "step": 8116 }, { "epoch": 2.6202025909035878, "grad_norm": 0.4500581342193888, "learning_rate": 6.453721563506779e-07, "loss": 0.2626, "step": 8117 }, { "epoch": 2.6205254449332096, "grad_norm": 0.41231462453689766, "learning_rate": 6.442891951015134e-07, "loss": 0.237, "step": 8118 }, { "epoch": 2.6208482989628314, "grad_norm": 0.4038930497140001, "learning_rate": 6.432071024614686e-07, "loss": 0.304, "step": 8119 }, { "epoch": 2.6211711529924533, "grad_norm": 0.35370405797010585, "learning_rate": 6.421258785676404e-07, "loss": 0.2501, "step": 8120 }, { "epoch": 2.621494007022075, "grad_norm": 0.33953947978138194, "learning_rate": 6.4104552355702e-07, "loss": 0.2393, "step": 8121 }, { "epoch": 2.621816861051697, "grad_norm": 0.37465219698477326, "learning_rate": 6.399660375664856e-07, "loss": 0.2781, "step": 8122 }, { "epoch": 2.622139715081319, "grad_norm": 0.3834911180105875, "learning_rate": 6.388874207328093e-07, "loss": 0.3047, "step": 8123 }, { "epoch": 2.6224625691109407, "grad_norm": 0.33912980298147155, "learning_rate": 6.378096731926469e-07, "loss": 0.2697, "step": 8124 }, { "epoch": 2.6227854231405625, "grad_norm": 0.3778359946651467, "learning_rate": 6.367327950825491e-07, "loss": 0.2573, "step": 8125 }, { "epoch": 2.6231082771701844, "grad_norm": 0.3831592202468397, "learning_rate": 6.356567865389543e-07, "loss": 0.2841, "step": 8126 }, { "epoch": 2.623431131199806, "grad_norm": 0.380636720257767, "learning_rate": 6.345816476981933e-07, "loss": 0.24, "step": 8127 }, { "epoch": 2.623753985229428, "grad_norm": 0.37617613215778994, "learning_rate": 6.335073786964823e-07, "loss": 0.2271, "step": 8128 }, { "epoch": 2.62407683925905, "grad_norm": 0.4012343615127607, "learning_rate": 6.324339796699307e-07, "loss": 0.2452, "step": 8129 }, { "epoch": 2.6243996932886717, "grad_norm": 0.37540392856480587, "learning_rate": 6.313614507545381e-07, "loss": 0.2152, "step": 8130 }, { "epoch": 2.6247225473182936, "grad_norm": 0.34763206609247804, "learning_rate": 6.3028979208619e-07, "loss": 0.2201, "step": 8131 }, { "epoch": 2.6250454013479154, "grad_norm": 0.4883496399545963, "learning_rate": 6.292190038006659e-07, "loss": 0.2676, "step": 8132 }, { "epoch": 2.6253682553775377, "grad_norm": 0.43604291637716064, "learning_rate": 6.281490860336309e-07, "loss": 0.2531, "step": 8133 }, { "epoch": 2.6256911094071596, "grad_norm": 0.5139841122757072, "learning_rate": 6.270800389206455e-07, "loss": 0.31, "step": 8134 }, { "epoch": 2.6260139634367814, "grad_norm": 0.35913863305235166, "learning_rate": 6.260118625971531e-07, "loss": 0.2403, "step": 8135 }, { "epoch": 2.6263368174664032, "grad_norm": 0.40615997311883323, "learning_rate": 6.249445571984921e-07, "loss": 0.2468, "step": 8136 }, { "epoch": 2.626659671496025, "grad_norm": 0.35582722831296143, "learning_rate": 6.238781228598877e-07, "loss": 0.2243, "step": 8137 }, { "epoch": 2.626982525525647, "grad_norm": 0.2962104677744298, "learning_rate": 6.228125597164578e-07, "loss": 0.237, "step": 8138 }, { "epoch": 2.6273053795552688, "grad_norm": 0.3639654550299303, "learning_rate": 6.217478679032041e-07, "loss": 0.2674, "step": 8139 }, { "epoch": 2.6276282335848906, "grad_norm": 0.43883468588330315, "learning_rate": 6.206840475550238e-07, "loss": 0.2549, "step": 8140 }, { "epoch": 2.6279510876145125, "grad_norm": 0.39477753161955725, "learning_rate": 6.196210988067033e-07, "loss": 0.2003, "step": 8141 }, { "epoch": 2.6282739416441343, "grad_norm": 0.3870064293440795, "learning_rate": 6.185590217929124e-07, "loss": 0.2519, "step": 8142 }, { "epoch": 2.628596795673756, "grad_norm": 0.41791994958576484, "learning_rate": 6.174978166482176e-07, "loss": 0.233, "step": 8143 }, { "epoch": 2.628919649703378, "grad_norm": 0.42734930977674906, "learning_rate": 6.164374835070729e-07, "loss": 0.275, "step": 8144 }, { "epoch": 2.629242503733, "grad_norm": 0.39723537229693523, "learning_rate": 6.153780225038186e-07, "loss": 0.2618, "step": 8145 }, { "epoch": 2.6295653577626217, "grad_norm": 0.5106855855366655, "learning_rate": 6.143194337726896e-07, "loss": 0.1757, "step": 8146 }, { "epoch": 2.6298882117922435, "grad_norm": 0.4653656642141686, "learning_rate": 6.132617174478045e-07, "loss": 0.2874, "step": 8147 }, { "epoch": 2.6302110658218654, "grad_norm": 0.4364078373822539, "learning_rate": 6.122048736631761e-07, "loss": 0.2762, "step": 8148 }, { "epoch": 2.630533919851487, "grad_norm": 0.38458902350583485, "learning_rate": 6.111489025527073e-07, "loss": 0.2664, "step": 8149 }, { "epoch": 2.630856773881109, "grad_norm": 0.32637825093997097, "learning_rate": 6.100938042501835e-07, "loss": 0.2442, "step": 8150 }, { "epoch": 2.631179627910731, "grad_norm": 0.38867474614765446, "learning_rate": 6.090395788892889e-07, "loss": 0.2081, "step": 8151 }, { "epoch": 2.6315024819403527, "grad_norm": 0.38698258750275827, "learning_rate": 6.079862266035887e-07, "loss": 0.2957, "step": 8152 }, { "epoch": 2.6318253359699746, "grad_norm": 0.40957907565517787, "learning_rate": 6.06933747526544e-07, "loss": 0.2616, "step": 8153 }, { "epoch": 2.6321481899995964, "grad_norm": 0.38673968106069645, "learning_rate": 6.058821417915006e-07, "loss": 0.2246, "step": 8154 }, { "epoch": 2.6324710440292183, "grad_norm": 0.456063302371948, "learning_rate": 6.04831409531696e-07, "loss": 0.2527, "step": 8155 }, { "epoch": 2.63279389805884, "grad_norm": 0.361968848745156, "learning_rate": 6.03781550880256e-07, "loss": 0.2732, "step": 8156 }, { "epoch": 2.633116752088462, "grad_norm": 0.3771805946657483, "learning_rate": 6.027325659701971e-07, "loss": 0.313, "step": 8157 }, { "epoch": 2.633439606118084, "grad_norm": 0.43833853193316796, "learning_rate": 6.016844549344236e-07, "loss": 0.2327, "step": 8158 }, { "epoch": 2.6337624601477057, "grad_norm": 0.38279335677911064, "learning_rate": 6.006372179057318e-07, "loss": 0.2706, "step": 8159 }, { "epoch": 2.6340853141773275, "grad_norm": 0.3524893347922074, "learning_rate": 5.995908550168022e-07, "loss": 0.2684, "step": 8160 }, { "epoch": 2.6344081682069493, "grad_norm": 0.36511983968461903, "learning_rate": 5.985453664002097e-07, "loss": 0.2527, "step": 8161 }, { "epoch": 2.634731022236571, "grad_norm": 0.437417364158682, "learning_rate": 5.975007521884141e-07, "loss": 0.2545, "step": 8162 }, { "epoch": 2.635053876266193, "grad_norm": 0.4358299764034365, "learning_rate": 5.964570125137695e-07, "loss": 0.2388, "step": 8163 }, { "epoch": 2.635376730295815, "grad_norm": 0.3873833678353282, "learning_rate": 5.954141475085134e-07, "loss": 0.2119, "step": 8164 }, { "epoch": 2.6356995843254367, "grad_norm": 0.3710066154928721, "learning_rate": 5.943721573047761e-07, "loss": 0.2214, "step": 8165 }, { "epoch": 2.6360224383550586, "grad_norm": 0.39823295776099427, "learning_rate": 5.933310420345786e-07, "loss": 0.3493, "step": 8166 }, { "epoch": 2.6363452923846804, "grad_norm": 0.4247581113901381, "learning_rate": 5.922908018298254e-07, "loss": 0.3017, "step": 8167 }, { "epoch": 2.6366681464143022, "grad_norm": 0.3010745365819989, "learning_rate": 5.912514368223168e-07, "loss": 0.2372, "step": 8168 }, { "epoch": 2.636991000443924, "grad_norm": 0.33613312575860954, "learning_rate": 5.902129471437359e-07, "loss": 0.2557, "step": 8169 }, { "epoch": 2.637313854473546, "grad_norm": 0.3755072505631129, "learning_rate": 5.891753329256608e-07, "loss": 0.2676, "step": 8170 }, { "epoch": 2.637636708503168, "grad_norm": 0.369875070394633, "learning_rate": 5.881385942995521e-07, "loss": 0.2061, "step": 8171 }, { "epoch": 2.6379595625327896, "grad_norm": 0.42285086285517565, "learning_rate": 5.871027313967664e-07, "loss": 0.2767, "step": 8172 }, { "epoch": 2.6382824165624115, "grad_norm": 0.3508156272427179, "learning_rate": 5.860677443485455e-07, "loss": 0.2978, "step": 8173 }, { "epoch": 2.6386052705920333, "grad_norm": 0.36611133232082665, "learning_rate": 5.850336332860192e-07, "loss": 0.2405, "step": 8174 }, { "epoch": 2.638928124621655, "grad_norm": 0.3853694181162026, "learning_rate": 5.84000398340209e-07, "loss": 0.2923, "step": 8175 }, { "epoch": 2.6392509786512774, "grad_norm": 0.4134236655656622, "learning_rate": 5.829680396420265e-07, "loss": 0.2239, "step": 8176 }, { "epoch": 2.6395738326808993, "grad_norm": 0.3793917704628371, "learning_rate": 5.819365573222654e-07, "loss": 0.3051, "step": 8177 }, { "epoch": 2.639896686710521, "grad_norm": 0.3196843483642784, "learning_rate": 5.80905951511618e-07, "loss": 0.2753, "step": 8178 }, { "epoch": 2.640219540740143, "grad_norm": 0.39427278278276057, "learning_rate": 5.798762223406573e-07, "loss": 0.3058, "step": 8179 }, { "epoch": 2.640542394769765, "grad_norm": 0.49009606638266623, "learning_rate": 5.788473699398506e-07, "loss": 0.2042, "step": 8180 }, { "epoch": 2.6408652487993867, "grad_norm": 0.31093859015170056, "learning_rate": 5.778193944395494e-07, "loss": 0.2343, "step": 8181 }, { "epoch": 2.6411881028290085, "grad_norm": 0.39987713999270397, "learning_rate": 5.767922959700004e-07, "loss": 0.3013, "step": 8182 }, { "epoch": 2.6415109568586304, "grad_norm": 0.2977848689003901, "learning_rate": 5.757660746613319e-07, "loss": 0.2302, "step": 8183 }, { "epoch": 2.641833810888252, "grad_norm": 0.4006879003096518, "learning_rate": 5.747407306435668e-07, "loss": 0.2576, "step": 8184 }, { "epoch": 2.642156664917874, "grad_norm": 0.362966498730754, "learning_rate": 5.73716264046616e-07, "loss": 0.2869, "step": 8185 }, { "epoch": 2.642479518947496, "grad_norm": 0.3843317458829998, "learning_rate": 5.726926750002748e-07, "loss": 0.2351, "step": 8186 }, { "epoch": 2.6428023729771177, "grad_norm": 0.37728572279901207, "learning_rate": 5.716699636342338e-07, "loss": 0.2252, "step": 8187 }, { "epoch": 2.6431252270067396, "grad_norm": 0.4233363988543874, "learning_rate": 5.706481300780658e-07, "loss": 0.2295, "step": 8188 }, { "epoch": 2.6434480810363614, "grad_norm": 0.40519358541284073, "learning_rate": 5.696271744612375e-07, "loss": 0.3066, "step": 8189 }, { "epoch": 2.6437709350659833, "grad_norm": 0.37618578166141303, "learning_rate": 5.686070969131027e-07, "loss": 0.2599, "step": 8190 }, { "epoch": 2.644093789095605, "grad_norm": 0.49856847165692575, "learning_rate": 5.675878975629048e-07, "loss": 0.2499, "step": 8191 }, { "epoch": 2.644416643125227, "grad_norm": 0.33826827064550313, "learning_rate": 5.66569576539773e-07, "loss": 0.2219, "step": 8192 }, { "epoch": 2.644739497154849, "grad_norm": 0.376397676274814, "learning_rate": 5.655521339727291e-07, "loss": 0.2663, "step": 8193 }, { "epoch": 2.6450623511844706, "grad_norm": 0.4521011377404161, "learning_rate": 5.645355699906791e-07, "loss": 0.2583, "step": 8194 }, { "epoch": 2.6453852052140925, "grad_norm": 0.37412502537201636, "learning_rate": 5.635198847224223e-07, "loss": 0.3053, "step": 8195 }, { "epoch": 2.6457080592437143, "grad_norm": 0.3635460223072807, "learning_rate": 5.625050782966438e-07, "loss": 0.3047, "step": 8196 }, { "epoch": 2.646030913273336, "grad_norm": 0.35315327989921985, "learning_rate": 5.614911508419185e-07, "loss": 0.2588, "step": 8197 }, { "epoch": 2.646353767302958, "grad_norm": 0.3913539293590334, "learning_rate": 5.60478102486709e-07, "loss": 0.312, "step": 8198 }, { "epoch": 2.64667662133258, "grad_norm": 0.3236076857226404, "learning_rate": 5.594659333593691e-07, "loss": 0.252, "step": 8199 }, { "epoch": 2.6469994753622017, "grad_norm": 0.35200192149848003, "learning_rate": 5.584546435881355e-07, "loss": 0.2334, "step": 8200 }, { "epoch": 2.647322329391824, "grad_norm": 0.3707935862585348, "learning_rate": 5.574442333011415e-07, "loss": 0.2416, "step": 8201 }, { "epoch": 2.647645183421446, "grad_norm": 0.4054477859631604, "learning_rate": 5.564347026264014e-07, "loss": 0.2833, "step": 8202 }, { "epoch": 2.6479680374510677, "grad_norm": 0.31851538122780115, "learning_rate": 5.554260516918228e-07, "loss": 0.2525, "step": 8203 }, { "epoch": 2.6482908914806895, "grad_norm": 0.38897693771908987, "learning_rate": 5.544182806251993e-07, "loss": 0.3587, "step": 8204 }, { "epoch": 2.6486137455103114, "grad_norm": 0.3666230438283307, "learning_rate": 5.534113895542157e-07, "loss": 0.2837, "step": 8205 }, { "epoch": 2.648936599539933, "grad_norm": 0.37813652852243657, "learning_rate": 5.524053786064445e-07, "loss": 0.2556, "step": 8206 }, { "epoch": 2.649259453569555, "grad_norm": 0.35506996668874885, "learning_rate": 5.514002479093425e-07, "loss": 0.2621, "step": 8207 }, { "epoch": 2.649582307599177, "grad_norm": 0.4024844996810088, "learning_rate": 5.503959975902617e-07, "loss": 0.275, "step": 8208 }, { "epoch": 2.6499051616287987, "grad_norm": 0.3979410686972286, "learning_rate": 5.493926277764372e-07, "loss": 0.2296, "step": 8209 }, { "epoch": 2.6502280156584206, "grad_norm": 0.34042166822715353, "learning_rate": 5.483901385949963e-07, "loss": 0.2542, "step": 8210 }, { "epoch": 2.6505508696880424, "grad_norm": 0.3968296116574713, "learning_rate": 5.473885301729501e-07, "loss": 0.3142, "step": 8211 }, { "epoch": 2.6508737237176643, "grad_norm": 0.4158539918718204, "learning_rate": 5.463878026372043e-07, "loss": 0.2551, "step": 8212 }, { "epoch": 2.651196577747286, "grad_norm": 0.40672290521255505, "learning_rate": 5.45387956114547e-07, "loss": 0.2899, "step": 8213 }, { "epoch": 2.651519431776908, "grad_norm": 0.3958475100265249, "learning_rate": 5.443889907316599e-07, "loss": 0.2735, "step": 8214 }, { "epoch": 2.65184228580653, "grad_norm": 0.36316121610308144, "learning_rate": 5.433909066151077e-07, "loss": 0.229, "step": 8215 }, { "epoch": 2.6521651398361517, "grad_norm": 0.4281010687778601, "learning_rate": 5.423937038913493e-07, "loss": 0.314, "step": 8216 }, { "epoch": 2.6524879938657735, "grad_norm": 0.3216104639858963, "learning_rate": 5.413973826867261e-07, "loss": 0.2401, "step": 8217 }, { "epoch": 2.6528108478953953, "grad_norm": 0.3856446027585075, "learning_rate": 5.40401943127472e-07, "loss": 0.2523, "step": 8218 }, { "epoch": 2.653133701925017, "grad_norm": 0.38346861523881204, "learning_rate": 5.394073853397089e-07, "loss": 0.228, "step": 8219 }, { "epoch": 2.653456555954639, "grad_norm": 0.4091277157119597, "learning_rate": 5.38413709449444e-07, "loss": 0.3398, "step": 8220 }, { "epoch": 2.653779409984261, "grad_norm": 0.37359583694727394, "learning_rate": 5.374209155825752e-07, "loss": 0.2616, "step": 8221 }, { "epoch": 2.6541022640138827, "grad_norm": 0.4016217711540157, "learning_rate": 5.364290038648883e-07, "loss": 0.2612, "step": 8222 }, { "epoch": 2.6544251180435046, "grad_norm": 0.3920163778562347, "learning_rate": 5.354379744220581e-07, "loss": 0.2955, "step": 8223 }, { "epoch": 2.6547479720731264, "grad_norm": 0.43235700947682515, "learning_rate": 5.344478273796447e-07, "loss": 0.2337, "step": 8224 }, { "epoch": 2.6550708261027482, "grad_norm": 0.3290776133521207, "learning_rate": 5.334585628631011e-07, "loss": 0.2303, "step": 8225 }, { "epoch": 2.65539368013237, "grad_norm": 0.3800198917507254, "learning_rate": 5.324701809977623e-07, "loss": 0.2801, "step": 8226 }, { "epoch": 2.655716534161992, "grad_norm": 0.38752923299359104, "learning_rate": 5.314826819088586e-07, "loss": 0.2882, "step": 8227 }, { "epoch": 2.656039388191614, "grad_norm": 0.3327544565629028, "learning_rate": 5.30496065721501e-07, "loss": 0.2389, "step": 8228 }, { "epoch": 2.6563622422212356, "grad_norm": 0.4367551895415022, "learning_rate": 5.295103325606953e-07, "loss": 0.2997, "step": 8229 }, { "epoch": 2.6566850962508575, "grad_norm": 0.3165371841044339, "learning_rate": 5.285254825513308e-07, "loss": 0.2511, "step": 8230 }, { "epoch": 2.6570079502804793, "grad_norm": 0.3822215411308749, "learning_rate": 5.275415158181885e-07, "loss": 0.2852, "step": 8231 }, { "epoch": 2.657330804310101, "grad_norm": 0.4356934004195573, "learning_rate": 5.265584324859335e-07, "loss": 0.2775, "step": 8232 }, { "epoch": 2.657653658339723, "grad_norm": 0.40482261623571214, "learning_rate": 5.25576232679123e-07, "loss": 0.2292, "step": 8233 }, { "epoch": 2.657976512369345, "grad_norm": 0.3835878695509016, "learning_rate": 5.245949165221984e-07, "loss": 0.2894, "step": 8234 }, { "epoch": 2.6582993663989667, "grad_norm": 0.3772173536224242, "learning_rate": 5.236144841394924e-07, "loss": 0.2627, "step": 8235 }, { "epoch": 2.6586222204285885, "grad_norm": 0.3656357548155923, "learning_rate": 5.226349356552243e-07, "loss": 0.2891, "step": 8236 }, { "epoch": 2.6589450744582104, "grad_norm": 0.38660019212429364, "learning_rate": 5.216562711935021e-07, "loss": 0.2545, "step": 8237 }, { "epoch": 2.6592679284878322, "grad_norm": 0.3894507862652017, "learning_rate": 5.206784908783193e-07, "loss": 0.2844, "step": 8238 }, { "epoch": 2.659590782517454, "grad_norm": 0.34342601588342286, "learning_rate": 5.197015948335609e-07, "loss": 0.2708, "step": 8239 }, { "epoch": 2.659913636547076, "grad_norm": 0.37913290281507667, "learning_rate": 5.187255831829995e-07, "loss": 0.2787, "step": 8240 }, { "epoch": 2.6602364905766978, "grad_norm": 0.37990272277704634, "learning_rate": 5.177504560502913e-07, "loss": 0.3003, "step": 8241 }, { "epoch": 2.6605593446063196, "grad_norm": 0.3603167175894175, "learning_rate": 5.167762135589866e-07, "loss": 0.2827, "step": 8242 }, { "epoch": 2.660882198635942, "grad_norm": 0.39232332902925676, "learning_rate": 5.158028558325182e-07, "loss": 0.1944, "step": 8243 }, { "epoch": 2.6612050526655637, "grad_norm": 0.3723970643797824, "learning_rate": 5.148303829942108e-07, "loss": 0.2788, "step": 8244 }, { "epoch": 2.6615279066951856, "grad_norm": 0.35130355605921765, "learning_rate": 5.138587951672735e-07, "loss": 0.2444, "step": 8245 }, { "epoch": 2.6618507607248074, "grad_norm": 0.33931785217933347, "learning_rate": 5.128880924748078e-07, "loss": 0.268, "step": 8246 }, { "epoch": 2.6621736147544293, "grad_norm": 0.3710224941732591, "learning_rate": 5.119182750397977e-07, "loss": 0.2799, "step": 8247 }, { "epoch": 2.662496468784051, "grad_norm": 0.4432995792959867, "learning_rate": 5.109493429851208e-07, "loss": 0.2207, "step": 8248 }, { "epoch": 2.662819322813673, "grad_norm": 0.38004163447528083, "learning_rate": 5.099812964335357e-07, "loss": 0.2041, "step": 8249 }, { "epoch": 2.663142176843295, "grad_norm": 0.4281406518746054, "learning_rate": 5.09014135507695e-07, "loss": 0.2469, "step": 8250 }, { "epoch": 2.6634650308729166, "grad_norm": 0.3609650029267467, "learning_rate": 5.080478603301356e-07, "loss": 0.2663, "step": 8251 }, { "epoch": 2.6637878849025385, "grad_norm": 0.37690764729647075, "learning_rate": 5.070824710232855e-07, "loss": 0.2817, "step": 8252 }, { "epoch": 2.6641107389321603, "grad_norm": 0.42390667196533827, "learning_rate": 5.061179677094552e-07, "loss": 0.2964, "step": 8253 }, { "epoch": 2.664433592961782, "grad_norm": 0.3473700989510802, "learning_rate": 5.051543505108477e-07, "loss": 0.2703, "step": 8254 }, { "epoch": 2.664756446991404, "grad_norm": 0.376482040149485, "learning_rate": 5.041916195495513e-07, "loss": 0.27, "step": 8255 }, { "epoch": 2.665079301021026, "grad_norm": 0.44071865462587845, "learning_rate": 5.032297749475434e-07, "loss": 0.2361, "step": 8256 }, { "epoch": 2.6654021550506477, "grad_norm": 0.3319850434981893, "learning_rate": 5.022688168266873e-07, "loss": 0.2614, "step": 8257 }, { "epoch": 2.6657250090802695, "grad_norm": 0.44394063337869993, "learning_rate": 5.013087453087356e-07, "loss": 0.2728, "step": 8258 }, { "epoch": 2.6660478631098914, "grad_norm": 0.3838894211252364, "learning_rate": 5.003495605153293e-07, "loss": 0.2773, "step": 8259 }, { "epoch": 2.6663707171395132, "grad_norm": 0.4023193211484546, "learning_rate": 4.993912625679931e-07, "loss": 0.2624, "step": 8260 }, { "epoch": 2.666693571169135, "grad_norm": 0.36346511364783307, "learning_rate": 4.984338515881446e-07, "loss": 0.2894, "step": 8261 }, { "epoch": 2.667016425198757, "grad_norm": 0.3215258380811906, "learning_rate": 4.974773276970842e-07, "loss": 0.2623, "step": 8262 }, { "epoch": 2.6673392792283788, "grad_norm": 0.4472822926925615, "learning_rate": 4.965216910160053e-07, "loss": 0.2478, "step": 8263 }, { "epoch": 2.6676621332580006, "grad_norm": 0.3491925017439787, "learning_rate": 4.955669416659817e-07, "loss": 0.2972, "step": 8264 }, { "epoch": 2.6679849872876225, "grad_norm": 0.3681406664314807, "learning_rate": 4.946130797679823e-07, "loss": 0.2057, "step": 8265 }, { "epoch": 2.6683078413172443, "grad_norm": 0.3105784900236587, "learning_rate": 4.936601054428575e-07, "loss": 0.2248, "step": 8266 }, { "epoch": 2.668630695346866, "grad_norm": 0.3818406079336815, "learning_rate": 4.927080188113495e-07, "loss": 0.2608, "step": 8267 }, { "epoch": 2.6689535493764884, "grad_norm": 0.4315379815873287, "learning_rate": 4.917568199940852e-07, "loss": 0.2222, "step": 8268 }, { "epoch": 2.6692764034061103, "grad_norm": 0.3666072180831491, "learning_rate": 4.908065091115826e-07, "loss": 0.2521, "step": 8269 }, { "epoch": 2.669599257435732, "grad_norm": 0.3893030937728807, "learning_rate": 4.89857086284242e-07, "loss": 0.2455, "step": 8270 }, { "epoch": 2.669922111465354, "grad_norm": 0.36632716742811416, "learning_rate": 4.88908551632356e-07, "loss": 0.2563, "step": 8271 }, { "epoch": 2.670244965494976, "grad_norm": 0.3964891365754927, "learning_rate": 4.879609052760997e-07, "loss": 0.2651, "step": 8272 }, { "epoch": 2.6705678195245977, "grad_norm": 0.4065750011320268, "learning_rate": 4.870141473355427e-07, "loss": 0.2481, "step": 8273 }, { "epoch": 2.6708906735542195, "grad_norm": 0.35047319422340134, "learning_rate": 4.860682779306342e-07, "loss": 0.276, "step": 8274 }, { "epoch": 2.6712135275838413, "grad_norm": 0.5695050381786763, "learning_rate": 4.851232971812172e-07, "loss": 0.2709, "step": 8275 }, { "epoch": 2.671536381613463, "grad_norm": 0.40123279815342994, "learning_rate": 4.841792052070179e-07, "loss": 0.2617, "step": 8276 }, { "epoch": 2.671859235643085, "grad_norm": 0.47494241873709403, "learning_rate": 4.832360021276514e-07, "loss": 0.2625, "step": 8277 }, { "epoch": 2.672182089672707, "grad_norm": 0.5397868557557777, "learning_rate": 4.822936880626219e-07, "loss": 0.3079, "step": 8278 }, { "epoch": 2.6725049437023287, "grad_norm": 0.33350039738502485, "learning_rate": 4.813522631313174e-07, "loss": 0.2417, "step": 8279 }, { "epoch": 2.6728277977319506, "grad_norm": 0.4003592284803381, "learning_rate": 4.804117274530168e-07, "loss": 0.2655, "step": 8280 }, { "epoch": 2.6731506517615724, "grad_norm": 0.44102436419146795, "learning_rate": 4.794720811468828e-07, "loss": 0.2533, "step": 8281 }, { "epoch": 2.6734735057911942, "grad_norm": 0.3513271263552698, "learning_rate": 4.785333243319678e-07, "loss": 0.2138, "step": 8282 }, { "epoch": 2.673796359820816, "grad_norm": 0.3619074675397845, "learning_rate": 4.775954571272118e-07, "loss": 0.2797, "step": 8283 }, { "epoch": 2.674119213850438, "grad_norm": 0.37982920807678616, "learning_rate": 4.76658479651442e-07, "loss": 0.2437, "step": 8284 }, { "epoch": 2.67444206788006, "grad_norm": 0.4030405450681465, "learning_rate": 4.7572239202337024e-07, "loss": 0.2509, "step": 8285 }, { "epoch": 2.6747649219096816, "grad_norm": 0.40381731469478044, "learning_rate": 4.747871943615992e-07, "loss": 0.2639, "step": 8286 }, { "epoch": 2.6750877759393035, "grad_norm": 0.30613562017861173, "learning_rate": 4.738528867846148e-07, "loss": 0.2345, "step": 8287 }, { "epoch": 2.6754106299689253, "grad_norm": 0.38483944790956204, "learning_rate": 4.729194694107958e-07, "loss": 0.2579, "step": 8288 }, { "epoch": 2.675733483998547, "grad_norm": 0.3725869508900712, "learning_rate": 4.719869423584025e-07, "loss": 0.2222, "step": 8289 }, { "epoch": 2.676056338028169, "grad_norm": 0.33726900487006267, "learning_rate": 4.7105530574558545e-07, "loss": 0.2706, "step": 8290 }, { "epoch": 2.676379192057791, "grad_norm": 0.3706675052685225, "learning_rate": 4.7012455969038193e-07, "loss": 0.2431, "step": 8291 }, { "epoch": 2.6767020460874127, "grad_norm": 0.41606968876141703, "learning_rate": 4.691947043107159e-07, "loss": 0.2608, "step": 8292 }, { "epoch": 2.6770249001170345, "grad_norm": 0.385456284377077, "learning_rate": 4.682657397243981e-07, "loss": 0.302, "step": 8293 }, { "epoch": 2.6773477541466564, "grad_norm": 0.4600800709728665, "learning_rate": 4.673376660491277e-07, "loss": 0.284, "step": 8294 }, { "epoch": 2.677670608176278, "grad_norm": 0.3849427403991954, "learning_rate": 4.664104834024924e-07, "loss": 0.2508, "step": 8295 }, { "epoch": 2.6779934622059, "grad_norm": 0.3605593458443482, "learning_rate": 4.654841919019606e-07, "loss": 0.2119, "step": 8296 }, { "epoch": 2.678316316235522, "grad_norm": 0.36636170452576233, "learning_rate": 4.645587916648969e-07, "loss": 0.2302, "step": 8297 }, { "epoch": 2.6786391702651438, "grad_norm": 0.38562633136914004, "learning_rate": 4.636342828085441e-07, "loss": 0.274, "step": 8298 }, { "epoch": 2.6789620242947656, "grad_norm": 0.3705194546873476, "learning_rate": 4.627106654500385e-07, "loss": 0.2967, "step": 8299 }, { "epoch": 2.6792848783243874, "grad_norm": 0.3562983581083007, "learning_rate": 4.617879397063998e-07, "loss": 0.2299, "step": 8300 }, { "epoch": 2.6796077323540093, "grad_norm": 0.43257171064781486, "learning_rate": 4.608661056945387e-07, "loss": 0.2907, "step": 8301 }, { "epoch": 2.679930586383631, "grad_norm": 0.39317519974640486, "learning_rate": 4.5994516353124765e-07, "loss": 0.2553, "step": 8302 }, { "epoch": 2.680253440413253, "grad_norm": 0.42509812862477453, "learning_rate": 4.590251133332099e-07, "loss": 0.173, "step": 8303 }, { "epoch": 2.680576294442875, "grad_norm": 0.3802382708132515, "learning_rate": 4.581059552169939e-07, "loss": 0.2974, "step": 8304 }, { "epoch": 2.6808991484724967, "grad_norm": 0.39421471165205546, "learning_rate": 4.571876892990573e-07, "loss": 0.2735, "step": 8305 }, { "epoch": 2.6812220025021185, "grad_norm": 0.3463093692613223, "learning_rate": 4.562703156957404e-07, "loss": 0.2424, "step": 8306 }, { "epoch": 2.6815448565317404, "grad_norm": 0.4369473923258659, "learning_rate": 4.5535383452327604e-07, "loss": 0.2701, "step": 8307 }, { "epoch": 2.681867710561362, "grad_norm": 0.3358198746243374, "learning_rate": 4.544382458977789e-07, "loss": 0.268, "step": 8308 }, { "epoch": 2.682190564590984, "grad_norm": 0.33880868869256653, "learning_rate": 4.535235499352544e-07, "loss": 0.2173, "step": 8309 }, { "epoch": 2.682513418620606, "grad_norm": 0.36927091303481263, "learning_rate": 4.526097467515916e-07, "loss": 0.2352, "step": 8310 }, { "epoch": 2.682836272650228, "grad_norm": 0.33868237191472095, "learning_rate": 4.5169683646257037e-07, "loss": 0.2393, "step": 8311 }, { "epoch": 2.68315912667985, "grad_norm": 0.38674689659425404, "learning_rate": 4.5078481918385234e-07, "loss": 0.2796, "step": 8312 }, { "epoch": 2.683481980709472, "grad_norm": 0.37576768706495717, "learning_rate": 4.498736950309901e-07, "loss": 0.2191, "step": 8313 }, { "epoch": 2.6838048347390937, "grad_norm": 0.44446116472724895, "learning_rate": 4.489634641194229e-07, "loss": 0.2855, "step": 8314 }, { "epoch": 2.6841276887687155, "grad_norm": 0.443145458761476, "learning_rate": 4.4805412656447443e-07, "loss": 0.2679, "step": 8315 }, { "epoch": 2.6844505427983374, "grad_norm": 0.41905304722574205, "learning_rate": 4.4714568248135827e-07, "loss": 0.3094, "step": 8316 }, { "epoch": 2.6847733968279592, "grad_norm": 0.3892774980745633, "learning_rate": 4.462381319851708e-07, "loss": 0.2867, "step": 8317 }, { "epoch": 2.685096250857581, "grad_norm": 0.35976435741531804, "learning_rate": 4.453314751908991e-07, "loss": 0.2422, "step": 8318 }, { "epoch": 2.685419104887203, "grad_norm": 0.3832597830591287, "learning_rate": 4.444257122134138e-07, "loss": 0.2932, "step": 8319 }, { "epoch": 2.6857419589168248, "grad_norm": 0.3686765738882611, "learning_rate": 4.435208431674756e-07, "loss": 0.2609, "step": 8320 }, { "epoch": 2.6860648129464466, "grad_norm": 0.35493572940547435, "learning_rate": 4.4261686816772867e-07, "loss": 0.2612, "step": 8321 }, { "epoch": 2.6863876669760685, "grad_norm": 0.41316026661661426, "learning_rate": 4.4171378732870636e-07, "loss": 0.2514, "step": 8322 }, { "epoch": 2.6867105210056903, "grad_norm": 0.3634311329244401, "learning_rate": 4.408116007648272e-07, "loss": 0.279, "step": 8323 }, { "epoch": 2.687033375035312, "grad_norm": 0.38727983301889085, "learning_rate": 4.3991030859039714e-07, "loss": 0.2523, "step": 8324 }, { "epoch": 2.687356229064934, "grad_norm": 0.37161750364843543, "learning_rate": 4.3900991091960827e-07, "loss": 0.2158, "step": 8325 }, { "epoch": 2.687679083094556, "grad_norm": 0.35346151902401635, "learning_rate": 4.38110407866541e-07, "loss": 0.3247, "step": 8326 }, { "epoch": 2.6880019371241777, "grad_norm": 0.464303192512684, "learning_rate": 4.372117995451591e-07, "loss": 0.2162, "step": 8327 }, { "epoch": 2.6883247911537995, "grad_norm": 0.38521982335491073, "learning_rate": 4.3631408606931574e-07, "loss": 0.2558, "step": 8328 }, { "epoch": 2.6886476451834214, "grad_norm": 0.4192873583112565, "learning_rate": 4.354172675527515e-07, "loss": 0.2868, "step": 8329 }, { "epoch": 2.688970499213043, "grad_norm": 0.3420206402995257, "learning_rate": 4.3452134410909145e-07, "loss": 0.2284, "step": 8330 }, { "epoch": 2.689293353242665, "grad_norm": 0.37989307431794744, "learning_rate": 4.3362631585184555e-07, "loss": 0.2467, "step": 8331 }, { "epoch": 2.689616207272287, "grad_norm": 0.35007155728248746, "learning_rate": 4.3273218289441556e-07, "loss": 0.2088, "step": 8332 }, { "epoch": 2.6899390613019087, "grad_norm": 0.3479916673312957, "learning_rate": 4.3183894535008594e-07, "loss": 0.2458, "step": 8333 }, { "epoch": 2.6902619153315306, "grad_norm": 0.3894135200268328, "learning_rate": 4.3094660333202865e-07, "loss": 0.2506, "step": 8334 }, { "epoch": 2.6905847693611524, "grad_norm": 0.34857036301862926, "learning_rate": 4.300551569533023e-07, "loss": 0.2523, "step": 8335 }, { "epoch": 2.6909076233907747, "grad_norm": 0.33841175348169344, "learning_rate": 4.2916460632685005e-07, "loss": 0.2262, "step": 8336 }, { "epoch": 2.6912304774203966, "grad_norm": 0.3922358333611762, "learning_rate": 4.282749515655074e-07, "loss": 0.2483, "step": 8337 }, { "epoch": 2.6915533314500184, "grad_norm": 0.3096230364256217, "learning_rate": 4.273861927819883e-07, "loss": 0.221, "step": 8338 }, { "epoch": 2.6918761854796402, "grad_norm": 0.37455009789138843, "learning_rate": 4.264983300889003e-07, "loss": 0.2348, "step": 8339 }, { "epoch": 2.692199039509262, "grad_norm": 0.30532359311020013, "learning_rate": 4.2561136359873157e-07, "loss": 0.2368, "step": 8340 }, { "epoch": 2.692521893538884, "grad_norm": 0.3472047607491473, "learning_rate": 4.247252934238624e-07, "loss": 0.2362, "step": 8341 }, { "epoch": 2.692844747568506, "grad_norm": 0.35699990710489893, "learning_rate": 4.2384011967655457e-07, "loss": 0.2721, "step": 8342 }, { "epoch": 2.6931676015981276, "grad_norm": 0.37329880830211015, "learning_rate": 4.229558424689592e-07, "loss": 0.2352, "step": 8343 }, { "epoch": 2.6934904556277495, "grad_norm": 0.457846645571161, "learning_rate": 4.220724619131125e-07, "loss": 0.3077, "step": 8344 }, { "epoch": 2.6938133096573713, "grad_norm": 0.3588446939745074, "learning_rate": 4.211899781209383e-07, "loss": 0.254, "step": 8345 }, { "epoch": 2.694136163686993, "grad_norm": 0.358015736102339, "learning_rate": 4.2030839120424545e-07, "loss": 0.2385, "step": 8346 }, { "epoch": 2.694459017716615, "grad_norm": 0.39510641027173504, "learning_rate": 4.1942770127473134e-07, "loss": 0.2598, "step": 8347 }, { "epoch": 2.694781871746237, "grad_norm": 0.46515467433515245, "learning_rate": 4.1854790844397585e-07, "loss": 0.2331, "step": 8348 }, { "epoch": 2.6951047257758587, "grad_norm": 0.35871384709939497, "learning_rate": 4.1766901282344977e-07, "loss": 0.2322, "step": 8349 }, { "epoch": 2.6954275798054805, "grad_norm": 0.44896670770151803, "learning_rate": 4.167910145245049e-07, "loss": 0.24, "step": 8350 }, { "epoch": 2.6957504338351024, "grad_norm": 0.38055412244752157, "learning_rate": 4.159139136583856e-07, "loss": 0.2925, "step": 8351 }, { "epoch": 2.696073287864724, "grad_norm": 0.37767039272166114, "learning_rate": 4.1503771033621797e-07, "loss": 0.2666, "step": 8352 }, { "epoch": 2.696396141894346, "grad_norm": 0.381400528986373, "learning_rate": 4.1416240466901567e-07, "loss": 0.2473, "step": 8353 }, { "epoch": 2.696718995923968, "grad_norm": 0.39425556300238895, "learning_rate": 4.1328799676768004e-07, "loss": 0.2357, "step": 8354 }, { "epoch": 2.6970418499535898, "grad_norm": 0.38598516718263504, "learning_rate": 4.124144867429941e-07, "loss": 0.2105, "step": 8355 }, { "epoch": 2.6973647039832116, "grad_norm": 0.4309862778900752, "learning_rate": 4.115418747056343e-07, "loss": 0.2175, "step": 8356 }, { "epoch": 2.6976875580128334, "grad_norm": 0.3479340867604791, "learning_rate": 4.1067016076615645e-07, "loss": 0.2018, "step": 8357 }, { "epoch": 2.6980104120424553, "grad_norm": 0.3995079007217438, "learning_rate": 4.097993450350079e-07, "loss": 0.3061, "step": 8358 }, { "epoch": 2.698333266072077, "grad_norm": 0.37406494296232556, "learning_rate": 4.089294276225172e-07, "loss": 0.2709, "step": 8359 }, { "epoch": 2.698656120101699, "grad_norm": 0.39759912888311133, "learning_rate": 4.080604086389036e-07, "loss": 0.2671, "step": 8360 }, { "epoch": 2.698978974131321, "grad_norm": 0.3435802927842299, "learning_rate": 4.0719228819426994e-07, "loss": 0.2804, "step": 8361 }, { "epoch": 2.6993018281609427, "grad_norm": 0.3676146719392127, "learning_rate": 4.063250663986065e-07, "loss": 0.2704, "step": 8362 }, { "epoch": 2.6996246821905645, "grad_norm": 0.38499338551291934, "learning_rate": 4.054587433617879e-07, "loss": 0.232, "step": 8363 }, { "epoch": 2.6999475362201864, "grad_norm": 0.39258342846687755, "learning_rate": 4.0459331919357717e-07, "loss": 0.2574, "step": 8364 }, { "epoch": 2.700270390249808, "grad_norm": 0.3798057044219868, "learning_rate": 4.037287940036216e-07, "loss": 0.2572, "step": 8365 }, { "epoch": 2.70059324427943, "grad_norm": 0.4098671180384193, "learning_rate": 4.0286516790145614e-07, "loss": 0.2851, "step": 8366 }, { "epoch": 2.700916098309052, "grad_norm": 0.39418198289961826, "learning_rate": 4.020024409964998e-07, "loss": 0.2781, "step": 8367 }, { "epoch": 2.7012389523386737, "grad_norm": 0.34819203787904923, "learning_rate": 4.011406133980602e-07, "loss": 0.2422, "step": 8368 }, { "epoch": 2.7015618063682956, "grad_norm": 0.4353838867827788, "learning_rate": 4.0027968521533007e-07, "loss": 0.2654, "step": 8369 }, { "epoch": 2.7018846603979174, "grad_norm": 0.38730063475953336, "learning_rate": 3.9941965655738546e-07, "loss": 0.2383, "step": 8370 }, { "epoch": 2.7022075144275393, "grad_norm": 0.3308169477122396, "learning_rate": 3.9856052753319326e-07, "loss": 0.2435, "step": 8371 }, { "epoch": 2.702530368457161, "grad_norm": 0.3776616339934081, "learning_rate": 3.977022982516024e-07, "loss": 0.1886, "step": 8372 }, { "epoch": 2.702853222486783, "grad_norm": 0.3648478498833274, "learning_rate": 3.968449688213499e-07, "loss": 0.2724, "step": 8373 }, { "epoch": 2.703176076516405, "grad_norm": 0.6485807605054359, "learning_rate": 3.959885393510582e-07, "loss": 0.2482, "step": 8374 }, { "epoch": 2.7034989305460266, "grad_norm": 0.3687258061183221, "learning_rate": 3.9513300994923534e-07, "loss": 0.2828, "step": 8375 }, { "epoch": 2.7038217845756485, "grad_norm": 0.3574547443501252, "learning_rate": 3.9427838072427723e-07, "loss": 0.2436, "step": 8376 }, { "epoch": 2.7041446386052703, "grad_norm": 0.4562574743151455, "learning_rate": 3.934246517844622e-07, "loss": 0.2764, "step": 8377 }, { "epoch": 2.7044674926348926, "grad_norm": 0.37444369121990395, "learning_rate": 3.92571823237958e-07, "loss": 0.2192, "step": 8378 }, { "epoch": 2.7047903466645145, "grad_norm": 0.343351854598857, "learning_rate": 3.9171989519281636e-07, "loss": 0.2279, "step": 8379 }, { "epoch": 2.7051132006941363, "grad_norm": 0.36802836772253805, "learning_rate": 3.9086886775697436e-07, "loss": 0.2725, "step": 8380 }, { "epoch": 2.705436054723758, "grad_norm": 0.4531236290933243, "learning_rate": 3.900187410382583e-07, "loss": 0.3295, "step": 8381 }, { "epoch": 2.70575890875338, "grad_norm": 0.3469999519681782, "learning_rate": 3.891695151443761e-07, "loss": 0.2496, "step": 8382 }, { "epoch": 2.706081762783002, "grad_norm": 0.4654435748482998, "learning_rate": 3.883211901829251e-07, "loss": 0.2885, "step": 8383 }, { "epoch": 2.7064046168126237, "grad_norm": 0.44304631603825667, "learning_rate": 3.874737662613842e-07, "loss": 0.2469, "step": 8384 }, { "epoch": 2.7067274708422455, "grad_norm": 0.31989812571793225, "learning_rate": 3.8662724348712433e-07, "loss": 0.2059, "step": 8385 }, { "epoch": 2.7070503248718674, "grad_norm": 0.42044396730860606, "learning_rate": 3.8578162196739544e-07, "loss": 0.2675, "step": 8386 }, { "epoch": 2.707373178901489, "grad_norm": 0.36967134585970574, "learning_rate": 3.849369018093385e-07, "loss": 0.2395, "step": 8387 }, { "epoch": 2.707696032931111, "grad_norm": 0.34411253330241676, "learning_rate": 3.840930831199788e-07, "loss": 0.2537, "step": 8388 }, { "epoch": 2.708018886960733, "grad_norm": 0.3729992816258083, "learning_rate": 3.8325016600622486e-07, "loss": 0.2422, "step": 8389 }, { "epoch": 2.7083417409903547, "grad_norm": 0.4077879830657112, "learning_rate": 3.824081505748764e-07, "loss": 0.2233, "step": 8390 }, { "epoch": 2.7086645950199766, "grad_norm": 0.31873668825183665, "learning_rate": 3.815670369326121e-07, "loss": 0.2458, "step": 8391 }, { "epoch": 2.7089874490495984, "grad_norm": 0.38293245332616255, "learning_rate": 3.8072682518600274e-07, "loss": 0.3202, "step": 8392 }, { "epoch": 2.7093103030792203, "grad_norm": 0.35680786465548453, "learning_rate": 3.7988751544149965e-07, "loss": 0.3176, "step": 8393 }, { "epoch": 2.709633157108842, "grad_norm": 0.43031852497075934, "learning_rate": 3.790491078054445e-07, "loss": 0.2607, "step": 8394 }, { "epoch": 2.709956011138464, "grad_norm": 0.3807623522342435, "learning_rate": 3.7821160238406066e-07, "loss": 0.2928, "step": 8395 }, { "epoch": 2.710278865168086, "grad_norm": 0.3053854807202077, "learning_rate": 3.773749992834599e-07, "loss": 0.2233, "step": 8396 }, { "epoch": 2.7106017191977076, "grad_norm": 0.34621514720724883, "learning_rate": 3.765392986096383e-07, "loss": 0.3179, "step": 8397 }, { "epoch": 2.7109245732273295, "grad_norm": 0.37440392131868755, "learning_rate": 3.757045004684778e-07, "loss": 0.209, "step": 8398 }, { "epoch": 2.7112474272569513, "grad_norm": 0.4358177424198082, "learning_rate": 3.748706049657463e-07, "loss": 0.2401, "step": 8399 }, { "epoch": 2.711570281286573, "grad_norm": 0.366981704131833, "learning_rate": 3.7403761220709765e-07, "loss": 0.2146, "step": 8400 }, { "epoch": 2.711893135316195, "grad_norm": 0.3970659222789045, "learning_rate": 3.732055222980699e-07, "loss": 0.2499, "step": 8401 }, { "epoch": 2.712215989345817, "grad_norm": 0.5126011221366799, "learning_rate": 3.723743353440895e-07, "loss": 0.2136, "step": 8402 }, { "epoch": 2.712538843375439, "grad_norm": 0.3661019167539218, "learning_rate": 3.7154405145046396e-07, "loss": 0.2387, "step": 8403 }, { "epoch": 2.712861697405061, "grad_norm": 0.43070917460881913, "learning_rate": 3.707146707223916e-07, "loss": 0.2468, "step": 8404 }, { "epoch": 2.713184551434683, "grad_norm": 0.3669274220469167, "learning_rate": 3.698861932649525e-07, "loss": 0.2652, "step": 8405 }, { "epoch": 2.7135074054643047, "grad_norm": 0.37641643083882736, "learning_rate": 3.6905861918311356e-07, "loss": 0.2529, "step": 8406 }, { "epoch": 2.7138302594939265, "grad_norm": 0.39756026338156203, "learning_rate": 3.682319485817276e-07, "loss": 0.262, "step": 8407 }, { "epoch": 2.7141531135235484, "grad_norm": 0.34116286561726095, "learning_rate": 3.674061815655333e-07, "loss": 0.2628, "step": 8408 }, { "epoch": 2.71447596755317, "grad_norm": 0.40074727898920676, "learning_rate": 3.665813182391528e-07, "loss": 0.2651, "step": 8409 }, { "epoch": 2.714798821582792, "grad_norm": 0.4175451453675688, "learning_rate": 3.657573587070967e-07, "loss": 0.258, "step": 8410 }, { "epoch": 2.715121675612414, "grad_norm": 0.4220360827832237, "learning_rate": 3.6493430307375896e-07, "loss": 0.2921, "step": 8411 }, { "epoch": 2.7154445296420358, "grad_norm": 0.37386800446910706, "learning_rate": 3.6411215144341954e-07, "loss": 0.2899, "step": 8412 }, { "epoch": 2.7157673836716576, "grad_norm": 0.38602363481132873, "learning_rate": 3.6329090392024424e-07, "loss": 0.2311, "step": 8413 }, { "epoch": 2.7160902377012794, "grad_norm": 0.3370478213128946, "learning_rate": 3.624705606082823e-07, "loss": 0.2773, "step": 8414 }, { "epoch": 2.7164130917309013, "grad_norm": 0.3644102429456883, "learning_rate": 3.6165112161147235e-07, "loss": 0.2724, "step": 8415 }, { "epoch": 2.716735945760523, "grad_norm": 0.47500445235212985, "learning_rate": 3.6083258703363456e-07, "loss": 0.272, "step": 8416 }, { "epoch": 2.717058799790145, "grad_norm": 0.3463934451795195, "learning_rate": 3.6001495697847773e-07, "loss": 0.2604, "step": 8417 }, { "epoch": 2.717381653819767, "grad_norm": 0.3660457956617593, "learning_rate": 3.5919823154959233e-07, "loss": 0.3248, "step": 8418 }, { "epoch": 2.7177045078493887, "grad_norm": 0.3174992797131989, "learning_rate": 3.583824108504588e-07, "loss": 0.2503, "step": 8419 }, { "epoch": 2.7180273618790105, "grad_norm": 0.376812320500849, "learning_rate": 3.575674949844379e-07, "loss": 0.2853, "step": 8420 }, { "epoch": 2.7183502159086324, "grad_norm": 0.3680963503500245, "learning_rate": 3.567534840547804e-07, "loss": 0.2536, "step": 8421 }, { "epoch": 2.718673069938254, "grad_norm": 0.37107223463692907, "learning_rate": 3.559403781646195e-07, "loss": 0.2308, "step": 8422 }, { "epoch": 2.718995923967876, "grad_norm": 0.37932987916622196, "learning_rate": 3.551281774169737e-07, "loss": 0.2659, "step": 8423 }, { "epoch": 2.719318777997498, "grad_norm": 0.43447344991636433, "learning_rate": 3.543168819147491e-07, "loss": 0.2884, "step": 8424 }, { "epoch": 2.7196416320271197, "grad_norm": 0.38089615399666277, "learning_rate": 3.535064917607358e-07, "loss": 0.2749, "step": 8425 }, { "epoch": 2.7199644860567416, "grad_norm": 0.40971659796443616, "learning_rate": 3.526970070576094e-07, "loss": 0.271, "step": 8426 }, { "epoch": 2.7202873400863634, "grad_norm": 0.35018500866614455, "learning_rate": 3.518884279079293e-07, "loss": 0.2727, "step": 8427 }, { "epoch": 2.7206101941159853, "grad_norm": 0.3745268466091174, "learning_rate": 3.51080754414142e-07, "loss": 0.2239, "step": 8428 }, { "epoch": 2.720933048145607, "grad_norm": 0.34536973036894975, "learning_rate": 3.5027398667857814e-07, "loss": 0.2465, "step": 8429 }, { "epoch": 2.721255902175229, "grad_norm": 0.7033697599211607, "learning_rate": 3.494681248034551e-07, "loss": 0.3455, "step": 8430 }, { "epoch": 2.721578756204851, "grad_norm": 0.3874311072920102, "learning_rate": 3.4866316889087284e-07, "loss": 0.2684, "step": 8431 }, { "epoch": 2.7219016102344726, "grad_norm": 0.32971547630138515, "learning_rate": 3.4785911904281987e-07, "loss": 0.2762, "step": 8432 }, { "epoch": 2.7222244642640945, "grad_norm": 0.4679646057012678, "learning_rate": 3.4705597536116627e-07, "loss": 0.2074, "step": 8433 }, { "epoch": 2.7225473182937163, "grad_norm": 0.4371858799060141, "learning_rate": 3.462537379476724e-07, "loss": 0.2841, "step": 8434 }, { "epoch": 2.722870172323338, "grad_norm": 0.41478584733707957, "learning_rate": 3.454524069039769e-07, "loss": 0.2431, "step": 8435 }, { "epoch": 2.72319302635296, "grad_norm": 0.38941942995805506, "learning_rate": 3.446519823316102e-07, "loss": 0.2502, "step": 8436 }, { "epoch": 2.723515880382582, "grad_norm": 0.3945278228289623, "learning_rate": 3.438524643319829e-07, "loss": 0.2655, "step": 8437 }, { "epoch": 2.7238387344122037, "grad_norm": 0.36231546091898614, "learning_rate": 3.4305385300639313e-07, "loss": 0.2764, "step": 8438 }, { "epoch": 2.7241615884418255, "grad_norm": 0.4361054591569362, "learning_rate": 3.422561484560249e-07, "loss": 0.2941, "step": 8439 }, { "epoch": 2.7244844424714474, "grad_norm": 0.3943713075655466, "learning_rate": 3.414593507819475e-07, "loss": 0.2682, "step": 8440 }, { "epoch": 2.7248072965010692, "grad_norm": 0.3718973677549862, "learning_rate": 3.406634600851108e-07, "loss": 0.235, "step": 8441 }, { "epoch": 2.725130150530691, "grad_norm": 0.3166879921858865, "learning_rate": 3.398684764663543e-07, "loss": 0.2226, "step": 8442 }, { "epoch": 2.725453004560313, "grad_norm": 0.3911350138619515, "learning_rate": 3.39074400026404e-07, "loss": 0.243, "step": 8443 }, { "epoch": 2.7257758585899348, "grad_norm": 0.42833001244663194, "learning_rate": 3.3828123086586365e-07, "loss": 0.2484, "step": 8444 }, { "epoch": 2.7260987126195566, "grad_norm": 0.40721651706097767, "learning_rate": 3.374889690852312e-07, "loss": 0.2365, "step": 8445 }, { "epoch": 2.726421566649179, "grad_norm": 0.35700984564575144, "learning_rate": 3.366976147848813e-07, "loss": 0.3176, "step": 8446 }, { "epoch": 2.7267444206788007, "grad_norm": 0.38883921692902185, "learning_rate": 3.359071680650805e-07, "loss": 0.2398, "step": 8447 }, { "epoch": 2.7270672747084226, "grad_norm": 0.3618903015366672, "learning_rate": 3.3511762902597445e-07, "loss": 0.2799, "step": 8448 }, { "epoch": 2.7273901287380444, "grad_norm": 0.397066301232012, "learning_rate": 3.343289977675998e-07, "loss": 0.2536, "step": 8449 }, { "epoch": 2.7277129827676663, "grad_norm": 0.3689252989004853, "learning_rate": 3.3354127438987245e-07, "loss": 0.2094, "step": 8450 }, { "epoch": 2.728035836797288, "grad_norm": 0.353448162109371, "learning_rate": 3.3275445899259766e-07, "loss": 0.2797, "step": 8451 }, { "epoch": 2.72835869082691, "grad_norm": 0.3857316973721198, "learning_rate": 3.319685516754614e-07, "loss": 0.2494, "step": 8452 }, { "epoch": 2.728681544856532, "grad_norm": 0.3628862920491952, "learning_rate": 3.3118355253803914e-07, "loss": 0.2948, "step": 8453 }, { "epoch": 2.7290043988861536, "grad_norm": 0.4190443352677732, "learning_rate": 3.3039946167979037e-07, "loss": 0.2671, "step": 8454 }, { "epoch": 2.7293272529157755, "grad_norm": 0.3193874636756282, "learning_rate": 3.2961627920005485e-07, "loss": 0.248, "step": 8455 }, { "epoch": 2.7296501069453973, "grad_norm": 0.46127166394812497, "learning_rate": 3.288340051980632e-07, "loss": 0.2886, "step": 8456 }, { "epoch": 2.729972960975019, "grad_norm": 0.3339454931279551, "learning_rate": 3.2805263977292775e-07, "loss": 0.2504, "step": 8457 }, { "epoch": 2.730295815004641, "grad_norm": 0.36513877582765936, "learning_rate": 3.272721830236461e-07, "loss": 0.2495, "step": 8458 }, { "epoch": 2.730618669034263, "grad_norm": 0.3591983109094419, "learning_rate": 3.264926350491032e-07, "loss": 0.221, "step": 8459 }, { "epoch": 2.7309415230638847, "grad_norm": 0.36401998071007435, "learning_rate": 3.257139959480626e-07, "loss": 0.2566, "step": 8460 }, { "epoch": 2.7312643770935066, "grad_norm": 0.41007800093051683, "learning_rate": 3.249362658191796e-07, "loss": 0.2341, "step": 8461 }, { "epoch": 2.7315872311231284, "grad_norm": 0.36787971753854437, "learning_rate": 3.241594447609927e-07, "loss": 0.2725, "step": 8462 }, { "epoch": 2.7319100851527502, "grad_norm": 0.3614774548713792, "learning_rate": 3.233835328719209e-07, "loss": 0.3005, "step": 8463 }, { "epoch": 2.732232939182372, "grad_norm": 0.40551617340008955, "learning_rate": 3.2260853025027374e-07, "loss": 0.2548, "step": 8464 }, { "epoch": 2.732555793211994, "grad_norm": 0.3050758665746197, "learning_rate": 3.218344369942411e-07, "loss": 0.2691, "step": 8465 }, { "epoch": 2.7328786472416158, "grad_norm": 0.39291128314986073, "learning_rate": 3.210612532019011e-07, "loss": 0.278, "step": 8466 }, { "epoch": 2.7332015012712376, "grad_norm": 0.36507593278759637, "learning_rate": 3.2028897897121297e-07, "loss": 0.2796, "step": 8467 }, { "epoch": 2.7335243553008595, "grad_norm": 0.3655167149364357, "learning_rate": 3.1951761440002583e-07, "loss": 0.2854, "step": 8468 }, { "epoch": 2.7338472093304813, "grad_norm": 0.3733768703718098, "learning_rate": 3.187471595860675e-07, "loss": 0.2351, "step": 8469 }, { "epoch": 2.734170063360103, "grad_norm": 0.34627795597552896, "learning_rate": 3.1797761462695475e-07, "loss": 0.2524, "step": 8470 }, { "epoch": 2.7344929173897254, "grad_norm": 0.32258199451092734, "learning_rate": 3.172089796201888e-07, "loss": 0.2536, "step": 8471 }, { "epoch": 2.7348157714193473, "grad_norm": 0.3560916016882395, "learning_rate": 3.1644125466315426e-07, "loss": 0.2439, "step": 8472 }, { "epoch": 2.735138625448969, "grad_norm": 0.36926181421039705, "learning_rate": 3.1567443985311917e-07, "loss": 0.2863, "step": 8473 }, { "epoch": 2.735461479478591, "grad_norm": 0.4600407629802622, "learning_rate": 3.1490853528724077e-07, "loss": 0.2735, "step": 8474 }, { "epoch": 2.735784333508213, "grad_norm": 0.355074627366165, "learning_rate": 3.141435410625548e-07, "loss": 0.2425, "step": 8475 }, { "epoch": 2.7361071875378347, "grad_norm": 0.40184232663452385, "learning_rate": 3.1337945727598796e-07, "loss": 0.2239, "step": 8476 }, { "epoch": 2.7364300415674565, "grad_norm": 0.36736865081515263, "learning_rate": 3.12616284024346e-07, "loss": 0.2318, "step": 8477 }, { "epoch": 2.7367528955970783, "grad_norm": 0.4182602215287195, "learning_rate": 3.11854021404325e-07, "loss": 0.2612, "step": 8478 }, { "epoch": 2.7370757496267, "grad_norm": 0.363161424013544, "learning_rate": 3.1109266951249936e-07, "loss": 0.2551, "step": 8479 }, { "epoch": 2.737398603656322, "grad_norm": 0.3217260396744401, "learning_rate": 3.1033222844533274e-07, "loss": 0.3151, "step": 8480 }, { "epoch": 2.737721457685944, "grad_norm": 0.4651730700401631, "learning_rate": 3.0957269829917313e-07, "loss": 0.2393, "step": 8481 }, { "epoch": 2.7380443117155657, "grad_norm": 0.43458674481981724, "learning_rate": 3.088140791702501e-07, "loss": 0.2637, "step": 8482 }, { "epoch": 2.7383671657451876, "grad_norm": 0.37306108034782254, "learning_rate": 3.0805637115468107e-07, "loss": 0.2477, "step": 8483 }, { "epoch": 2.7386900197748094, "grad_norm": 0.4789061005714262, "learning_rate": 3.0729957434846496e-07, "loss": 0.2317, "step": 8484 }, { "epoch": 2.7390128738044313, "grad_norm": 0.4010034456741181, "learning_rate": 3.065436888474876e-07, "loss": 0.2654, "step": 8485 }, { "epoch": 2.739335727834053, "grad_norm": 0.3703020710403709, "learning_rate": 3.0578871474751995e-07, "loss": 0.2386, "step": 8486 }, { "epoch": 2.739658581863675, "grad_norm": 0.3583504133642302, "learning_rate": 3.050346521442146e-07, "loss": 0.2657, "step": 8487 }, { "epoch": 2.739981435893297, "grad_norm": 0.36557352194506765, "learning_rate": 3.0428150113311013e-07, "loss": 0.2527, "step": 8488 }, { "epoch": 2.7403042899229186, "grad_norm": 0.4072537336340159, "learning_rate": 3.0352926180963193e-07, "loss": 0.2729, "step": 8489 }, { "epoch": 2.7406271439525405, "grad_norm": 0.35528906526085374, "learning_rate": 3.027779342690845e-07, "loss": 0.2389, "step": 8490 }, { "epoch": 2.7409499979821623, "grad_norm": 0.31463180806570124, "learning_rate": 3.0202751860666343e-07, "loss": 0.2221, "step": 8491 }, { "epoch": 2.741272852011784, "grad_norm": 0.3690415159435754, "learning_rate": 3.012780149174418e-07, "loss": 0.2533, "step": 8492 }, { "epoch": 2.741595706041406, "grad_norm": 0.35518465524411474, "learning_rate": 3.0052942329638374e-07, "loss": 0.257, "step": 8493 }, { "epoch": 2.741918560071028, "grad_norm": 0.37376030874772037, "learning_rate": 2.997817438383324e-07, "loss": 0.2887, "step": 8494 }, { "epoch": 2.7422414141006497, "grad_norm": 0.39116143523846814, "learning_rate": 2.990349766380196e-07, "loss": 0.2841, "step": 8495 }, { "epoch": 2.7425642681302715, "grad_norm": 0.3558478439528097, "learning_rate": 2.982891217900588e-07, "loss": 0.2323, "step": 8496 }, { "epoch": 2.7428871221598934, "grad_norm": 0.35859157722839696, "learning_rate": 2.975441793889494e-07, "loss": 0.2795, "step": 8497 }, { "epoch": 2.7432099761895152, "grad_norm": 0.29972803001481774, "learning_rate": 2.968001495290734e-07, "loss": 0.2273, "step": 8498 }, { "epoch": 2.743532830219137, "grad_norm": 0.396826989963508, "learning_rate": 2.9605703230469877e-07, "loss": 0.2473, "step": 8499 }, { "epoch": 2.743855684248759, "grad_norm": 0.40032655416895496, "learning_rate": 2.953148278099785e-07, "loss": 0.2379, "step": 8500 }, { "epoch": 2.7441785382783808, "grad_norm": 0.4942067779433705, "learning_rate": 2.945735361389473e-07, "loss": 0.2196, "step": 8501 }, { "epoch": 2.7445013923080026, "grad_norm": 0.426432397754376, "learning_rate": 2.9383315738552683e-07, "loss": 0.2496, "step": 8502 }, { "epoch": 2.7448242463376245, "grad_norm": 0.4198247093520912, "learning_rate": 2.930936916435212e-07, "loss": 0.2529, "step": 8503 }, { "epoch": 2.7451471003672463, "grad_norm": 0.4521176510007039, "learning_rate": 2.9235513900662205e-07, "loss": 0.2675, "step": 8504 }, { "epoch": 2.745469954396868, "grad_norm": 0.3564821777009923, "learning_rate": 2.9161749956839955e-07, "loss": 0.268, "step": 8505 }, { "epoch": 2.74579280842649, "grad_norm": 0.4542431766274777, "learning_rate": 2.908807734223148e-07, "loss": 0.2323, "step": 8506 }, { "epoch": 2.746115662456112, "grad_norm": 0.4385415065369198, "learning_rate": 2.9014496066170735e-07, "loss": 0.2903, "step": 8507 }, { "epoch": 2.7464385164857337, "grad_norm": 0.3725304586209191, "learning_rate": 2.894100613798059e-07, "loss": 0.2689, "step": 8508 }, { "epoch": 2.7467613705153555, "grad_norm": 0.38233391536507716, "learning_rate": 2.886760756697193e-07, "loss": 0.2796, "step": 8509 }, { "epoch": 2.7470842245449774, "grad_norm": 0.3683684478963732, "learning_rate": 2.8794300362444403e-07, "loss": 0.2594, "step": 8510 }, { "epoch": 2.747407078574599, "grad_norm": 0.3743931542163407, "learning_rate": 2.8721084533685826e-07, "loss": 0.2418, "step": 8511 }, { "epoch": 2.747729932604221, "grad_norm": 0.38232691035619887, "learning_rate": 2.864796008997261e-07, "loss": 0.2398, "step": 8512 }, { "epoch": 2.7480527866338433, "grad_norm": 0.3355326272648059, "learning_rate": 2.857492704056944e-07, "loss": 0.2394, "step": 8513 }, { "epoch": 2.748375640663465, "grad_norm": 0.32999662266107405, "learning_rate": 2.8501985394729564e-07, "loss": 0.2361, "step": 8514 }, { "epoch": 2.748698494693087, "grad_norm": 0.40595304015518846, "learning_rate": 2.8429135161694597e-07, "loss": 0.261, "step": 8515 }, { "epoch": 2.749021348722709, "grad_norm": 0.3407468889573092, "learning_rate": 2.8356376350694406e-07, "loss": 0.2707, "step": 8516 }, { "epoch": 2.7493442027523307, "grad_norm": 0.4158774250604009, "learning_rate": 2.82837089709477e-07, "loss": 0.2178, "step": 8517 }, { "epoch": 2.7496670567819526, "grad_norm": 0.3655830705328708, "learning_rate": 2.8211133031661116e-07, "loss": 0.2391, "step": 8518 }, { "epoch": 2.7499899108115744, "grad_norm": 0.3724902785596385, "learning_rate": 2.813864854203013e-07, "loss": 0.293, "step": 8519 }, { "epoch": 2.7503127648411962, "grad_norm": 0.36087734849487346, "learning_rate": 2.806625551123823e-07, "loss": 0.2823, "step": 8520 }, { "epoch": 2.750635618870818, "grad_norm": 0.33287401332102196, "learning_rate": 2.7993953948457746e-07, "loss": 0.2349, "step": 8521 }, { "epoch": 2.75095847290044, "grad_norm": 0.33454591330546835, "learning_rate": 2.792174386284885e-07, "loss": 0.2646, "step": 8522 }, { "epoch": 2.7512813269300618, "grad_norm": 0.3243505373812377, "learning_rate": 2.7849625263560804e-07, "loss": 0.2088, "step": 8523 }, { "epoch": 2.7516041809596836, "grad_norm": 0.4358113241606935, "learning_rate": 2.777759815973063e-07, "loss": 0.2915, "step": 8524 }, { "epoch": 2.7519270349893055, "grad_norm": 0.4262108478775512, "learning_rate": 2.770566256048429e-07, "loss": 0.2384, "step": 8525 }, { "epoch": 2.7522498890189273, "grad_norm": 0.39072276624395763, "learning_rate": 2.7633818474935817e-07, "loss": 0.2921, "step": 8526 }, { "epoch": 2.752572743048549, "grad_norm": 0.3870748981594585, "learning_rate": 2.756206591218785e-07, "loss": 0.2753, "step": 8527 }, { "epoch": 2.752895597078171, "grad_norm": 0.3744211474715083, "learning_rate": 2.74904048813312e-07, "loss": 0.2485, "step": 8528 }, { "epoch": 2.753218451107793, "grad_norm": 0.3137863086765198, "learning_rate": 2.741883539144535e-07, "loss": 0.2464, "step": 8529 }, { "epoch": 2.7535413051374147, "grad_norm": 0.35444680210619217, "learning_rate": 2.734735745159789e-07, "loss": 0.2822, "step": 8530 }, { "epoch": 2.7538641591670365, "grad_norm": 0.36080308505673636, "learning_rate": 2.7275971070845144e-07, "loss": 0.2699, "step": 8531 }, { "epoch": 2.7541870131966584, "grad_norm": 0.4415543899829759, "learning_rate": 2.7204676258231727e-07, "loss": 0.2522, "step": 8532 }, { "epoch": 2.75450986722628, "grad_norm": 0.33870174184188084, "learning_rate": 2.7133473022790315e-07, "loss": 0.2772, "step": 8533 }, { "epoch": 2.754832721255902, "grad_norm": 0.3567753226425624, "learning_rate": 2.7062361373542455e-07, "loss": 0.2725, "step": 8534 }, { "epoch": 2.755155575285524, "grad_norm": 0.3660957815071949, "learning_rate": 2.6991341319497847e-07, "loss": 0.2589, "step": 8535 }, { "epoch": 2.7554784293151457, "grad_norm": 0.3709248761959188, "learning_rate": 2.6920412869654796e-07, "loss": 0.2568, "step": 8536 }, { "epoch": 2.7558012833447676, "grad_norm": 0.366875784167158, "learning_rate": 2.6849576032999533e-07, "loss": 0.2796, "step": 8537 }, { "epoch": 2.7561241373743894, "grad_norm": 0.37019830711500146, "learning_rate": 2.6778830818507293e-07, "loss": 0.2697, "step": 8538 }, { "epoch": 2.7564469914040117, "grad_norm": 0.35689281148616897, "learning_rate": 2.670817723514116e-07, "loss": 0.2434, "step": 8539 }, { "epoch": 2.7567698454336336, "grad_norm": 0.3958414805456983, "learning_rate": 2.6637615291853047e-07, "loss": 0.2431, "step": 8540 }, { "epoch": 2.7570926994632554, "grad_norm": 0.4202393666208984, "learning_rate": 2.656714499758281e-07, "loss": 0.2396, "step": 8541 }, { "epoch": 2.7574155534928773, "grad_norm": 0.37401297181116416, "learning_rate": 2.6496766361259224e-07, "loss": 0.2362, "step": 8542 }, { "epoch": 2.757738407522499, "grad_norm": 0.4394809305762297, "learning_rate": 2.642647939179882e-07, "loss": 0.2796, "step": 8543 }, { "epoch": 2.758061261552121, "grad_norm": 0.44079903340519566, "learning_rate": 2.6356284098107225e-07, "loss": 0.2262, "step": 8544 }, { "epoch": 2.758384115581743, "grad_norm": 0.3861646927819888, "learning_rate": 2.6286180489077827e-07, "loss": 0.2616, "step": 8545 }, { "epoch": 2.7587069696113646, "grad_norm": 0.7079335173044362, "learning_rate": 2.621616857359285e-07, "loss": 0.2724, "step": 8546 }, { "epoch": 2.7590298236409865, "grad_norm": 0.3752154819154828, "learning_rate": 2.6146248360522454e-07, "loss": 0.2369, "step": 8547 }, { "epoch": 2.7593526776706083, "grad_norm": 0.428981750558845, "learning_rate": 2.6076419858725635e-07, "loss": 0.2752, "step": 8548 }, { "epoch": 2.75967553170023, "grad_norm": 0.4024783826783202, "learning_rate": 2.600668307704948e-07, "loss": 0.25, "step": 8549 }, { "epoch": 2.759998385729852, "grad_norm": 0.42218488812255217, "learning_rate": 2.593703802432976e-07, "loss": 0.2794, "step": 8550 }, { "epoch": 2.760321239759474, "grad_norm": 0.3693019880188706, "learning_rate": 2.5867484709390073e-07, "loss": 0.2697, "step": 8551 }, { "epoch": 2.7606440937890957, "grad_norm": 0.4141846838363174, "learning_rate": 2.579802314104296e-07, "loss": 0.2481, "step": 8552 }, { "epoch": 2.7609669478187175, "grad_norm": 0.4068476086519833, "learning_rate": 2.5728653328088964e-07, "loss": 0.2215, "step": 8553 }, { "epoch": 2.7612898018483394, "grad_norm": 0.4104975024188075, "learning_rate": 2.565937527931722e-07, "loss": 0.2037, "step": 8554 }, { "epoch": 2.7616126558779612, "grad_norm": 0.4020187297627421, "learning_rate": 2.559018900350521e-07, "loss": 0.2689, "step": 8555 }, { "epoch": 2.761935509907583, "grad_norm": 0.3652089417435037, "learning_rate": 2.5521094509418584e-07, "loss": 0.2414, "step": 8556 }, { "epoch": 2.762258363937205, "grad_norm": 0.4198278920178834, "learning_rate": 2.545209180581176e-07, "loss": 0.1854, "step": 8557 }, { "epoch": 2.7625812179668268, "grad_norm": 0.34935632463053135, "learning_rate": 2.5383180901426995e-07, "loss": 0.2665, "step": 8558 }, { "epoch": 2.7629040719964486, "grad_norm": 0.40416342477220885, "learning_rate": 2.5314361804995467e-07, "loss": 0.2906, "step": 8559 }, { "epoch": 2.7632269260260705, "grad_norm": 0.3850272372371294, "learning_rate": 2.524563452523629e-07, "loss": 0.1965, "step": 8560 }, { "epoch": 2.7635497800556923, "grad_norm": 0.3629235467599107, "learning_rate": 2.517699907085716e-07, "loss": 0.2752, "step": 8561 }, { "epoch": 2.763872634085314, "grad_norm": 0.2997345328627433, "learning_rate": 2.5108455450554124e-07, "loss": 0.2401, "step": 8562 }, { "epoch": 2.764195488114936, "grad_norm": 0.36242442182024376, "learning_rate": 2.5040003673011486e-07, "loss": 0.229, "step": 8563 }, { "epoch": 2.764518342144558, "grad_norm": 0.3947612287092326, "learning_rate": 2.4971643746902055e-07, "loss": 0.2552, "step": 8564 }, { "epoch": 2.7648411961741797, "grad_norm": 0.3900445584881216, "learning_rate": 2.4903375680887074e-07, "loss": 0.2722, "step": 8565 }, { "epoch": 2.7651640502038015, "grad_norm": 0.3012016193914732, "learning_rate": 2.483519948361579e-07, "loss": 0.2438, "step": 8566 }, { "epoch": 2.7654869042334234, "grad_norm": 0.3409945397022553, "learning_rate": 2.4767115163726205e-07, "loss": 0.2689, "step": 8567 }, { "epoch": 2.765809758263045, "grad_norm": 0.3863939643338846, "learning_rate": 2.469912272984426e-07, "loss": 0.2369, "step": 8568 }, { "epoch": 2.766132612292667, "grad_norm": 0.4052953240087219, "learning_rate": 2.4631222190584805e-07, "loss": 0.2424, "step": 8569 }, { "epoch": 2.766455466322289, "grad_norm": 0.4156057666195063, "learning_rate": 2.456341355455055e-07, "loss": 0.2687, "step": 8570 }, { "epoch": 2.7667783203519107, "grad_norm": 0.3679690108521547, "learning_rate": 2.4495696830332785e-07, "loss": 0.2654, "step": 8571 }, { "epoch": 2.7671011743815326, "grad_norm": 0.3561771158948616, "learning_rate": 2.4428072026511234e-07, "loss": 0.2448, "step": 8572 }, { "epoch": 2.7674240284111544, "grad_norm": 0.37847751416121034, "learning_rate": 2.4360539151653705e-07, "loss": 0.2252, "step": 8573 }, { "epoch": 2.7677468824407763, "grad_norm": 0.403635260796568, "learning_rate": 2.4293098214316687e-07, "loss": 0.2472, "step": 8574 }, { "epoch": 2.768069736470398, "grad_norm": 0.358623644828904, "learning_rate": 2.4225749223044686e-07, "loss": 0.2879, "step": 8575 }, { "epoch": 2.76839259050002, "grad_norm": 0.4042503712235711, "learning_rate": 2.415849218637087e-07, "loss": 0.2258, "step": 8576 }, { "epoch": 2.768715444529642, "grad_norm": 0.44239723844897205, "learning_rate": 2.4091327112816504e-07, "loss": 0.2617, "step": 8577 }, { "epoch": 2.7690382985592636, "grad_norm": 0.34757927575776726, "learning_rate": 2.4024254010891366e-07, "loss": 0.2496, "step": 8578 }, { "epoch": 2.7693611525888855, "grad_norm": 0.36754614379284467, "learning_rate": 2.3957272889093495e-07, "loss": 0.3055, "step": 8579 }, { "epoch": 2.7696840066185073, "grad_norm": 0.322770804468086, "learning_rate": 2.389038375590935e-07, "loss": 0.2601, "step": 8580 }, { "epoch": 2.7700068606481296, "grad_norm": 0.4417767980722774, "learning_rate": 2.382358661981357e-07, "loss": 0.2917, "step": 8581 }, { "epoch": 2.7703297146777515, "grad_norm": 0.30661363846375994, "learning_rate": 2.375688148926955e-07, "loss": 0.2579, "step": 8582 }, { "epoch": 2.7706525687073733, "grad_norm": 0.34606125106200497, "learning_rate": 2.3690268372728446e-07, "loss": 0.2656, "step": 8583 }, { "epoch": 2.770975422736995, "grad_norm": 0.3628123251786504, "learning_rate": 2.3623747278630176e-07, "loss": 0.2071, "step": 8584 }, { "epoch": 2.771298276766617, "grad_norm": 0.3891532587422681, "learning_rate": 2.355731821540283e-07, "loss": 0.2453, "step": 8585 }, { "epoch": 2.771621130796239, "grad_norm": 0.42419892092752526, "learning_rate": 2.3490981191462923e-07, "loss": 0.2642, "step": 8586 }, { "epoch": 2.7719439848258607, "grad_norm": 0.4698895063872318, "learning_rate": 2.3424736215215238e-07, "loss": 0.3003, "step": 8587 }, { "epoch": 2.7722668388554825, "grad_norm": 0.35981488128204314, "learning_rate": 2.3358583295052887e-07, "loss": 0.2732, "step": 8588 }, { "epoch": 2.7725896928851044, "grad_norm": 0.34683023576016164, "learning_rate": 2.3292522439357422e-07, "loss": 0.2601, "step": 8589 }, { "epoch": 2.772912546914726, "grad_norm": 0.43784121992260816, "learning_rate": 2.3226553656498562e-07, "loss": 0.255, "step": 8590 }, { "epoch": 2.773235400944348, "grad_norm": 0.3585706808349533, "learning_rate": 2.316067695483462e-07, "loss": 0.2478, "step": 8591 }, { "epoch": 2.77355825497397, "grad_norm": 0.32216198201341906, "learning_rate": 2.3094892342711922e-07, "loss": 0.2678, "step": 8592 }, { "epoch": 2.7738811090035917, "grad_norm": 0.3820360607112315, "learning_rate": 2.3029199828465381e-07, "loss": 0.2215, "step": 8593 }, { "epoch": 2.7742039630332136, "grad_norm": 0.3905310760311807, "learning_rate": 2.296359942041809e-07, "loss": 0.2814, "step": 8594 }, { "epoch": 2.7745268170628354, "grad_norm": 0.32106160855041516, "learning_rate": 2.289809112688157e-07, "loss": 0.2778, "step": 8595 }, { "epoch": 2.7748496710924573, "grad_norm": 0.36516590005457844, "learning_rate": 2.2832674956155597e-07, "loss": 0.2669, "step": 8596 }, { "epoch": 2.775172525122079, "grad_norm": 0.38622816972808044, "learning_rate": 2.2767350916528455e-07, "loss": 0.2689, "step": 8597 }, { "epoch": 2.775495379151701, "grad_norm": 0.3512863683658451, "learning_rate": 2.2702119016276363e-07, "loss": 0.1867, "step": 8598 }, { "epoch": 2.775818233181323, "grad_norm": 0.3628881119772211, "learning_rate": 2.263697926366437e-07, "loss": 0.2575, "step": 8599 }, { "epoch": 2.7761410872109447, "grad_norm": 0.4303684357435603, "learning_rate": 2.2571931666945378e-07, "loss": 0.2689, "step": 8600 }, { "epoch": 2.7764639412405665, "grad_norm": 0.388705785838432, "learning_rate": 2.250697623436096e-07, "loss": 0.2866, "step": 8601 }, { "epoch": 2.7767867952701883, "grad_norm": 0.3955574954382923, "learning_rate": 2.2442112974140782e-07, "loss": 0.2323, "step": 8602 }, { "epoch": 2.77710964929981, "grad_norm": 0.39654894739349, "learning_rate": 2.2377341894503107e-07, "loss": 0.2744, "step": 8603 }, { "epoch": 2.777432503329432, "grad_norm": 0.36303470794442877, "learning_rate": 2.2312663003654033e-07, "loss": 0.2412, "step": 8604 }, { "epoch": 2.777755357359054, "grad_norm": 0.3816547609947303, "learning_rate": 2.2248076309788673e-07, "loss": 0.243, "step": 8605 }, { "epoch": 2.778078211388676, "grad_norm": 0.38769247601396445, "learning_rate": 2.218358182108973e-07, "loss": 0.2505, "step": 8606 }, { "epoch": 2.778401065418298, "grad_norm": 0.38454427265399477, "learning_rate": 2.2119179545728758e-07, "loss": 0.2065, "step": 8607 }, { "epoch": 2.77872391944792, "grad_norm": 0.3688837228634934, "learning_rate": 2.2054869491865305e-07, "loss": 0.2413, "step": 8608 }, { "epoch": 2.7790467734775417, "grad_norm": 0.35910957971735263, "learning_rate": 2.1990651667647522e-07, "loss": 0.226, "step": 8609 }, { "epoch": 2.7793696275071635, "grad_norm": 0.38632885992273663, "learning_rate": 2.1926526081211652e-07, "loss": 0.2625, "step": 8610 }, { "epoch": 2.7796924815367854, "grad_norm": 0.3525892609749948, "learning_rate": 2.1862492740682276e-07, "loss": 0.2887, "step": 8611 }, { "epoch": 2.7800153355664072, "grad_norm": 0.3735298077783849, "learning_rate": 2.1798551654172328e-07, "loss": 0.2558, "step": 8612 }, { "epoch": 2.780338189596029, "grad_norm": 0.4313319132745871, "learning_rate": 2.1734702829783154e-07, "loss": 0.2827, "step": 8613 }, { "epoch": 2.780661043625651, "grad_norm": 0.39330644680090177, "learning_rate": 2.1670946275604286e-07, "loss": 0.2875, "step": 8614 }, { "epoch": 2.7809838976552728, "grad_norm": 0.38416861277888353, "learning_rate": 2.160728199971343e-07, "loss": 0.3448, "step": 8615 }, { "epoch": 2.7813067516848946, "grad_norm": 0.3929897855814226, "learning_rate": 2.1543710010177047e-07, "loss": 0.2951, "step": 8616 }, { "epoch": 2.7816296057145165, "grad_norm": 0.3841690419509959, "learning_rate": 2.1480230315049365e-07, "loss": 0.2833, "step": 8617 }, { "epoch": 2.7819524597441383, "grad_norm": 0.35320720822542456, "learning_rate": 2.141684292237328e-07, "loss": 0.2725, "step": 8618 }, { "epoch": 2.78227531377376, "grad_norm": 0.37393130630380683, "learning_rate": 2.135354784017987e-07, "loss": 0.222, "step": 8619 }, { "epoch": 2.782598167803382, "grad_norm": 0.337024035809972, "learning_rate": 2.1290345076488637e-07, "loss": 0.2292, "step": 8620 }, { "epoch": 2.782921021833004, "grad_norm": 0.4262259584142916, "learning_rate": 2.1227234639307093e-07, "loss": 0.2709, "step": 8621 }, { "epoch": 2.7832438758626257, "grad_norm": 0.39919238652064504, "learning_rate": 2.1164216536631337e-07, "loss": 0.3153, "step": 8622 }, { "epoch": 2.7835667298922475, "grad_norm": 0.39347127911204893, "learning_rate": 2.1101290776445736e-07, "loss": 0.2584, "step": 8623 }, { "epoch": 2.7838895839218694, "grad_norm": 0.4049795910964727, "learning_rate": 2.103845736672283e-07, "loss": 0.2696, "step": 8624 }, { "epoch": 2.784212437951491, "grad_norm": 0.37780649969582913, "learning_rate": 2.0975716315423498e-07, "loss": 0.269, "step": 8625 }, { "epoch": 2.784535291981113, "grad_norm": 0.4649933190262639, "learning_rate": 2.0913067630496885e-07, "loss": 0.2374, "step": 8626 }, { "epoch": 2.784858146010735, "grad_norm": 0.36580932459465965, "learning_rate": 2.0850511319880722e-07, "loss": 0.2034, "step": 8627 }, { "epoch": 2.7851810000403567, "grad_norm": 0.37085140453940657, "learning_rate": 2.0788047391500586e-07, "loss": 0.2816, "step": 8628 }, { "epoch": 2.7855038540699786, "grad_norm": 0.4005679309464883, "learning_rate": 2.0725675853270814e-07, "loss": 0.2577, "step": 8629 }, { "epoch": 2.7858267080996004, "grad_norm": 0.4316578844529304, "learning_rate": 2.0663396713093502e-07, "loss": 0.327, "step": 8630 }, { "epoch": 2.7861495621292223, "grad_norm": 0.37793354049881256, "learning_rate": 2.0601209978859587e-07, "loss": 0.3106, "step": 8631 }, { "epoch": 2.786472416158844, "grad_norm": 0.3967514372983697, "learning_rate": 2.053911565844785e-07, "loss": 0.2525, "step": 8632 }, { "epoch": 2.786795270188466, "grad_norm": 0.34694818365649055, "learning_rate": 2.0477113759725664e-07, "loss": 0.2478, "step": 8633 }, { "epoch": 2.787118124218088, "grad_norm": 0.3885445350838399, "learning_rate": 2.0415204290548495e-07, "loss": 0.2376, "step": 8634 }, { "epoch": 2.7874409782477096, "grad_norm": 0.40540663597545057, "learning_rate": 2.035338725876032e-07, "loss": 0.3148, "step": 8635 }, { "epoch": 2.7877638322773315, "grad_norm": 0.3835530743989045, "learning_rate": 2.0291662672193122e-07, "loss": 0.2593, "step": 8636 }, { "epoch": 2.7880866863069533, "grad_norm": 0.30348322927560173, "learning_rate": 2.0230030538667483e-07, "loss": 0.2279, "step": 8637 }, { "epoch": 2.788409540336575, "grad_norm": 0.4537425450485844, "learning_rate": 2.0168490865991985e-07, "loss": 0.2697, "step": 8638 }, { "epoch": 2.788732394366197, "grad_norm": 0.4108074655749082, "learning_rate": 2.0107043661963642e-07, "loss": 0.2755, "step": 8639 }, { "epoch": 2.789055248395819, "grad_norm": 0.352959752500209, "learning_rate": 2.004568893436773e-07, "loss": 0.2913, "step": 8640 }, { "epoch": 2.7893781024254407, "grad_norm": 0.3920059371991167, "learning_rate": 1.9984426690977774e-07, "loss": 0.2329, "step": 8641 }, { "epoch": 2.7897009564550626, "grad_norm": 0.42391862258244734, "learning_rate": 1.992325693955574e-07, "loss": 0.2396, "step": 8642 }, { "epoch": 2.7900238104846844, "grad_norm": 0.4627504924662782, "learning_rate": 1.9862179687851673e-07, "loss": 0.2406, "step": 8643 }, { "epoch": 2.7903466645143062, "grad_norm": 0.3851950584915443, "learning_rate": 1.9801194943603968e-07, "loss": 0.2606, "step": 8644 }, { "epoch": 2.790669518543928, "grad_norm": 0.4145242824771128, "learning_rate": 1.9740302714539276e-07, "loss": 0.2608, "step": 8645 }, { "epoch": 2.79099237257355, "grad_norm": 0.3672204597729597, "learning_rate": 1.9679503008372678e-07, "loss": 0.2107, "step": 8646 }, { "epoch": 2.7913152266031718, "grad_norm": 0.39253808497026615, "learning_rate": 1.9618795832807345e-07, "loss": 0.2608, "step": 8647 }, { "epoch": 2.791638080632794, "grad_norm": 0.3875554249855679, "learning_rate": 1.955818119553479e-07, "loss": 0.2085, "step": 8648 }, { "epoch": 2.791960934662416, "grad_norm": 0.3668993711750195, "learning_rate": 1.9497659104234705e-07, "loss": 0.2954, "step": 8649 }, { "epoch": 2.7922837886920377, "grad_norm": 0.3530459041551228, "learning_rate": 1.943722956657537e-07, "loss": 0.2283, "step": 8650 }, { "epoch": 2.7926066427216596, "grad_norm": 0.3582503693726849, "learning_rate": 1.9376892590212914e-07, "loss": 0.2462, "step": 8651 }, { "epoch": 2.7929294967512814, "grad_norm": 0.3754921261994573, "learning_rate": 1.9316648182792052e-07, "loss": 0.2559, "step": 8652 }, { "epoch": 2.7932523507809033, "grad_norm": 0.3702979528783443, "learning_rate": 1.925649635194568e-07, "loss": 0.2142, "step": 8653 }, { "epoch": 2.793575204810525, "grad_norm": 0.4057249073179477, "learning_rate": 1.9196437105294952e-07, "loss": 0.2513, "step": 8654 }, { "epoch": 2.793898058840147, "grad_norm": 0.376907669484673, "learning_rate": 1.9136470450449195e-07, "loss": 0.2468, "step": 8655 }, { "epoch": 2.794220912869769, "grad_norm": 0.36672780745178346, "learning_rate": 1.9076596395006251e-07, "loss": 0.2434, "step": 8656 }, { "epoch": 2.7945437668993907, "grad_norm": 0.37754236779888345, "learning_rate": 1.9016814946551885e-07, "loss": 0.2217, "step": 8657 }, { "epoch": 2.7948666209290125, "grad_norm": 0.35660935312498104, "learning_rate": 1.895712611266054e-07, "loss": 0.2516, "step": 8658 }, { "epoch": 2.7951894749586343, "grad_norm": 0.33609820269133517, "learning_rate": 1.8897529900894494e-07, "loss": 0.225, "step": 8659 }, { "epoch": 2.795512328988256, "grad_norm": 0.4056300046270629, "learning_rate": 1.8838026318804796e-07, "loss": 0.2989, "step": 8660 }, { "epoch": 2.795835183017878, "grad_norm": 0.3630435095972934, "learning_rate": 1.8778615373930164e-07, "loss": 0.2681, "step": 8661 }, { "epoch": 2.7961580370475, "grad_norm": 0.48369851411580955, "learning_rate": 1.8719297073798076e-07, "loss": 0.2686, "step": 8662 }, { "epoch": 2.7964808910771217, "grad_norm": 0.3695014831819619, "learning_rate": 1.8660071425923937e-07, "loss": 0.3154, "step": 8663 }, { "epoch": 2.7968037451067436, "grad_norm": 0.32176523264974505, "learning_rate": 1.860093843781166e-07, "loss": 0.258, "step": 8664 }, { "epoch": 2.7971265991363654, "grad_norm": 0.3318460449548335, "learning_rate": 1.854189811695342e-07, "loss": 0.2232, "step": 8665 }, { "epoch": 2.7974494531659873, "grad_norm": 0.4016629598356086, "learning_rate": 1.8482950470829236e-07, "loss": 0.247, "step": 8666 }, { "epoch": 2.797772307195609, "grad_norm": 0.3898242943595096, "learning_rate": 1.8424095506908044e-07, "loss": 0.2538, "step": 8667 }, { "epoch": 2.798095161225231, "grad_norm": 0.35281205120571996, "learning_rate": 1.8365333232646304e-07, "loss": 0.2635, "step": 8668 }, { "epoch": 2.798418015254853, "grad_norm": 0.3405887267546128, "learning_rate": 1.8306663655489468e-07, "loss": 0.2381, "step": 8669 }, { "epoch": 2.7987408692844746, "grad_norm": 0.3030888762585204, "learning_rate": 1.8248086782870683e-07, "loss": 0.2296, "step": 8670 }, { "epoch": 2.7990637233140965, "grad_norm": 0.35774567338392926, "learning_rate": 1.818960262221167e-07, "loss": 0.2099, "step": 8671 }, { "epoch": 2.7993865773437183, "grad_norm": 0.3245925228640496, "learning_rate": 1.8131211180922257e-07, "loss": 0.2821, "step": 8672 }, { "epoch": 2.79970943137334, "grad_norm": 0.36569640316576185, "learning_rate": 1.8072912466400443e-07, "loss": 0.2886, "step": 8673 }, { "epoch": 2.8000322854029625, "grad_norm": 0.37712106806842816, "learning_rate": 1.8014706486032734e-07, "loss": 0.2623, "step": 8674 }, { "epoch": 2.8003551394325843, "grad_norm": 0.3677435396646306, "learning_rate": 1.7956593247193732e-07, "loss": 0.2933, "step": 8675 }, { "epoch": 2.800677993462206, "grad_norm": 0.4343776341222303, "learning_rate": 1.7898572757246296e-07, "loss": 0.2532, "step": 8676 }, { "epoch": 2.801000847491828, "grad_norm": 0.36924671983229623, "learning_rate": 1.7840645023541545e-07, "loss": 0.2794, "step": 8677 }, { "epoch": 2.80132370152145, "grad_norm": 0.41963855931081345, "learning_rate": 1.7782810053418695e-07, "loss": 0.2453, "step": 8678 }, { "epoch": 2.8016465555510717, "grad_norm": 0.39819402735715553, "learning_rate": 1.7725067854205635e-07, "loss": 0.225, "step": 8679 }, { "epoch": 2.8019694095806935, "grad_norm": 0.4033918289022958, "learning_rate": 1.7667418433218008e-07, "loss": 0.2557, "step": 8680 }, { "epoch": 2.8022922636103154, "grad_norm": 0.44294506354740965, "learning_rate": 1.7609861797760057e-07, "loss": 0.2724, "step": 8681 }, { "epoch": 2.802615117639937, "grad_norm": 0.3434758456880629, "learning_rate": 1.755239795512395e-07, "loss": 0.2435, "step": 8682 }, { "epoch": 2.802937971669559, "grad_norm": 0.36346654760989844, "learning_rate": 1.7495026912590444e-07, "loss": 0.2428, "step": 8683 }, { "epoch": 2.803260825699181, "grad_norm": 0.5108436749924621, "learning_rate": 1.743774867742831e-07, "loss": 0.2363, "step": 8684 }, { "epoch": 2.8035836797288027, "grad_norm": 0.32393292843836613, "learning_rate": 1.7380563256894576e-07, "loss": 0.2424, "step": 8685 }, { "epoch": 2.8039065337584246, "grad_norm": 0.4069247670324493, "learning_rate": 1.7323470658234698e-07, "loss": 0.2909, "step": 8686 }, { "epoch": 2.8042293877880464, "grad_norm": 0.4438273306649338, "learning_rate": 1.726647088868197e-07, "loss": 0.2608, "step": 8687 }, { "epoch": 2.8045522418176683, "grad_norm": 0.40300241702593576, "learning_rate": 1.7209563955458447e-07, "loss": 0.2612, "step": 8688 }, { "epoch": 2.80487509584729, "grad_norm": 0.38151366682591314, "learning_rate": 1.7152749865774115e-07, "loss": 0.2658, "step": 8689 }, { "epoch": 2.805197949876912, "grad_norm": 0.35371303453618513, "learning_rate": 1.7096028626827047e-07, "loss": 0.2747, "step": 8690 }, { "epoch": 2.805520803906534, "grad_norm": 0.3908515098599415, "learning_rate": 1.7039400245803915e-07, "loss": 0.3157, "step": 8691 }, { "epoch": 2.8058436579361556, "grad_norm": 0.4084929870974385, "learning_rate": 1.698286472987956e-07, "loss": 0.2772, "step": 8692 }, { "epoch": 2.8061665119657775, "grad_norm": 0.36985234285726093, "learning_rate": 1.692642208621667e-07, "loss": 0.3046, "step": 8693 }, { "epoch": 2.8064893659953993, "grad_norm": 0.38936489547586745, "learning_rate": 1.687007232196669e-07, "loss": 0.3053, "step": 8694 }, { "epoch": 2.806812220025021, "grad_norm": 0.3979479045951246, "learning_rate": 1.681381544426891e-07, "loss": 0.2288, "step": 8695 }, { "epoch": 2.807135074054643, "grad_norm": 0.38810761002473076, "learning_rate": 1.6757651460251127e-07, "loss": 0.2503, "step": 8696 }, { "epoch": 2.807457928084265, "grad_norm": 0.415450715389038, "learning_rate": 1.670158037702907e-07, "loss": 0.2085, "step": 8697 }, { "epoch": 2.8077807821138867, "grad_norm": 0.3213122827367995, "learning_rate": 1.6645602201707049e-07, "loss": 0.264, "step": 8698 }, { "epoch": 2.8081036361435086, "grad_norm": 0.3783494562226556, "learning_rate": 1.6589716941377313e-07, "loss": 0.2182, "step": 8699 }, { "epoch": 2.8084264901731304, "grad_norm": 0.3631091578217331, "learning_rate": 1.6533924603120447e-07, "loss": 0.2005, "step": 8700 }, { "epoch": 2.8087493442027522, "grad_norm": 0.35458463966650655, "learning_rate": 1.6478225194005298e-07, "loss": 0.2697, "step": 8701 }, { "epoch": 2.809072198232374, "grad_norm": 0.34578832589056424, "learning_rate": 1.6422618721088883e-07, "loss": 0.2473, "step": 8702 }, { "epoch": 2.809395052261996, "grad_norm": 0.3589521265116951, "learning_rate": 1.636710519141657e-07, "loss": 0.2507, "step": 8703 }, { "epoch": 2.8097179062916178, "grad_norm": 0.4114311972340948, "learning_rate": 1.631168461202165e-07, "loss": 0.2319, "step": 8704 }, { "epoch": 2.8100407603212396, "grad_norm": 0.4970348407723204, "learning_rate": 1.6256356989926002e-07, "loss": 0.2886, "step": 8705 }, { "epoch": 2.8103636143508615, "grad_norm": 0.3648974367718549, "learning_rate": 1.620112233213944e-07, "loss": 0.2799, "step": 8706 }, { "epoch": 2.8106864683804833, "grad_norm": 0.39202568204006083, "learning_rate": 1.6145980645660195e-07, "loss": 0.2777, "step": 8707 }, { "epoch": 2.811009322410105, "grad_norm": 0.42553826164077685, "learning_rate": 1.6090931937474685e-07, "loss": 0.2813, "step": 8708 }, { "epoch": 2.811332176439727, "grad_norm": 0.39400896144337855, "learning_rate": 1.6035976214557407e-07, "loss": 0.2609, "step": 8709 }, { "epoch": 2.811655030469349, "grad_norm": 0.4084137533039829, "learning_rate": 1.5981113483871213e-07, "loss": 0.2917, "step": 8710 }, { "epoch": 2.8119778844989707, "grad_norm": 0.3886794888284465, "learning_rate": 1.5926343752367207e-07, "loss": 0.2866, "step": 8711 }, { "epoch": 2.8123007385285925, "grad_norm": 0.5107193587512181, "learning_rate": 1.587166702698442e-07, "loss": 0.2858, "step": 8712 }, { "epoch": 2.8126235925582144, "grad_norm": 0.3871518395015063, "learning_rate": 1.5817083314650565e-07, "loss": 0.2868, "step": 8713 }, { "epoch": 2.812946446587836, "grad_norm": 0.3681220507168618, "learning_rate": 1.5762592622281192e-07, "loss": 0.2194, "step": 8714 }, { "epoch": 2.813269300617458, "grad_norm": 0.3567026561068728, "learning_rate": 1.5708194956780274e-07, "loss": 0.2158, "step": 8715 }, { "epoch": 2.8135921546470803, "grad_norm": 0.35357332173305717, "learning_rate": 1.5653890325039804e-07, "loss": 0.2685, "step": 8716 }, { "epoch": 2.813915008676702, "grad_norm": 0.404309305306324, "learning_rate": 1.5599678733940275e-07, "loss": 0.2856, "step": 8717 }, { "epoch": 2.814237862706324, "grad_norm": 0.39222065055869765, "learning_rate": 1.554556019035003e-07, "loss": 0.1968, "step": 8718 }, { "epoch": 2.814560716735946, "grad_norm": 0.3841624561754729, "learning_rate": 1.549153470112591e-07, "loss": 0.2482, "step": 8719 }, { "epoch": 2.8148835707655677, "grad_norm": 0.36340915607918306, "learning_rate": 1.5437602273112862e-07, "loss": 0.2569, "step": 8720 }, { "epoch": 2.8152064247951896, "grad_norm": 0.38694337510930255, "learning_rate": 1.5383762913144083e-07, "loss": 0.2422, "step": 8721 }, { "epoch": 2.8155292788248114, "grad_norm": 0.4190267293301178, "learning_rate": 1.5330016628040954e-07, "loss": 0.2705, "step": 8722 }, { "epoch": 2.8158521328544333, "grad_norm": 0.3951300368586001, "learning_rate": 1.5276363424612943e-07, "loss": 0.2178, "step": 8723 }, { "epoch": 2.816174986884055, "grad_norm": 0.3853337709113258, "learning_rate": 1.5222803309657946e-07, "loss": 0.2821, "step": 8724 }, { "epoch": 2.816497840913677, "grad_norm": 0.369822722899836, "learning_rate": 1.5169336289961866e-07, "loss": 0.2702, "step": 8725 }, { "epoch": 2.816820694943299, "grad_norm": 0.4259862194998393, "learning_rate": 1.5115962372299119e-07, "loss": 0.3157, "step": 8726 }, { "epoch": 2.8171435489729206, "grad_norm": 0.370375412307245, "learning_rate": 1.5062681563431797e-07, "loss": 0.2551, "step": 8727 }, { "epoch": 2.8174664030025425, "grad_norm": 0.3398947343947343, "learning_rate": 1.5009493870110747e-07, "loss": 0.2892, "step": 8728 }, { "epoch": 2.8177892570321643, "grad_norm": 0.4376721511196755, "learning_rate": 1.495639929907458e-07, "loss": 0.2463, "step": 8729 }, { "epoch": 2.818112111061786, "grad_norm": 0.36386039786578067, "learning_rate": 1.49033978570505e-07, "loss": 0.3011, "step": 8730 }, { "epoch": 2.818434965091408, "grad_norm": 0.324976361370614, "learning_rate": 1.4850489550753548e-07, "loss": 0.2549, "step": 8731 }, { "epoch": 2.81875781912103, "grad_norm": 0.32254613394260473, "learning_rate": 1.47976743868872e-07, "loss": 0.1967, "step": 8732 }, { "epoch": 2.8190806731506517, "grad_norm": 0.36384251058473116, "learning_rate": 1.47449523721431e-07, "loss": 0.2611, "step": 8733 }, { "epoch": 2.8194035271802735, "grad_norm": 0.4127001445495456, "learning_rate": 1.4692323513201072e-07, "loss": 0.2509, "step": 8734 }, { "epoch": 2.8197263812098954, "grad_norm": 0.3455731692382347, "learning_rate": 1.4639787816729032e-07, "loss": 0.2677, "step": 8735 }, { "epoch": 2.8200492352395172, "grad_norm": 0.33079033453011863, "learning_rate": 1.4587345289383241e-07, "loss": 0.2402, "step": 8736 }, { "epoch": 2.820372089269139, "grad_norm": 0.3473348240476679, "learning_rate": 1.4534995937808049e-07, "loss": 0.2531, "step": 8737 }, { "epoch": 2.820694943298761, "grad_norm": 0.3860327243316952, "learning_rate": 1.4482739768636066e-07, "loss": 0.2638, "step": 8738 }, { "epoch": 2.8210177973283828, "grad_norm": 0.3749121545125715, "learning_rate": 1.443057678848808e-07, "loss": 0.2689, "step": 8739 }, { "epoch": 2.8213406513580046, "grad_norm": 0.3989339722956062, "learning_rate": 1.437850700397314e-07, "loss": 0.3156, "step": 8740 }, { "epoch": 2.821663505387627, "grad_norm": 0.374348130589686, "learning_rate": 1.4326530421688384e-07, "loss": 0.2473, "step": 8741 }, { "epoch": 2.8219863594172487, "grad_norm": 0.3866389768839793, "learning_rate": 1.427464704821904e-07, "loss": 0.2717, "step": 8742 }, { "epoch": 2.8223092134468706, "grad_norm": 0.353141537705785, "learning_rate": 1.4222856890138776e-07, "loss": 0.2517, "step": 8743 }, { "epoch": 2.8226320674764924, "grad_norm": 0.3802228986933802, "learning_rate": 1.4171159954009332e-07, "loss": 0.2139, "step": 8744 }, { "epoch": 2.8229549215061143, "grad_norm": 0.35396432882360634, "learning_rate": 1.4119556246380728e-07, "loss": 0.228, "step": 8745 }, { "epoch": 2.823277775535736, "grad_norm": 0.37761403548756467, "learning_rate": 1.406804577379081e-07, "loss": 0.2987, "step": 8746 }, { "epoch": 2.823600629565358, "grad_norm": 0.40847563576995677, "learning_rate": 1.4016628542766197e-07, "loss": 0.2512, "step": 8747 }, { "epoch": 2.82392348359498, "grad_norm": 0.37348932199852103, "learning_rate": 1.3965304559821174e-07, "loss": 0.2712, "step": 8748 }, { "epoch": 2.8242463376246016, "grad_norm": 0.3552682103476531, "learning_rate": 1.3914073831458456e-07, "loss": 0.2175, "step": 8749 }, { "epoch": 2.8245691916542235, "grad_norm": 0.32847374842642607, "learning_rate": 1.3862936364168938e-07, "loss": 0.2555, "step": 8750 }, { "epoch": 2.8248920456838453, "grad_norm": 0.33870046706928686, "learning_rate": 1.3811892164431684e-07, "loss": 0.2358, "step": 8751 }, { "epoch": 2.825214899713467, "grad_norm": 0.41551036055486973, "learning_rate": 1.3760941238713854e-07, "loss": 0.2551, "step": 8752 }, { "epoch": 2.825537753743089, "grad_norm": 0.3513331458996669, "learning_rate": 1.3710083593471035e-07, "loss": 0.2074, "step": 8753 }, { "epoch": 2.825860607772711, "grad_norm": 0.34866089349206103, "learning_rate": 1.3659319235146573e-07, "loss": 0.2302, "step": 8754 }, { "epoch": 2.8261834618023327, "grad_norm": 0.3719638872343039, "learning_rate": 1.3608648170172406e-07, "loss": 0.2723, "step": 8755 }, { "epoch": 2.8265063158319546, "grad_norm": 0.33463797006563883, "learning_rate": 1.3558070404968397e-07, "loss": 0.2379, "step": 8756 }, { "epoch": 2.8268291698615764, "grad_norm": 0.3407093599737048, "learning_rate": 1.3507585945942758e-07, "loss": 0.2598, "step": 8757 }, { "epoch": 2.8271520238911982, "grad_norm": 0.37971015535008723, "learning_rate": 1.3457194799491867e-07, "loss": 0.2339, "step": 8758 }, { "epoch": 2.82747487792082, "grad_norm": 0.37046186318261126, "learning_rate": 1.3406896971999955e-07, "loss": 0.2699, "step": 8759 }, { "epoch": 2.827797731950442, "grad_norm": 0.4020892991719789, "learning_rate": 1.3356692469840004e-07, "loss": 0.2104, "step": 8760 }, { "epoch": 2.8281205859800638, "grad_norm": 0.3704079008644411, "learning_rate": 1.3306581299372594e-07, "loss": 0.2803, "step": 8761 }, { "epoch": 2.8284434400096856, "grad_norm": 0.36439484155783575, "learning_rate": 1.3256563466946897e-07, "loss": 0.2932, "step": 8762 }, { "epoch": 2.8287662940393075, "grad_norm": 0.32951379541104714, "learning_rate": 1.320663897890001e-07, "loss": 0.2368, "step": 8763 }, { "epoch": 2.8290891480689293, "grad_norm": 0.32601173398740285, "learning_rate": 1.3156807841557374e-07, "loss": 0.213, "step": 8764 }, { "epoch": 2.829412002098551, "grad_norm": 0.38748172719356216, "learning_rate": 1.3107070061232435e-07, "loss": 0.234, "step": 8765 }, { "epoch": 2.829734856128173, "grad_norm": 0.39510565175571677, "learning_rate": 1.3057425644226988e-07, "loss": 0.2484, "step": 8766 }, { "epoch": 2.830057710157795, "grad_norm": 0.3703373768998607, "learning_rate": 1.300787459683092e-07, "loss": 0.3017, "step": 8767 }, { "epoch": 2.8303805641874167, "grad_norm": 0.43385221787679495, "learning_rate": 1.2958416925322125e-07, "loss": 0.2868, "step": 8768 }, { "epoch": 2.8307034182170385, "grad_norm": 0.49448170764606086, "learning_rate": 1.2909052635967007e-07, "loss": 0.2035, "step": 8769 }, { "epoch": 2.8310262722466604, "grad_norm": 0.3454741731279595, "learning_rate": 1.2859781735019976e-07, "loss": 0.2463, "step": 8770 }, { "epoch": 2.831349126276282, "grad_norm": 0.3513453621116421, "learning_rate": 1.2810604228723376e-07, "loss": 0.2161, "step": 8771 }, { "epoch": 2.831671980305904, "grad_norm": 0.4360001464227337, "learning_rate": 1.276152012330814e-07, "loss": 0.3091, "step": 8772 }, { "epoch": 2.831994834335526, "grad_norm": 0.4663350747058632, "learning_rate": 1.2712529424992957e-07, "loss": 0.2695, "step": 8773 }, { "epoch": 2.8323176883651477, "grad_norm": 0.4100956879867727, "learning_rate": 1.2663632139985032e-07, "loss": 0.2461, "step": 8774 }, { "epoch": 2.8326405423947696, "grad_norm": 0.3265182414888957, "learning_rate": 1.2614828274479656e-07, "loss": 0.2727, "step": 8775 }, { "epoch": 2.8329633964243914, "grad_norm": 0.43870262981644953, "learning_rate": 1.2566117834659967e-07, "loss": 0.28, "step": 8776 }, { "epoch": 2.8332862504540133, "grad_norm": 0.40481148221287716, "learning_rate": 1.2517500826697699e-07, "loss": 0.2205, "step": 8777 }, { "epoch": 2.833609104483635, "grad_norm": 0.3935182846953806, "learning_rate": 1.24689772567525e-07, "loss": 0.2261, "step": 8778 }, { "epoch": 2.833931958513257, "grad_norm": 0.3898042276498701, "learning_rate": 1.242054713097221e-07, "loss": 0.3225, "step": 8779 }, { "epoch": 2.834254812542879, "grad_norm": 0.4081692110457609, "learning_rate": 1.2372210455492917e-07, "loss": 0.3105, "step": 8780 }, { "epoch": 2.8345776665725007, "grad_norm": 0.34379291599247747, "learning_rate": 1.2323967236438805e-07, "loss": 0.275, "step": 8781 }, { "epoch": 2.8349005206021225, "grad_norm": 0.3279828460690989, "learning_rate": 1.2275817479922153e-07, "loss": 0.2611, "step": 8782 }, { "epoch": 2.835223374631745, "grad_norm": 0.39736288674425807, "learning_rate": 1.2227761192043497e-07, "loss": 0.2759, "step": 8783 }, { "epoch": 2.8355462286613666, "grad_norm": 0.47364364911615237, "learning_rate": 1.2179798378891548e-07, "loss": 0.2571, "step": 8784 }, { "epoch": 2.8358690826909885, "grad_norm": 0.39135872472595407, "learning_rate": 1.21319290465432e-07, "loss": 0.2436, "step": 8785 }, { "epoch": 2.8361919367206103, "grad_norm": 0.4092233855875901, "learning_rate": 1.208415320106318e-07, "loss": 0.2491, "step": 8786 }, { "epoch": 2.836514790750232, "grad_norm": 0.31294709113593366, "learning_rate": 1.20364708485049e-07, "loss": 0.2379, "step": 8787 }, { "epoch": 2.836837644779854, "grad_norm": 0.3389594252447578, "learning_rate": 1.1988881994909523e-07, "loss": 0.2642, "step": 8788 }, { "epoch": 2.837160498809476, "grad_norm": 0.34788319840279014, "learning_rate": 1.194138664630648e-07, "loss": 0.2348, "step": 8789 }, { "epoch": 2.8374833528390977, "grad_norm": 0.3648485894609377, "learning_rate": 1.189398480871337e-07, "loss": 0.2759, "step": 8790 }, { "epoch": 2.8378062068687195, "grad_norm": 0.3651565280558437, "learning_rate": 1.1846676488135977e-07, "loss": 0.2256, "step": 8791 }, { "epoch": 2.8381290608983414, "grad_norm": 0.40585094231947966, "learning_rate": 1.1799461690568086e-07, "loss": 0.2379, "step": 8792 }, { "epoch": 2.8384519149279632, "grad_norm": 0.36921321414188096, "learning_rate": 1.1752340421991831e-07, "loss": 0.2194, "step": 8793 }, { "epoch": 2.838774768957585, "grad_norm": 0.48961591244934277, "learning_rate": 1.1705312688377517e-07, "loss": 0.2709, "step": 8794 }, { "epoch": 2.839097622987207, "grad_norm": 0.3745952576837891, "learning_rate": 1.165837849568338e-07, "loss": 0.29, "step": 8795 }, { "epoch": 2.8394204770168288, "grad_norm": 0.3606403554373714, "learning_rate": 1.1611537849855913e-07, "loss": 0.2674, "step": 8796 }, { "epoch": 2.8397433310464506, "grad_norm": 0.39056336031945205, "learning_rate": 1.1564790756829785e-07, "loss": 0.2086, "step": 8797 }, { "epoch": 2.8400661850760724, "grad_norm": 0.3832597518622155, "learning_rate": 1.1518137222527759e-07, "loss": 0.263, "step": 8798 }, { "epoch": 2.8403890391056943, "grad_norm": 0.3370147819225989, "learning_rate": 1.1471577252860854e-07, "loss": 0.2214, "step": 8799 }, { "epoch": 2.840711893135316, "grad_norm": 0.4491519720121743, "learning_rate": 1.1425110853728021e-07, "loss": 0.2979, "step": 8800 }, { "epoch": 2.841034747164938, "grad_norm": 0.4002268444261363, "learning_rate": 1.137873803101655e-07, "loss": 0.2535, "step": 8801 }, { "epoch": 2.84135760119456, "grad_norm": 0.34922774128060163, "learning_rate": 1.1332458790601908e-07, "loss": 0.2477, "step": 8802 }, { "epoch": 2.8416804552241817, "grad_norm": 0.3597004620614129, "learning_rate": 1.1286273138347486e-07, "loss": 0.3092, "step": 8803 }, { "epoch": 2.8420033092538035, "grad_norm": 0.3350968259361932, "learning_rate": 1.1240181080105022e-07, "loss": 0.2372, "step": 8804 }, { "epoch": 2.8423261632834254, "grad_norm": 0.3810659627235078, "learning_rate": 1.1194182621714177e-07, "loss": 0.2818, "step": 8805 }, { "epoch": 2.842649017313047, "grad_norm": 0.32635345074194144, "learning_rate": 1.1148277769003124e-07, "loss": 0.2694, "step": 8806 }, { "epoch": 2.842971871342669, "grad_norm": 0.33347611809167926, "learning_rate": 1.1102466527787791e-07, "loss": 0.2632, "step": 8807 }, { "epoch": 2.843294725372291, "grad_norm": 0.34872080113455783, "learning_rate": 1.1056748903872455e-07, "loss": 0.2518, "step": 8808 }, { "epoch": 2.843617579401913, "grad_norm": 0.3950623983203806, "learning_rate": 1.1011124903049314e-07, "loss": 0.2951, "step": 8809 }, { "epoch": 2.843940433431535, "grad_norm": 0.42811062071327866, "learning_rate": 1.096559453109916e-07, "loss": 0.2394, "step": 8810 }, { "epoch": 2.844263287461157, "grad_norm": 0.4828842803976769, "learning_rate": 1.0920157793790381e-07, "loss": 0.2729, "step": 8811 }, { "epoch": 2.8445861414907787, "grad_norm": 0.3558440273474862, "learning_rate": 1.0874814696879787e-07, "loss": 0.2542, "step": 8812 }, { "epoch": 2.8449089955204006, "grad_norm": 0.40952929264430515, "learning_rate": 1.082956524611245e-07, "loss": 0.25, "step": 8813 }, { "epoch": 2.8452318495500224, "grad_norm": 0.4479195631587194, "learning_rate": 1.07844094472212e-07, "loss": 0.21, "step": 8814 }, { "epoch": 2.8455547035796442, "grad_norm": 0.3349481667149899, "learning_rate": 1.0739347305927377e-07, "loss": 0.277, "step": 8815 }, { "epoch": 2.845877557609266, "grad_norm": 0.37761243066035427, "learning_rate": 1.0694378827940166e-07, "loss": 0.3018, "step": 8816 }, { "epoch": 2.846200411638888, "grad_norm": 0.3644217062363294, "learning_rate": 1.0649504018957173e-07, "loss": 0.2268, "step": 8817 }, { "epoch": 2.8465232656685098, "grad_norm": 0.3349968499041283, "learning_rate": 1.060472288466377e-07, "loss": 0.2155, "step": 8818 }, { "epoch": 2.8468461196981316, "grad_norm": 0.44487535167564973, "learning_rate": 1.0560035430733833e-07, "loss": 0.3151, "step": 8819 }, { "epoch": 2.8471689737277535, "grad_norm": 0.42613022797885386, "learning_rate": 1.0515441662829083e-07, "loss": 0.2715, "step": 8820 }, { "epoch": 2.8474918277573753, "grad_norm": 0.3888031712561564, "learning_rate": 1.0470941586599669e-07, "loss": 0.2429, "step": 8821 }, { "epoch": 2.847814681786997, "grad_norm": 0.4092070960660928, "learning_rate": 1.0426535207683413e-07, "loss": 0.2949, "step": 8822 }, { "epoch": 2.848137535816619, "grad_norm": 0.3344921066359588, "learning_rate": 1.0382222531706815e-07, "loss": 0.2376, "step": 8823 }, { "epoch": 2.848460389846241, "grad_norm": 0.3063137330791311, "learning_rate": 1.0338003564284049e-07, "loss": 0.2435, "step": 8824 }, { "epoch": 2.8487832438758627, "grad_norm": 0.5455072072961091, "learning_rate": 1.0293878311017635e-07, "loss": 0.2468, "step": 8825 }, { "epoch": 2.8491060979054845, "grad_norm": 0.42204501741871475, "learning_rate": 1.0249846777498268e-07, "loss": 0.2777, "step": 8826 }, { "epoch": 2.8494289519351064, "grad_norm": 0.34619374930671976, "learning_rate": 1.0205908969304567e-07, "loss": 0.2417, "step": 8827 }, { "epoch": 2.849751805964728, "grad_norm": 0.36511043213847166, "learning_rate": 1.0162064892003414e-07, "loss": 0.3188, "step": 8828 }, { "epoch": 2.85007465999435, "grad_norm": 0.35775896809457447, "learning_rate": 1.0118314551149866e-07, "loss": 0.2921, "step": 8829 }, { "epoch": 2.850397514023972, "grad_norm": 0.3739732199227687, "learning_rate": 1.0074657952286903e-07, "loss": 0.2301, "step": 8830 }, { "epoch": 2.8507203680535937, "grad_norm": 0.38753647837236777, "learning_rate": 1.0031095100945936e-07, "loss": 0.2798, "step": 8831 }, { "epoch": 2.8510432220832156, "grad_norm": 0.41446256477945637, "learning_rate": 9.987626002646216e-08, "loss": 0.2789, "step": 8832 }, { "epoch": 2.8513660761128374, "grad_norm": 0.3995570748705052, "learning_rate": 9.944250662895171e-08, "loss": 0.2232, "step": 8833 }, { "epoch": 2.8516889301424593, "grad_norm": 0.34014615533138637, "learning_rate": 9.900969087188571e-08, "loss": 0.2478, "step": 8834 }, { "epoch": 2.852011784172081, "grad_norm": 0.356402426540038, "learning_rate": 9.857781281009865e-08, "loss": 0.2268, "step": 8835 }, { "epoch": 2.852334638201703, "grad_norm": 0.3274103660498253, "learning_rate": 9.814687249831173e-08, "loss": 0.2662, "step": 8836 }, { "epoch": 2.852657492231325, "grad_norm": 0.42887540174631034, "learning_rate": 9.771686999112212e-08, "loss": 0.2756, "step": 8837 }, { "epoch": 2.8529803462609467, "grad_norm": 0.400439907259734, "learning_rate": 9.728780534301206e-08, "loss": 0.2249, "step": 8838 }, { "epoch": 2.8533032002905685, "grad_norm": 0.6855519940926387, "learning_rate": 9.685967860834227e-08, "loss": 0.2483, "step": 8839 }, { "epoch": 2.8536260543201903, "grad_norm": 0.3552208431623158, "learning_rate": 9.643248984135683e-08, "loss": 0.255, "step": 8840 }, { "epoch": 2.853948908349812, "grad_norm": 0.32550709057107546, "learning_rate": 9.600623909617912e-08, "loss": 0.2641, "step": 8841 }, { "epoch": 2.854271762379434, "grad_norm": 0.37790385844407737, "learning_rate": 9.558092642681593e-08, "loss": 0.2589, "step": 8842 }, { "epoch": 2.854594616409056, "grad_norm": 0.380024522881122, "learning_rate": 9.51565518871525e-08, "loss": 0.2584, "step": 8843 }, { "epoch": 2.8549174704386777, "grad_norm": 0.3480524786400942, "learning_rate": 9.473311553095665e-08, "loss": 0.2931, "step": 8844 }, { "epoch": 2.8552403244682996, "grad_norm": 0.3868146429553256, "learning_rate": 9.431061741187713e-08, "loss": 0.2438, "step": 8845 }, { "epoch": 2.8555631784979214, "grad_norm": 0.4334289383199244, "learning_rate": 9.388905758344446e-08, "loss": 0.247, "step": 8846 }, { "epoch": 2.8558860325275433, "grad_norm": 0.4146547602463916, "learning_rate": 9.346843609906925e-08, "loss": 0.227, "step": 8847 }, { "epoch": 2.856208886557165, "grad_norm": 0.380984578568316, "learning_rate": 9.304875301204385e-08, "loss": 0.2628, "step": 8848 }, { "epoch": 2.856531740586787, "grad_norm": 0.38261614442550074, "learning_rate": 9.263000837554241e-08, "loss": 0.2489, "step": 8849 }, { "epoch": 2.856854594616409, "grad_norm": 0.41317951978871637, "learning_rate": 9.221220224261668e-08, "loss": 0.2473, "step": 8850 }, { "epoch": 2.857177448646031, "grad_norm": 0.40082322404050597, "learning_rate": 9.17953346662051e-08, "loss": 0.2819, "step": 8851 }, { "epoch": 2.857500302675653, "grad_norm": 0.4221876506981164, "learning_rate": 9.137940569912212e-08, "loss": 0.2607, "step": 8852 }, { "epoch": 2.8578231567052748, "grad_norm": 0.4307096502592729, "learning_rate": 9.096441539406725e-08, "loss": 0.261, "step": 8853 }, { "epoch": 2.8581460107348966, "grad_norm": 0.36920612181001583, "learning_rate": 9.055036380361676e-08, "loss": 0.2789, "step": 8854 }, { "epoch": 2.8584688647645184, "grad_norm": 0.4148138929109271, "learning_rate": 9.013725098023284e-08, "loss": 0.2589, "step": 8855 }, { "epoch": 2.8587917187941403, "grad_norm": 0.3777034976578622, "learning_rate": 8.972507697625448e-08, "loss": 0.2841, "step": 8856 }, { "epoch": 2.859114572823762, "grad_norm": 0.3849637142759451, "learning_rate": 8.93138418439049e-08, "loss": 0.2285, "step": 8857 }, { "epoch": 2.859437426853384, "grad_norm": 0.37167039059783114, "learning_rate": 8.890354563528574e-08, "loss": 0.2766, "step": 8858 }, { "epoch": 2.859760280883006, "grad_norm": 0.3982419491683084, "learning_rate": 8.849418840238294e-08, "loss": 0.2093, "step": 8859 }, { "epoch": 2.8600831349126277, "grad_norm": 0.3729904405828439, "learning_rate": 8.808577019705916e-08, "loss": 0.2243, "step": 8860 }, { "epoch": 2.8604059889422495, "grad_norm": 0.34140175563409764, "learning_rate": 8.767829107106135e-08, "loss": 0.2545, "step": 8861 }, { "epoch": 2.8607288429718714, "grad_norm": 0.40350975687914303, "learning_rate": 8.727175107601653e-08, "loss": 0.3154, "step": 8862 }, { "epoch": 2.861051697001493, "grad_norm": 0.46762169781030555, "learning_rate": 8.686615026343353e-08, "loss": 0.2605, "step": 8863 }, { "epoch": 2.861374551031115, "grad_norm": 0.38131584746516506, "learning_rate": 8.646148868470039e-08, "loss": 0.2892, "step": 8864 }, { "epoch": 2.861697405060737, "grad_norm": 0.3750871772658065, "learning_rate": 8.605776639108775e-08, "loss": 0.2393, "step": 8865 }, { "epoch": 2.8620202590903587, "grad_norm": 0.3885713100659034, "learning_rate": 8.56549834337464e-08, "loss": 0.2805, "step": 8866 }, { "epoch": 2.8623431131199806, "grad_norm": 0.386648235010736, "learning_rate": 8.52531398637088e-08, "loss": 0.2643, "step": 8867 }, { "epoch": 2.8626659671496024, "grad_norm": 0.3798647801493757, "learning_rate": 8.485223573188761e-08, "loss": 0.2799, "step": 8868 }, { "epoch": 2.8629888211792243, "grad_norm": 0.34583168498958583, "learning_rate": 8.445227108907716e-08, "loss": 0.2955, "step": 8869 }, { "epoch": 2.863311675208846, "grad_norm": 0.37160357700204766, "learning_rate": 8.405324598595193e-08, "loss": 0.2623, "step": 8870 }, { "epoch": 2.863634529238468, "grad_norm": 0.33252052188728576, "learning_rate": 8.365516047306732e-08, "loss": 0.2328, "step": 8871 }, { "epoch": 2.86395738326809, "grad_norm": 0.3372552062092405, "learning_rate": 8.325801460086213e-08, "loss": 0.2392, "step": 8872 }, { "epoch": 2.8642802372977116, "grad_norm": 0.3925513820784258, "learning_rate": 8.286180841965196e-08, "loss": 0.2769, "step": 8873 }, { "epoch": 2.8646030913273335, "grad_norm": 0.39313369102826623, "learning_rate": 8.246654197963749e-08, "loss": 0.3363, "step": 8874 }, { "epoch": 2.8649259453569553, "grad_norm": 0.41586919777935516, "learning_rate": 8.207221533089699e-08, "loss": 0.2649, "step": 8875 }, { "epoch": 2.8652487993865776, "grad_norm": 0.4027615413775472, "learning_rate": 8.167882852339215e-08, "loss": 0.2624, "step": 8876 }, { "epoch": 2.8655716534161995, "grad_norm": 0.3332373884542743, "learning_rate": 8.128638160696394e-08, "loss": 0.2251, "step": 8877 }, { "epoch": 2.8658945074458213, "grad_norm": 0.40963475534397925, "learning_rate": 8.08948746313351e-08, "loss": 0.2988, "step": 8878 }, { "epoch": 2.866217361475443, "grad_norm": 0.3911701987931904, "learning_rate": 8.050430764610844e-08, "loss": 0.2832, "step": 8879 }, { "epoch": 2.866540215505065, "grad_norm": 0.3975576596779798, "learning_rate": 8.011468070076938e-08, "loss": 0.2439, "step": 8880 }, { "epoch": 2.866863069534687, "grad_norm": 0.35819668927380655, "learning_rate": 7.972599384468177e-08, "loss": 0.2538, "step": 8881 }, { "epoch": 2.8671859235643087, "grad_norm": 0.5062860216679793, "learning_rate": 7.93382471270937e-08, "loss": 0.2339, "step": 8882 }, { "epoch": 2.8675087775939305, "grad_norm": 0.43733237537003133, "learning_rate": 7.895144059712921e-08, "loss": 0.2047, "step": 8883 }, { "epoch": 2.8678316316235524, "grad_norm": 0.40686662847779426, "learning_rate": 7.856557430379912e-08, "loss": 0.2733, "step": 8884 }, { "epoch": 2.868154485653174, "grad_norm": 0.3911677088457818, "learning_rate": 7.818064829599014e-08, "loss": 0.2766, "step": 8885 }, { "epoch": 2.868477339682796, "grad_norm": 0.4133375400935979, "learning_rate": 7.779666262247243e-08, "loss": 0.2678, "step": 8886 }, { "epoch": 2.868800193712418, "grad_norm": 0.3696444142520767, "learning_rate": 7.741361733189789e-08, "loss": 0.2952, "step": 8887 }, { "epoch": 2.8691230477420397, "grad_norm": 0.4779624238985153, "learning_rate": 7.703151247279522e-08, "loss": 0.3209, "step": 8888 }, { "epoch": 2.8694459017716616, "grad_norm": 0.34093719897598246, "learning_rate": 7.665034809357818e-08, "loss": 0.2476, "step": 8889 }, { "epoch": 2.8697687558012834, "grad_norm": 0.3256021549406001, "learning_rate": 7.627012424253982e-08, "loss": 0.2624, "step": 8890 }, { "epoch": 2.8700916098309053, "grad_norm": 0.3894616155850445, "learning_rate": 7.589084096785325e-08, "loss": 0.2421, "step": 8891 }, { "epoch": 2.870414463860527, "grad_norm": 0.42830133396987585, "learning_rate": 7.551249831757339e-08, "loss": 0.2803, "step": 8892 }, { "epoch": 2.870737317890149, "grad_norm": 0.39629334806011046, "learning_rate": 7.513509633963605e-08, "loss": 0.3033, "step": 8893 }, { "epoch": 2.871060171919771, "grad_norm": 0.41213733717736517, "learning_rate": 7.475863508185715e-08, "loss": 0.2541, "step": 8894 }, { "epoch": 2.8713830259493927, "grad_norm": 0.3768835647704981, "learning_rate": 7.438311459193437e-08, "loss": 0.2664, "step": 8895 }, { "epoch": 2.8717058799790145, "grad_norm": 0.383361493795979, "learning_rate": 7.400853491744547e-08, "loss": 0.233, "step": 8896 }, { "epoch": 2.8720287340086363, "grad_norm": 0.3596184811266577, "learning_rate": 7.363489610584834e-08, "loss": 0.2306, "step": 8897 }, { "epoch": 2.872351588038258, "grad_norm": 0.4006284366732032, "learning_rate": 7.326219820448343e-08, "loss": 0.3026, "step": 8898 }, { "epoch": 2.87267444206788, "grad_norm": 0.3559258159996078, "learning_rate": 7.289044126057131e-08, "loss": 0.2922, "step": 8899 }, { "epoch": 2.872997296097502, "grad_norm": 0.36771144151771373, "learning_rate": 7.251962532121181e-08, "loss": 0.2641, "step": 8900 }, { "epoch": 2.8733201501271237, "grad_norm": 0.4315668495926073, "learning_rate": 7.214975043338817e-08, "loss": 0.2523, "step": 8901 }, { "epoch": 2.8736430041567456, "grad_norm": 0.3446565647223176, "learning_rate": 7.178081664396207e-08, "loss": 0.2308, "step": 8902 }, { "epoch": 2.8739658581863674, "grad_norm": 0.41792744705812207, "learning_rate": 7.141282399967697e-08, "loss": 0.1947, "step": 8903 }, { "epoch": 2.8742887122159892, "grad_norm": 0.3854464752027699, "learning_rate": 7.104577254715805e-08, "loss": 0.3017, "step": 8904 }, { "epoch": 2.874611566245611, "grad_norm": 0.3535855511074068, "learning_rate": 7.067966233290813e-08, "loss": 0.3074, "step": 8905 }, { "epoch": 2.874934420275233, "grad_norm": 0.4288677811277097, "learning_rate": 7.031449340331592e-08, "loss": 0.2099, "step": 8906 }, { "epoch": 2.875257274304855, "grad_norm": 0.33601969476110316, "learning_rate": 6.995026580464442e-08, "loss": 0.2403, "step": 8907 }, { "epoch": 2.8755801283344766, "grad_norm": 0.40616454808663216, "learning_rate": 6.958697958304339e-08, "loss": 0.2595, "step": 8908 }, { "epoch": 2.8759029823640985, "grad_norm": 0.38676061458771915, "learning_rate": 6.922463478453933e-08, "loss": 0.2466, "step": 8909 }, { "epoch": 2.8762258363937203, "grad_norm": 0.35157596036182664, "learning_rate": 6.88632314550422e-08, "loss": 0.27, "step": 8910 }, { "epoch": 2.876548690423342, "grad_norm": 0.36874338051622096, "learning_rate": 6.850276964033953e-08, "loss": 0.2688, "step": 8911 }, { "epoch": 2.876871544452964, "grad_norm": 0.3686814621235449, "learning_rate": 6.814324938610228e-08, "loss": 0.2399, "step": 8912 }, { "epoch": 2.877194398482586, "grad_norm": 0.3751696467990544, "learning_rate": 6.77846707378807e-08, "loss": 0.2483, "step": 8913 }, { "epoch": 2.8775172525122077, "grad_norm": 0.35829828033997624, "learning_rate": 6.742703374110759e-08, "loss": 0.2467, "step": 8914 }, { "epoch": 2.8778401065418295, "grad_norm": 0.4213279803524928, "learning_rate": 6.707033844109339e-08, "loss": 0.2917, "step": 8915 }, { "epoch": 2.8781629605714514, "grad_norm": 0.361889642277985, "learning_rate": 6.671458488303195e-08, "loss": 0.2303, "step": 8916 }, { "epoch": 2.8784858146010732, "grad_norm": 0.39682231531756146, "learning_rate": 6.635977311199553e-08, "loss": 0.3013, "step": 8917 }, { "epoch": 2.8788086686306955, "grad_norm": 0.3972592736694088, "learning_rate": 6.600590317293986e-08, "loss": 0.2966, "step": 8918 }, { "epoch": 2.8791315226603174, "grad_norm": 0.35040649684162556, "learning_rate": 6.565297511069907e-08, "loss": 0.279, "step": 8919 }, { "epoch": 2.879454376689939, "grad_norm": 0.40484767173039904, "learning_rate": 6.53009889699882e-08, "loss": 0.2584, "step": 8920 }, { "epoch": 2.879777230719561, "grad_norm": 0.41237095301592636, "learning_rate": 6.494994479540411e-08, "loss": 0.225, "step": 8921 }, { "epoch": 2.880100084749183, "grad_norm": 0.36256536599250405, "learning_rate": 6.459984263142287e-08, "loss": 0.2672, "step": 8922 }, { "epoch": 2.8804229387788047, "grad_norm": 0.35669755706290873, "learning_rate": 6.425068252240318e-08, "loss": 0.2787, "step": 8923 }, { "epoch": 2.8807457928084266, "grad_norm": 0.3904952121698954, "learning_rate": 6.39024645125813e-08, "loss": 0.2803, "step": 8924 }, { "epoch": 2.8810686468380484, "grad_norm": 0.4704115399979275, "learning_rate": 6.355518864607779e-08, "loss": 0.2316, "step": 8925 }, { "epoch": 2.8813915008676703, "grad_norm": 0.376791627596903, "learning_rate": 6.320885496689078e-08, "loss": 0.2739, "step": 8926 }, { "epoch": 2.881714354897292, "grad_norm": 0.3492317439990225, "learning_rate": 6.286346351890182e-08, "loss": 0.2636, "step": 8927 }, { "epoch": 2.882037208926914, "grad_norm": 0.3515542168038523, "learning_rate": 6.251901434586926e-08, "loss": 0.2137, "step": 8928 }, { "epoch": 2.882360062956536, "grad_norm": 0.33674369375647273, "learning_rate": 6.217550749143652e-08, "loss": 0.2803, "step": 8929 }, { "epoch": 2.8826829169861576, "grad_norm": 0.4327652837272456, "learning_rate": 6.183294299912378e-08, "loss": 0.26, "step": 8930 }, { "epoch": 2.8830057710157795, "grad_norm": 0.389957723850432, "learning_rate": 6.149132091233551e-08, "loss": 0.2674, "step": 8931 }, { "epoch": 2.8833286250454013, "grad_norm": 0.3334256769507574, "learning_rate": 6.115064127435205e-08, "loss": 0.264, "step": 8932 }, { "epoch": 2.883651479075023, "grad_norm": 0.35490007196352974, "learning_rate": 6.08109041283389e-08, "loss": 0.2539, "step": 8933 }, { "epoch": 2.883974333104645, "grad_norm": 0.3031362598200539, "learning_rate": 6.047210951733994e-08, "loss": 0.2528, "step": 8934 }, { "epoch": 2.884297187134267, "grad_norm": 0.3521411962812369, "learning_rate": 6.013425748428003e-08, "loss": 0.2567, "step": 8935 }, { "epoch": 2.8846200411638887, "grad_norm": 0.37504472615375206, "learning_rate": 5.979734807196324e-08, "loss": 0.2461, "step": 8936 }, { "epoch": 2.8849428951935105, "grad_norm": 0.3703024232687422, "learning_rate": 5.946138132307793e-08, "loss": 0.2733, "step": 8937 }, { "epoch": 2.8852657492231324, "grad_norm": 0.3843238215657151, "learning_rate": 5.912635728018922e-08, "loss": 0.2477, "step": 8938 }, { "epoch": 2.8855886032527542, "grad_norm": 0.41310964176218534, "learning_rate": 5.8792275985743995e-08, "loss": 0.2977, "step": 8939 }, { "epoch": 2.885911457282376, "grad_norm": 0.3606508606910252, "learning_rate": 5.845913748207005e-08, "loss": 0.2477, "step": 8940 }, { "epoch": 2.886234311311998, "grad_norm": 0.3383421553750155, "learning_rate": 5.812694181137529e-08, "loss": 0.2395, "step": 8941 }, { "epoch": 2.8865571653416198, "grad_norm": 0.33285552151748415, "learning_rate": 5.779568901575022e-08, "loss": 0.2943, "step": 8942 }, { "epoch": 2.8868800193712416, "grad_norm": 0.3644798506443593, "learning_rate": 5.74653791371621e-08, "loss": 0.2791, "step": 8943 }, { "epoch": 2.887202873400864, "grad_norm": 0.466427667832663, "learning_rate": 5.7136012217462455e-08, "loss": 0.2208, "step": 8944 }, { "epoch": 2.8875257274304857, "grad_norm": 0.4121861064429905, "learning_rate": 5.680758829837957e-08, "loss": 0.318, "step": 8945 }, { "epoch": 2.8878485814601076, "grad_norm": 0.3228928910663548, "learning_rate": 5.6480107421525984e-08, "loss": 0.223, "step": 8946 }, { "epoch": 2.8881714354897294, "grad_norm": 0.6002401431232331, "learning_rate": 5.615356962839185e-08, "loss": 0.3172, "step": 8947 }, { "epoch": 2.8884942895193513, "grad_norm": 0.35354247380146503, "learning_rate": 5.5827974960349915e-08, "loss": 0.2784, "step": 8948 }, { "epoch": 2.888817143548973, "grad_norm": 0.33602113691244406, "learning_rate": 5.5503323458652166e-08, "loss": 0.2428, "step": 8949 }, { "epoch": 2.889139997578595, "grad_norm": 0.3758614762798173, "learning_rate": 5.517961516443237e-08, "loss": 0.357, "step": 8950 }, { "epoch": 2.889462851608217, "grad_norm": 0.3850505739393165, "learning_rate": 5.48568501187019e-08, "loss": 0.3052, "step": 8951 }, { "epoch": 2.8897857056378387, "grad_norm": 0.34160755435116213, "learning_rate": 5.453502836235719e-08, "loss": 0.2246, "step": 8952 }, { "epoch": 2.8901085596674605, "grad_norm": 0.392280556905742, "learning_rate": 5.4214149936170634e-08, "loss": 0.2114, "step": 8953 }, { "epoch": 2.8904314136970823, "grad_norm": 0.4338812070040062, "learning_rate": 5.3894214880798044e-08, "loss": 0.2779, "step": 8954 }, { "epoch": 2.890754267726704, "grad_norm": 0.37092728632865213, "learning_rate": 5.35752232367745e-08, "loss": 0.1765, "step": 8955 }, { "epoch": 2.891077121756326, "grad_norm": 0.35029332065259905, "learning_rate": 5.325717504451599e-08, "loss": 0.2713, "step": 8956 }, { "epoch": 2.891399975785948, "grad_norm": 0.37928915481905645, "learning_rate": 5.2940070344317807e-08, "loss": 0.2861, "step": 8957 }, { "epoch": 2.8917228298155697, "grad_norm": 0.38004886783968966, "learning_rate": 5.262390917635862e-08, "loss": 0.2536, "step": 8958 }, { "epoch": 2.8920456838451916, "grad_norm": 0.3790531251702652, "learning_rate": 5.23086915806939e-08, "loss": 0.2836, "step": 8959 }, { "epoch": 2.8923685378748134, "grad_norm": 0.40804730956432167, "learning_rate": 5.19944175972617e-08, "loss": 0.2645, "step": 8960 }, { "epoch": 2.8926913919044352, "grad_norm": 0.4021334664386471, "learning_rate": 5.1681087265881e-08, "loss": 0.2569, "step": 8961 }, { "epoch": 2.893014245934057, "grad_norm": 0.3251913160570221, "learning_rate": 5.136870062624921e-08, "loss": 0.2793, "step": 8962 }, { "epoch": 2.893337099963679, "grad_norm": 0.39381447704407474, "learning_rate": 5.1057257717946335e-08, "loss": 0.2844, "step": 8963 }, { "epoch": 2.893659953993301, "grad_norm": 0.4208761635023336, "learning_rate": 5.074675858042999e-08, "loss": 0.2753, "step": 8964 }, { "epoch": 2.8939828080229226, "grad_norm": 0.38158716486070043, "learning_rate": 5.043720325304202e-08, "loss": 0.2289, "step": 8965 }, { "epoch": 2.8943056620525445, "grad_norm": 0.44885121005397366, "learning_rate": 5.0128591775001064e-08, "loss": 0.2452, "step": 8966 }, { "epoch": 2.8946285160821663, "grad_norm": 0.39584232487690346, "learning_rate": 4.982092418540835e-08, "loss": 0.2628, "step": 8967 }, { "epoch": 2.894951370111788, "grad_norm": 0.3633052726815982, "learning_rate": 4.951420052324518e-08, "loss": 0.2913, "step": 8968 }, { "epoch": 2.89527422414141, "grad_norm": 0.41546653657729904, "learning_rate": 4.920842082737381e-08, "loss": 0.2563, "step": 8969 }, { "epoch": 2.895597078171032, "grad_norm": 0.3148583921394831, "learning_rate": 4.890358513653409e-08, "loss": 0.2468, "step": 8970 }, { "epoch": 2.8959199322006537, "grad_norm": 0.322900637345062, "learning_rate": 4.859969348934929e-08, "loss": 0.1991, "step": 8971 }, { "epoch": 2.8962427862302755, "grad_norm": 0.36216033675851655, "learning_rate": 4.829674592432276e-08, "loss": 0.2397, "step": 8972 }, { "epoch": 2.8965656402598974, "grad_norm": 0.337811364000998, "learning_rate": 4.799474247983632e-08, "loss": 0.2292, "step": 8973 }, { "epoch": 2.8968884942895192, "grad_norm": 0.44112804708690034, "learning_rate": 4.769368319415435e-08, "loss": 0.2147, "step": 8974 }, { "epoch": 2.897211348319141, "grad_norm": 0.6567716012016556, "learning_rate": 4.7393568105420505e-08, "loss": 0.2255, "step": 8975 }, { "epoch": 2.897534202348763, "grad_norm": 0.349778749763065, "learning_rate": 4.709439725165854e-08, "loss": 0.2805, "step": 8976 }, { "epoch": 2.8978570563783848, "grad_norm": 0.42033817881628843, "learning_rate": 4.679617067077313e-08, "loss": 0.2904, "step": 8977 }, { "epoch": 2.8981799104080066, "grad_norm": 0.41331084646684013, "learning_rate": 4.649888840054906e-08, "loss": 0.291, "step": 8978 }, { "epoch": 2.8985027644376284, "grad_norm": 0.37543908714384033, "learning_rate": 4.6202550478652014e-08, "loss": 0.2507, "step": 8979 }, { "epoch": 2.8988256184672503, "grad_norm": 0.3427920369129643, "learning_rate": 4.5907156942627814e-08, "loss": 0.2379, "step": 8980 }, { "epoch": 2.899148472496872, "grad_norm": 0.35705110762365183, "learning_rate": 4.561270782990068e-08, "loss": 0.2603, "step": 8981 }, { "epoch": 2.899471326526494, "grad_norm": 0.4023108291003804, "learning_rate": 4.53192031777791e-08, "loss": 0.2653, "step": 8982 }, { "epoch": 2.899794180556116, "grad_norm": 0.39994173102987135, "learning_rate": 4.502664302344833e-08, "loss": 0.2811, "step": 8983 }, { "epoch": 2.9001170345857377, "grad_norm": 0.4076317232431328, "learning_rate": 4.4735027403976226e-08, "loss": 0.2619, "step": 8984 }, { "epoch": 2.9004398886153595, "grad_norm": 0.35026028178604035, "learning_rate": 4.444435635630906e-08, "loss": 0.2611, "step": 8985 }, { "epoch": 2.900762742644982, "grad_norm": 0.38008681981030423, "learning_rate": 4.41546299172757e-08, "loss": 0.3373, "step": 8986 }, { "epoch": 2.9010855966746036, "grad_norm": 0.3578432093665545, "learning_rate": 4.386584812358263e-08, "loss": 0.2178, "step": 8987 }, { "epoch": 2.9014084507042255, "grad_norm": 0.3720267882386812, "learning_rate": 4.357801101181974e-08, "loss": 0.3005, "step": 8988 }, { "epoch": 2.9017313047338473, "grad_norm": 0.33674572207166575, "learning_rate": 4.329111861845369e-08, "loss": 0.2478, "step": 8989 }, { "epoch": 2.902054158763469, "grad_norm": 0.3672753473034797, "learning_rate": 4.300517097983542e-08, "loss": 0.2768, "step": 8990 }, { "epoch": 2.902377012793091, "grad_norm": 0.4091501191055875, "learning_rate": 4.2720168132192596e-08, "loss": 0.2541, "step": 8991 }, { "epoch": 2.902699866822713, "grad_norm": 0.38593233463450344, "learning_rate": 4.243611011163551e-08, "loss": 0.2268, "step": 8992 }, { "epoch": 2.9030227208523347, "grad_norm": 0.37803624849821016, "learning_rate": 4.21529969541537e-08, "loss": 0.2736, "step": 8993 }, { "epoch": 2.9033455748819565, "grad_norm": 0.3765049450127974, "learning_rate": 4.187082869561681e-08, "loss": 0.295, "step": 8994 }, { "epoch": 2.9036684289115784, "grad_norm": 0.38489498933673505, "learning_rate": 4.158960537177542e-08, "loss": 0.2748, "step": 8995 }, { "epoch": 2.9039912829412002, "grad_norm": 0.36815801779992746, "learning_rate": 4.130932701826018e-08, "loss": 0.2864, "step": 8996 }, { "epoch": 2.904314136970822, "grad_norm": 0.3852714348658794, "learning_rate": 4.1029993670581855e-08, "loss": 0.2403, "step": 8997 }, { "epoch": 2.904636991000444, "grad_norm": 0.31428612842802234, "learning_rate": 4.075160536413214e-08, "loss": 0.2424, "step": 8998 }, { "epoch": 2.9049598450300658, "grad_norm": 0.4459047084501844, "learning_rate": 4.047416213418198e-08, "loss": 0.2763, "step": 8999 }, { "epoch": 2.9052826990596876, "grad_norm": 0.39101301810428235, "learning_rate": 4.019766401588243e-08, "loss": 0.2853, "step": 9000 }, { "epoch": 2.9056055530893095, "grad_norm": 0.3130973871366949, "learning_rate": 3.992211104426713e-08, "loss": 0.2519, "step": 9001 }, { "epoch": 2.9059284071189313, "grad_norm": 0.38951485925338103, "learning_rate": 3.9647503254246484e-08, "loss": 0.2734, "step": 9002 }, { "epoch": 2.906251261148553, "grad_norm": 0.420873761609598, "learning_rate": 3.9373840680613507e-08, "loss": 0.285, "step": 9003 }, { "epoch": 2.906574115178175, "grad_norm": 0.36827348269338556, "learning_rate": 3.910112335804128e-08, "loss": 0.2636, "step": 9004 }, { "epoch": 2.906896969207797, "grad_norm": 0.3383838914620819, "learning_rate": 3.882935132108217e-08, "loss": 0.2867, "step": 9005 }, { "epoch": 2.9072198232374187, "grad_norm": 0.38822122083575805, "learning_rate": 3.8558524604170306e-08, "loss": 0.293, "step": 9006 }, { "epoch": 2.9075426772670405, "grad_norm": 0.37149170625190403, "learning_rate": 3.8288643241618234e-08, "loss": 0.2861, "step": 9007 }, { "epoch": 2.9078655312966624, "grad_norm": 0.3823630023981308, "learning_rate": 3.801970726761944e-08, "loss": 0.2832, "step": 9008 }, { "epoch": 2.908188385326284, "grad_norm": 0.393284877914958, "learning_rate": 3.7751716716247495e-08, "loss": 0.2129, "step": 9009 }, { "epoch": 2.908511239355906, "grad_norm": 0.3323597576686501, "learning_rate": 3.748467162145774e-08, "loss": 0.2575, "step": 9010 }, { "epoch": 2.9088340933855283, "grad_norm": 0.38722164155794436, "learning_rate": 3.721857201708312e-08, "loss": 0.2806, "step": 9011 }, { "epoch": 2.90915694741515, "grad_norm": 0.39121844385723026, "learning_rate": 3.695341793683915e-08, "loss": 0.215, "step": 9012 }, { "epoch": 2.909479801444772, "grad_norm": 0.35898097814678365, "learning_rate": 3.668920941431897e-08, "loss": 0.2402, "step": 9013 }, { "epoch": 2.909802655474394, "grad_norm": 0.8884892569944612, "learning_rate": 3.642594648299913e-08, "loss": 0.2049, "step": 9014 }, { "epoch": 2.9101255095040157, "grad_norm": 0.4055585893991671, "learning_rate": 3.6163629176233784e-08, "loss": 0.2701, "step": 9015 }, { "epoch": 2.9104483635336376, "grad_norm": 0.39086132388566996, "learning_rate": 3.590225752725801e-08, "loss": 0.2176, "step": 9016 }, { "epoch": 2.9107712175632594, "grad_norm": 0.2887046841700733, "learning_rate": 3.564183156918782e-08, "loss": 0.2321, "step": 9017 }, { "epoch": 2.9110940715928812, "grad_norm": 0.3588689286059514, "learning_rate": 3.538235133501766e-08, "loss": 0.2739, "step": 9018 }, { "epoch": 2.911416925622503, "grad_norm": 0.3940060072752342, "learning_rate": 3.512381685762539e-08, "loss": 0.2587, "step": 9019 }, { "epoch": 2.911739779652125, "grad_norm": 0.35692004880097566, "learning_rate": 3.4866228169764814e-08, "loss": 0.2408, "step": 9020 }, { "epoch": 2.912062633681747, "grad_norm": 0.39971872143303866, "learning_rate": 3.460958530407399e-08, "loss": 0.2676, "step": 9021 }, { "epoch": 2.9123854877113686, "grad_norm": 0.40041048432721127, "learning_rate": 3.435388829306774e-08, "loss": 0.24, "step": 9022 }, { "epoch": 2.9127083417409905, "grad_norm": 0.38600243511877386, "learning_rate": 3.4099137169142647e-08, "loss": 0.2589, "step": 9023 }, { "epoch": 2.9130311957706123, "grad_norm": 0.38189872108481654, "learning_rate": 3.3845331964577065e-08, "loss": 0.2122, "step": 9024 }, { "epoch": 2.913354049800234, "grad_norm": 0.3880608839193593, "learning_rate": 3.359247271152527e-08, "loss": 0.2675, "step": 9025 }, { "epoch": 2.913676903829856, "grad_norm": 0.3755111950126748, "learning_rate": 3.33405594420258e-08, "loss": 0.2566, "step": 9026 }, { "epoch": 2.913999757859478, "grad_norm": 0.4097457911334784, "learning_rate": 3.308959218799479e-08, "loss": 0.3062, "step": 9027 }, { "epoch": 2.9143226118890997, "grad_norm": 0.34630830062608803, "learning_rate": 3.283957098123014e-08, "loss": 0.2448, "step": 9028 }, { "epoch": 2.9146454659187215, "grad_norm": 0.5438833939584412, "learning_rate": 3.2590495853409006e-08, "loss": 0.2694, "step": 9029 }, { "epoch": 2.9149683199483434, "grad_norm": 0.3683184438157704, "learning_rate": 3.2342366836088656e-08, "loss": 0.2679, "step": 9030 }, { "epoch": 2.9152911739779652, "grad_norm": 0.3940381816166161, "learning_rate": 3.209518396070643e-08, "loss": 0.2413, "step": 9031 }, { "epoch": 2.915614028007587, "grad_norm": 0.3937368023010012, "learning_rate": 3.184894725858145e-08, "loss": 0.3098, "step": 9032 }, { "epoch": 2.915936882037209, "grad_norm": 0.42991799983218787, "learning_rate": 3.1603656760909594e-08, "loss": 0.2277, "step": 9033 }, { "epoch": 2.9162597360668308, "grad_norm": 0.3343345016940893, "learning_rate": 3.1359312498771e-08, "loss": 0.2982, "step": 9034 }, { "epoch": 2.9165825900964526, "grad_norm": 0.36509080136872174, "learning_rate": 3.111591450312173e-08, "loss": 0.2794, "step": 9035 }, { "epoch": 2.9169054441260744, "grad_norm": 0.3586433652785181, "learning_rate": 3.087346280480047e-08, "loss": 0.2852, "step": 9036 }, { "epoch": 2.9172282981556963, "grad_norm": 0.3634398635585951, "learning_rate": 3.063195743452679e-08, "loss": 0.2201, "step": 9037 }, { "epoch": 2.917551152185318, "grad_norm": 0.4221756525470699, "learning_rate": 3.03913984228979e-08, "loss": 0.2576, "step": 9038 }, { "epoch": 2.91787400621494, "grad_norm": 0.39880391555766215, "learning_rate": 3.015178580039274e-08, "loss": 0.2615, "step": 9039 }, { "epoch": 2.918196860244562, "grad_norm": 0.3415404936289436, "learning_rate": 2.991311959736953e-08, "loss": 0.2211, "step": 9040 }, { "epoch": 2.9185197142741837, "grad_norm": 0.2986571653467644, "learning_rate": 2.9675399844066586e-08, "loss": 0.2353, "step": 9041 }, { "epoch": 2.9188425683038055, "grad_norm": 0.36205544008327056, "learning_rate": 2.9438626570603976e-08, "loss": 0.2578, "step": 9042 }, { "epoch": 2.9191654223334274, "grad_norm": 0.3746217992516455, "learning_rate": 2.9202799806979375e-08, "loss": 0.2483, "step": 9043 }, { "epoch": 2.919488276363049, "grad_norm": 0.3787014989230633, "learning_rate": 2.896791958307221e-08, "loss": 0.2827, "step": 9044 }, { "epoch": 2.919811130392671, "grad_norm": 0.3288495830637952, "learning_rate": 2.8733985928642014e-08, "loss": 0.2243, "step": 9045 }, { "epoch": 2.920133984422293, "grad_norm": 0.3344475843929039, "learning_rate": 2.8500998873326743e-08, "loss": 0.2461, "step": 9046 }, { "epoch": 2.9204568384519147, "grad_norm": 0.3151390364391515, "learning_rate": 2.826895844664612e-08, "loss": 0.2462, "step": 9047 }, { "epoch": 2.9207796924815366, "grad_norm": 0.3918269197081774, "learning_rate": 2.8037864677999124e-08, "loss": 0.2641, "step": 9048 }, { "epoch": 2.9211025465111584, "grad_norm": 0.3844485962289996, "learning_rate": 2.780771759666484e-08, "loss": 0.2632, "step": 9049 }, { "epoch": 2.9214254005407803, "grad_norm": 0.3639907883663253, "learning_rate": 2.7578517231803268e-08, "loss": 0.241, "step": 9050 }, { "epoch": 2.921748254570402, "grad_norm": 0.3229807649662195, "learning_rate": 2.7350263612453685e-08, "loss": 0.2789, "step": 9051 }, { "epoch": 2.922071108600024, "grad_norm": 0.3722362983626105, "learning_rate": 2.7122956767535457e-08, "loss": 0.2867, "step": 9052 }, { "epoch": 2.9223939626296462, "grad_norm": 0.45775844192255644, "learning_rate": 2.6896596725847212e-08, "loss": 0.2587, "step": 9053 }, { "epoch": 2.922716816659268, "grad_norm": 0.4559950720992855, "learning_rate": 2.667118351606934e-08, "loss": 0.2684, "step": 9054 }, { "epoch": 2.92303967068889, "grad_norm": 0.41413204103931445, "learning_rate": 2.6446717166761502e-08, "loss": 0.2644, "step": 9055 }, { "epoch": 2.9233625247185118, "grad_norm": 0.35252475565698715, "learning_rate": 2.622319770636261e-08, "loss": 0.2583, "step": 9056 }, { "epoch": 2.9236853787481336, "grad_norm": 0.40810084059365986, "learning_rate": 2.6000625163191682e-08, "loss": 0.25, "step": 9057 }, { "epoch": 2.9240082327777555, "grad_norm": 0.41895711162595156, "learning_rate": 2.5778999565450322e-08, "loss": 0.2906, "step": 9058 }, { "epoch": 2.9243310868073773, "grad_norm": 0.4732416494296091, "learning_rate": 2.5558320941216905e-08, "loss": 0.2357, "step": 9059 }, { "epoch": 2.924653940836999, "grad_norm": 0.4095422048613954, "learning_rate": 2.5338589318451566e-08, "loss": 0.2467, "step": 9060 }, { "epoch": 2.924976794866621, "grad_norm": 0.37050821673721046, "learning_rate": 2.5119804724992867e-08, "loss": 0.3019, "step": 9061 }, { "epoch": 2.925299648896243, "grad_norm": 0.34721981465133495, "learning_rate": 2.4901967188561968e-08, "loss": 0.2475, "step": 9062 }, { "epoch": 2.9256225029258647, "grad_norm": 0.3823270255821741, "learning_rate": 2.468507673675846e-08, "loss": 0.2438, "step": 9063 }, { "epoch": 2.9259453569554865, "grad_norm": 0.32400128577342335, "learning_rate": 2.446913339706036e-08, "loss": 0.2538, "step": 9064 }, { "epoch": 2.9262682109851084, "grad_norm": 0.6211501318360312, "learning_rate": 2.425413719682995e-08, "loss": 0.2081, "step": 9065 }, { "epoch": 2.92659106501473, "grad_norm": 0.4202930136867577, "learning_rate": 2.4040088163304608e-08, "loss": 0.2587, "step": 9066 }, { "epoch": 2.926913919044352, "grad_norm": 0.8485898424232102, "learning_rate": 2.382698632360597e-08, "loss": 0.2508, "step": 9067 }, { "epoch": 2.927236773073974, "grad_norm": 0.345268782144602, "learning_rate": 2.3614831704732444e-08, "loss": 0.206, "step": 9068 }, { "epoch": 2.9275596271035957, "grad_norm": 0.32642803051329994, "learning_rate": 2.340362433356419e-08, "loss": 0.2436, "step": 9069 }, { "epoch": 2.9278824811332176, "grad_norm": 0.3824877761646829, "learning_rate": 2.3193364236860635e-08, "loss": 0.2419, "step": 9070 }, { "epoch": 2.9282053351628394, "grad_norm": 0.3698964083770731, "learning_rate": 2.298405144126131e-08, "loss": 0.1947, "step": 9071 }, { "epoch": 2.9285281891924613, "grad_norm": 0.3120550336440699, "learning_rate": 2.2775685973286663e-08, "loss": 0.2462, "step": 9072 }, { "epoch": 2.928851043222083, "grad_norm": 0.3976437181787362, "learning_rate": 2.256826785933641e-08, "loss": 0.2536, "step": 9073 }, { "epoch": 2.929173897251705, "grad_norm": 0.3741280568350634, "learning_rate": 2.2361797125688698e-08, "loss": 0.2973, "step": 9074 }, { "epoch": 2.929496751281327, "grad_norm": 0.4046026662626115, "learning_rate": 2.2156273798504267e-08, "loss": 0.2374, "step": 9075 }, { "epoch": 2.9298196053109486, "grad_norm": 0.34666083428848005, "learning_rate": 2.1951697903822287e-08, "loss": 0.2213, "step": 9076 }, { "epoch": 2.9301424593405705, "grad_norm": 0.4071241326297583, "learning_rate": 2.1748069467562858e-08, "loss": 0.256, "step": 9077 }, { "epoch": 2.9304653133701923, "grad_norm": 0.31244450383947747, "learning_rate": 2.154538851552451e-08, "loss": 0.2432, "step": 9078 }, { "epoch": 2.9307881673998146, "grad_norm": 0.367504990395866, "learning_rate": 2.1343655073386705e-08, "loss": 0.2514, "step": 9079 }, { "epoch": 2.9311110214294365, "grad_norm": 0.3974617656251734, "learning_rate": 2.114286916670899e-08, "loss": 0.2432, "step": 9080 }, { "epoch": 2.9314338754590583, "grad_norm": 0.3313593755209004, "learning_rate": 2.094303082093102e-08, "loss": 0.2619, "step": 9081 }, { "epoch": 2.93175672948868, "grad_norm": 0.3723803269301384, "learning_rate": 2.0744140061371708e-08, "loss": 0.2342, "step": 9082 }, { "epoch": 2.932079583518302, "grad_norm": 0.3502730515537596, "learning_rate": 2.05461969132309e-08, "loss": 0.2445, "step": 9083 }, { "epoch": 2.932402437547924, "grad_norm": 0.35182071029998374, "learning_rate": 2.0349201401586027e-08, "loss": 0.2659, "step": 9084 }, { "epoch": 2.9327252915775457, "grad_norm": 0.3749864008530407, "learning_rate": 2.0153153551397962e-08, "loss": 0.2662, "step": 9085 }, { "epoch": 2.9330481456071675, "grad_norm": 0.3739436325730362, "learning_rate": 1.9958053387505172e-08, "loss": 0.3385, "step": 9086 }, { "epoch": 2.9333709996367894, "grad_norm": 0.3462248083220116, "learning_rate": 1.976390093462621e-08, "loss": 0.2584, "step": 9087 }, { "epoch": 2.933693853666411, "grad_norm": 0.406050506139648, "learning_rate": 1.957069621735974e-08, "loss": 0.2572, "step": 9088 }, { "epoch": 2.934016707696033, "grad_norm": 0.3623755740776241, "learning_rate": 1.9378439260185345e-08, "loss": 0.3094, "step": 9089 }, { "epoch": 2.934339561725655, "grad_norm": 0.4302851835963993, "learning_rate": 1.9187130087461036e-08, "loss": 0.2546, "step": 9090 }, { "epoch": 2.9346624157552768, "grad_norm": 0.37187976071075507, "learning_rate": 1.899676872342493e-08, "loss": 0.2661, "step": 9091 }, { "epoch": 2.9349852697848986, "grad_norm": 0.3777605063446487, "learning_rate": 1.880735519219773e-08, "loss": 0.2277, "step": 9092 }, { "epoch": 2.9353081238145204, "grad_norm": 0.3686587431406024, "learning_rate": 1.861888951777524e-08, "loss": 0.2527, "step": 9093 }, { "epoch": 2.9356309778441423, "grad_norm": 0.36533199720987286, "learning_rate": 1.8431371724037527e-08, "loss": 0.2631, "step": 9094 }, { "epoch": 2.935953831873764, "grad_norm": 0.39652945609208373, "learning_rate": 1.8244801834742253e-08, "loss": 0.2615, "step": 9095 }, { "epoch": 2.936276685903386, "grad_norm": 0.3807573803941398, "learning_rate": 1.8059179873527176e-08, "loss": 0.2714, "step": 9096 }, { "epoch": 2.936599539933008, "grad_norm": 0.4316866028766658, "learning_rate": 1.7874505863910983e-08, "loss": 0.2606, "step": 9097 }, { "epoch": 2.9369223939626297, "grad_norm": 0.32782935816401476, "learning_rate": 1.7690779829291626e-08, "loss": 0.241, "step": 9098 }, { "epoch": 2.9372452479922515, "grad_norm": 0.36906114611246327, "learning_rate": 1.7508001792946314e-08, "loss": 0.2793, "step": 9099 }, { "epoch": 2.9375681020218734, "grad_norm": 0.43396267993104815, "learning_rate": 1.7326171778034028e-08, "loss": 0.2058, "step": 9100 }, { "epoch": 2.937890956051495, "grad_norm": 0.3946343160867108, "learning_rate": 1.7145289807590503e-08, "loss": 0.2727, "step": 9101 }, { "epoch": 2.938213810081117, "grad_norm": 0.37747573904253545, "learning_rate": 1.6965355904534908e-08, "loss": 0.273, "step": 9102 }, { "epoch": 2.938536664110739, "grad_norm": 0.42824400725836714, "learning_rate": 1.6786370091664004e-08, "loss": 0.2702, "step": 9103 }, { "epoch": 2.9388595181403607, "grad_norm": 0.3512262918663316, "learning_rate": 1.6608332391654656e-08, "loss": 0.2532, "step": 9104 }, { "epoch": 2.9391823721699826, "grad_norm": 0.34829702516160954, "learning_rate": 1.643124282706382e-08, "loss": 0.2582, "step": 9105 }, { "epoch": 2.9395052261996044, "grad_norm": 0.4033656078148016, "learning_rate": 1.625510142033021e-08, "loss": 0.2293, "step": 9106 }, { "epoch": 2.9398280802292263, "grad_norm": 0.3246877612764487, "learning_rate": 1.6079908193768477e-08, "loss": 0.2346, "step": 9107 }, { "epoch": 2.940150934258848, "grad_norm": 0.40777491865140236, "learning_rate": 1.59056631695767e-08, "loss": 0.2593, "step": 9108 }, { "epoch": 2.94047378828847, "grad_norm": 0.3588177415312733, "learning_rate": 1.573236636983222e-08, "loss": 0.2539, "step": 9109 }, { "epoch": 2.940796642318092, "grad_norm": 0.4953385751692422, "learning_rate": 1.5560017816489135e-08, "loss": 0.2392, "step": 9110 }, { "epoch": 2.9411194963477136, "grad_norm": 0.4280258116327429, "learning_rate": 1.538861753138582e-08, "loss": 0.2692, "step": 9111 }, { "epoch": 2.9414423503773355, "grad_norm": 0.3653145175003994, "learning_rate": 1.5218165536237395e-08, "loss": 0.2608, "step": 9112 }, { "epoch": 2.9417652044069573, "grad_norm": 0.38348683572178294, "learning_rate": 1.504866185263992e-08, "loss": 0.2952, "step": 9113 }, { "epoch": 2.942088058436579, "grad_norm": 0.6017392821963302, "learning_rate": 1.4880106502070389e-08, "loss": 0.2689, "step": 9114 }, { "epoch": 2.942410912466201, "grad_norm": 0.39821159468699907, "learning_rate": 1.4712499505883382e-08, "loss": 0.2217, "step": 9115 }, { "epoch": 2.942733766495823, "grad_norm": 0.35102064812297396, "learning_rate": 1.454584088531441e-08, "loss": 0.2853, "step": 9116 }, { "epoch": 2.9430566205254447, "grad_norm": 0.3343483127380246, "learning_rate": 1.4380130661479918e-08, "loss": 0.2735, "step": 9117 }, { "epoch": 2.9433794745550665, "grad_norm": 0.4022566217925706, "learning_rate": 1.4215368855374778e-08, "loss": 0.2848, "step": 9118 }, { "epoch": 2.9437023285846884, "grad_norm": 0.39443223985823406, "learning_rate": 1.4051555487873125e-08, "loss": 0.273, "step": 9119 }, { "epoch": 2.9440251826143102, "grad_norm": 0.33588957941721514, "learning_rate": 1.3888690579730856e-08, "loss": 0.2318, "step": 9120 }, { "epoch": 2.9443480366439325, "grad_norm": 0.3264890323690539, "learning_rate": 1.3726774151583132e-08, "loss": 0.2744, "step": 9121 }, { "epoch": 2.9446708906735544, "grad_norm": 0.3712821526406683, "learning_rate": 1.3565806223942711e-08, "loss": 0.2209, "step": 9122 }, { "epoch": 2.944993744703176, "grad_norm": 0.319400153609268, "learning_rate": 1.3405786817206612e-08, "loss": 0.2225, "step": 9123 }, { "epoch": 2.945316598732798, "grad_norm": 0.40381318008915074, "learning_rate": 1.3246715951646949e-08, "loss": 0.3316, "step": 9124 }, { "epoch": 2.94563945276242, "grad_norm": 0.3738474871242947, "learning_rate": 1.3088593647418435e-08, "loss": 0.2725, "step": 9125 }, { "epoch": 2.9459623067920417, "grad_norm": 0.3882402514010944, "learning_rate": 1.2931419924555043e-08, "loss": 0.2419, "step": 9126 }, { "epoch": 2.9462851608216636, "grad_norm": 0.40685078540924646, "learning_rate": 1.2775194802970015e-08, "loss": 0.2606, "step": 9127 }, { "epoch": 2.9466080148512854, "grad_norm": 0.38134638036810686, "learning_rate": 1.2619918302457511e-08, "loss": 0.2236, "step": 9128 }, { "epoch": 2.9469308688809073, "grad_norm": 0.36906012816036093, "learning_rate": 1.2465590442690966e-08, "loss": 0.3066, "step": 9129 }, { "epoch": 2.947253722910529, "grad_norm": 0.33327523465532827, "learning_rate": 1.231221124322307e-08, "loss": 0.2433, "step": 9130 }, { "epoch": 2.947576576940151, "grad_norm": 0.4101772816440727, "learning_rate": 1.2159780723486613e-08, "loss": 0.2692, "step": 9131 }, { "epoch": 2.947899430969773, "grad_norm": 0.35428846208224596, "learning_rate": 1.2008298902795311e-08, "loss": 0.2898, "step": 9132 }, { "epoch": 2.9482222849993946, "grad_norm": 0.3327472722069696, "learning_rate": 1.1857765800340482e-08, "loss": 0.2085, "step": 9133 }, { "epoch": 2.9485451390290165, "grad_norm": 0.37419429190576575, "learning_rate": 1.1708181435194366e-08, "loss": 0.2722, "step": 9134 }, { "epoch": 2.9488679930586383, "grad_norm": 0.3242564249044304, "learning_rate": 1.1559545826310137e-08, "loss": 0.2264, "step": 9135 }, { "epoch": 2.94919084708826, "grad_norm": 0.3966966243266253, "learning_rate": 1.1411858992519397e-08, "loss": 0.332, "step": 9136 }, { "epoch": 2.949513701117882, "grad_norm": 0.37297033383473366, "learning_rate": 1.1265120952533848e-08, "loss": 0.2571, "step": 9137 }, { "epoch": 2.949836555147504, "grad_norm": 0.3738956227204803, "learning_rate": 1.1119331724944449e-08, "loss": 0.2644, "step": 9138 }, { "epoch": 2.9501594091771257, "grad_norm": 0.3739840330977211, "learning_rate": 1.097449132822309e-08, "loss": 0.2389, "step": 9139 }, { "epoch": 2.9504822632067476, "grad_norm": 0.3725429330596136, "learning_rate": 1.0830599780720929e-08, "loss": 0.2833, "step": 9140 }, { "epoch": 2.9508051172363694, "grad_norm": 0.3699334685762278, "learning_rate": 1.068765710066838e-08, "loss": 0.2805, "step": 9141 }, { "epoch": 2.9511279712659912, "grad_norm": 0.3590204896401814, "learning_rate": 1.0545663306175956e-08, "loss": 0.2763, "step": 9142 }, { "epoch": 2.951450825295613, "grad_norm": 0.34322724021721024, "learning_rate": 1.04046184152351e-08, "loss": 0.2192, "step": 9143 }, { "epoch": 2.951773679325235, "grad_norm": 0.35920979000174286, "learning_rate": 1.0264522445714852e-08, "loss": 0.2859, "step": 9144 }, { "epoch": 2.9520965333548568, "grad_norm": 0.37276729294155936, "learning_rate": 1.012537541536518e-08, "loss": 0.265, "step": 9145 }, { "epoch": 2.952419387384479, "grad_norm": 0.34534428251034704, "learning_rate": 9.987177341816978e-09, "loss": 0.2787, "step": 9146 }, { "epoch": 2.952742241414101, "grad_norm": 0.3481935661877274, "learning_rate": 9.849928242578742e-09, "loss": 0.2146, "step": 9147 }, { "epoch": 2.9530650954437228, "grad_norm": 0.5650905866224113, "learning_rate": 9.713628135039898e-09, "loss": 0.2745, "step": 9148 }, { "epoch": 2.9533879494733446, "grad_norm": 0.35256590943813887, "learning_rate": 9.578277036469962e-09, "loss": 0.2061, "step": 9149 }, { "epoch": 2.9537108035029664, "grad_norm": 0.3577289960285122, "learning_rate": 9.443874964017718e-09, "loss": 0.204, "step": 9150 }, { "epoch": 2.9540336575325883, "grad_norm": 0.3448883698281043, "learning_rate": 9.31042193471121e-09, "loss": 0.2411, "step": 9151 }, { "epoch": 2.95435651156221, "grad_norm": 0.3726743758758021, "learning_rate": 9.177917965458582e-09, "loss": 0.2855, "step": 9152 }, { "epoch": 2.954679365591832, "grad_norm": 0.3915117775456217, "learning_rate": 9.046363073048069e-09, "loss": 0.2373, "step": 9153 }, { "epoch": 2.955002219621454, "grad_norm": 0.4084938066866495, "learning_rate": 8.915757274148006e-09, "loss": 0.2297, "step": 9154 }, { "epoch": 2.9553250736510757, "grad_norm": 0.4186061134164974, "learning_rate": 8.786100585305989e-09, "loss": 0.3399, "step": 9155 }, { "epoch": 2.9556479276806975, "grad_norm": 0.3863484112742662, "learning_rate": 8.65739302294888e-09, "loss": 0.2962, "step": 9156 }, { "epoch": 2.9559707817103194, "grad_norm": 0.3491622581487061, "learning_rate": 8.529634603383629e-09, "loss": 0.2395, "step": 9157 }, { "epoch": 2.956293635739941, "grad_norm": 0.3631523520745583, "learning_rate": 8.402825342797294e-09, "loss": 0.2702, "step": 9158 }, { "epoch": 2.956616489769563, "grad_norm": 0.37086299009228485, "learning_rate": 8.27696525725702e-09, "loss": 0.2409, "step": 9159 }, { "epoch": 2.956939343799185, "grad_norm": 0.4264153581848314, "learning_rate": 8.152054362707551e-09, "loss": 0.2558, "step": 9160 }, { "epoch": 2.9572621978288067, "grad_norm": 0.3892516924922193, "learning_rate": 8.028092674976229e-09, "loss": 0.2699, "step": 9161 }, { "epoch": 2.9575850518584286, "grad_norm": 0.40717674183074626, "learning_rate": 7.905080209768822e-09, "loss": 0.2487, "step": 9162 }, { "epoch": 2.9579079058880504, "grad_norm": 0.39461814187041355, "learning_rate": 7.783016982670366e-09, "loss": 0.2703, "step": 9163 }, { "epoch": 2.9582307599176723, "grad_norm": 0.37905327117031007, "learning_rate": 7.661903009146821e-09, "loss": 0.3268, "step": 9164 }, { "epoch": 2.958553613947294, "grad_norm": 0.3714186008889184, "learning_rate": 7.541738304541746e-09, "loss": 0.2152, "step": 9165 }, { "epoch": 2.958876467976916, "grad_norm": 0.3250084963106742, "learning_rate": 7.422522884081295e-09, "loss": 0.3125, "step": 9166 }, { "epoch": 2.959199322006538, "grad_norm": 0.33795606212237156, "learning_rate": 7.304256762868389e-09, "loss": 0.2779, "step": 9167 }, { "epoch": 2.9595221760361596, "grad_norm": 0.40106004471056217, "learning_rate": 7.186939955889371e-09, "loss": 0.2543, "step": 9168 }, { "epoch": 2.9598450300657815, "grad_norm": 0.40921986990580583, "learning_rate": 7.070572478005688e-09, "loss": 0.2102, "step": 9169 }, { "epoch": 2.9601678840954033, "grad_norm": 0.3604769622716221, "learning_rate": 6.955154343963044e-09, "loss": 0.2607, "step": 9170 }, { "epoch": 2.960490738125025, "grad_norm": 0.313187603580833, "learning_rate": 6.840685568383076e-09, "loss": 0.2216, "step": 9171 }, { "epoch": 2.960813592154647, "grad_norm": 0.39393225598890835, "learning_rate": 6.727166165770016e-09, "loss": 0.3135, "step": 9172 }, { "epoch": 2.961136446184269, "grad_norm": 0.39207724147961937, "learning_rate": 6.614596150505691e-09, "loss": 0.2102, "step": 9173 }, { "epoch": 2.9614593002138907, "grad_norm": 0.4212372067237154, "learning_rate": 6.502975536853695e-09, "loss": 0.2474, "step": 9174 }, { "epoch": 2.9617821542435125, "grad_norm": 0.417914150436947, "learning_rate": 6.392304338955213e-09, "loss": 0.2625, "step": 9175 }, { "epoch": 2.9621050082731344, "grad_norm": 0.38965810301319825, "learning_rate": 6.282582570832363e-09, "loss": 0.3134, "step": 9176 }, { "epoch": 2.9624278623027562, "grad_norm": 0.31919301982611314, "learning_rate": 6.173810246386525e-09, "loss": 0.2553, "step": 9177 }, { "epoch": 2.962750716332378, "grad_norm": 0.43452111221206163, "learning_rate": 6.065987379400006e-09, "loss": 0.2409, "step": 9178 }, { "epoch": 2.963073570362, "grad_norm": 0.3888922127950601, "learning_rate": 5.9591139835327156e-09, "loss": 0.2992, "step": 9179 }, { "epoch": 2.9633964243916218, "grad_norm": 0.3661505699070527, "learning_rate": 5.853190072325487e-09, "loss": 0.2504, "step": 9180 }, { "epoch": 2.9637192784212436, "grad_norm": 0.3157040690433249, "learning_rate": 5.7482156592000844e-09, "loss": 0.2516, "step": 9181 }, { "epoch": 2.9640421324508655, "grad_norm": 0.3857416092069441, "learning_rate": 5.644190757454204e-09, "loss": 0.2539, "step": 9182 }, { "epoch": 2.9643649864804873, "grad_norm": 0.390550171468176, "learning_rate": 5.541115380269801e-09, "loss": 0.2579, "step": 9183 }, { "epoch": 2.964687840510109, "grad_norm": 0.4089706504736463, "learning_rate": 5.438989540705597e-09, "loss": 0.2198, "step": 9184 }, { "epoch": 2.965010694539731, "grad_norm": 0.34333699971401743, "learning_rate": 5.3378132517004075e-09, "loss": 0.2787, "step": 9185 }, { "epoch": 2.965333548569353, "grad_norm": 0.3266836376615198, "learning_rate": 5.237586526073979e-09, "loss": 0.2487, "step": 9186 }, { "epoch": 2.9656564025989747, "grad_norm": 0.3829994271879467, "learning_rate": 5.138309376524486e-09, "loss": 0.2909, "step": 9187 }, { "epoch": 2.965979256628597, "grad_norm": 0.41885177245974886, "learning_rate": 5.039981815629368e-09, "loss": 0.2392, "step": 9188 }, { "epoch": 2.966302110658219, "grad_norm": 0.34621562032095216, "learning_rate": 4.942603855847827e-09, "loss": 0.2314, "step": 9189 }, { "epoch": 2.9666249646878406, "grad_norm": 0.3840074164618648, "learning_rate": 4.846175509517492e-09, "loss": 0.2865, "step": 9190 }, { "epoch": 2.9669478187174625, "grad_norm": 0.3459943739900793, "learning_rate": 4.7506967888552574e-09, "loss": 0.2748, "step": 9191 }, { "epoch": 2.9672706727470843, "grad_norm": 0.3644769753124676, "learning_rate": 4.656167705957281e-09, "loss": 0.2811, "step": 9192 }, { "epoch": 2.967593526776706, "grad_norm": 0.3138750983803602, "learning_rate": 4.562588272802315e-09, "loss": 0.2584, "step": 9193 }, { "epoch": 2.967916380806328, "grad_norm": 0.36777546930506805, "learning_rate": 4.469958501245041e-09, "loss": 0.2835, "step": 9194 }, { "epoch": 2.96823923483595, "grad_norm": 0.33657970219837297, "learning_rate": 4.378278403021907e-09, "loss": 0.269, "step": 9195 }, { "epoch": 2.9685620888655717, "grad_norm": 0.3489349411032428, "learning_rate": 4.287547989748619e-09, "loss": 0.303, "step": 9196 }, { "epoch": 2.9688849428951936, "grad_norm": 0.3764264134664114, "learning_rate": 4.197767272920983e-09, "loss": 0.2704, "step": 9197 }, { "epoch": 2.9692077969248154, "grad_norm": 0.44343754266521296, "learning_rate": 4.108936263914065e-09, "loss": 0.3449, "step": 9198 }, { "epoch": 2.9695306509544372, "grad_norm": 0.3623467988712096, "learning_rate": 4.021054973982197e-09, "loss": 0.2515, "step": 9199 }, { "epoch": 2.969853504984059, "grad_norm": 0.3736728452481022, "learning_rate": 3.934123414260637e-09, "loss": 0.2619, "step": 9200 }, { "epoch": 2.970176359013681, "grad_norm": 0.39125854841725427, "learning_rate": 3.848141595762245e-09, "loss": 0.2667, "step": 9201 }, { "epoch": 2.9704992130433028, "grad_norm": 0.41098304305777783, "learning_rate": 3.763109529382469e-09, "loss": 0.2656, "step": 9202 }, { "epoch": 2.9708220670729246, "grad_norm": 0.37041767963675637, "learning_rate": 3.6790272258926948e-09, "loss": 0.2997, "step": 9203 }, { "epoch": 2.9711449211025465, "grad_norm": 0.366734338441636, "learning_rate": 3.5958946959477323e-09, "loss": 0.2728, "step": 9204 }, { "epoch": 2.9714677751321683, "grad_norm": 0.41712223398138687, "learning_rate": 3.513711950079157e-09, "loss": 0.2738, "step": 9205 }, { "epoch": 2.97179062916179, "grad_norm": 0.40432245253138693, "learning_rate": 3.4324789987011385e-09, "loss": 0.2288, "step": 9206 }, { "epoch": 2.972113483191412, "grad_norm": 0.33736152979784567, "learning_rate": 3.3521958521037788e-09, "loss": 0.2575, "step": 9207 }, { "epoch": 2.972436337221034, "grad_norm": 0.3934257075467065, "learning_rate": 3.272862520459774e-09, "loss": 0.2549, "step": 9208 }, { "epoch": 2.9727591912506557, "grad_norm": 0.3254713826659799, "learning_rate": 3.1944790138202508e-09, "loss": 0.2813, "step": 9209 }, { "epoch": 2.9730820452802775, "grad_norm": 0.34858223032069047, "learning_rate": 3.117045342117264e-09, "loss": 0.2724, "step": 9210 }, { "epoch": 2.9734048993098994, "grad_norm": 0.4295351457881493, "learning_rate": 3.040561515160467e-09, "loss": 0.2529, "step": 9211 }, { "epoch": 2.973727753339521, "grad_norm": 0.32267185251842057, "learning_rate": 2.9650275426404414e-09, "loss": 0.2183, "step": 9212 }, { "epoch": 2.974050607369143, "grad_norm": 0.38295861720997, "learning_rate": 2.8904434341270324e-09, "loss": 0.2367, "step": 9213 }, { "epoch": 2.9743734613987654, "grad_norm": 0.3798667922162063, "learning_rate": 2.8168091990710134e-09, "loss": 0.2488, "step": 9214 }, { "epoch": 2.974696315428387, "grad_norm": 0.3641610342012179, "learning_rate": 2.744124846800755e-09, "loss": 0.2244, "step": 9215 }, { "epoch": 2.975019169458009, "grad_norm": 0.4411612571736725, "learning_rate": 2.6723903865255585e-09, "loss": 0.2524, "step": 9216 }, { "epoch": 2.975342023487631, "grad_norm": 0.605746796602594, "learning_rate": 2.601605827333986e-09, "loss": 0.2674, "step": 9217 }, { "epoch": 2.9756648775172527, "grad_norm": 0.36312987984019973, "learning_rate": 2.5317711781946975e-09, "loss": 0.2448, "step": 9218 }, { "epoch": 2.9759877315468746, "grad_norm": 0.3788477411199751, "learning_rate": 2.462886447954782e-09, "loss": 0.3426, "step": 9219 }, { "epoch": 2.9763105855764964, "grad_norm": 0.3038057809067203, "learning_rate": 2.394951645343091e-09, "loss": 0.2379, "step": 9220 }, { "epoch": 2.9766334396061183, "grad_norm": 0.30316070989777966, "learning_rate": 2.327966778966073e-09, "loss": 0.2548, "step": 9221 }, { "epoch": 2.97695629363574, "grad_norm": 0.37581635546543357, "learning_rate": 2.2619318573102732e-09, "loss": 0.2217, "step": 9222 }, { "epoch": 2.977279147665362, "grad_norm": 0.40040265196736274, "learning_rate": 2.1968468887431647e-09, "loss": 0.2666, "step": 9223 }, { "epoch": 2.977602001694984, "grad_norm": 0.41352727031670977, "learning_rate": 2.1327118815106517e-09, "loss": 0.2332, "step": 9224 }, { "epoch": 2.9779248557246056, "grad_norm": 0.4273370302422266, "learning_rate": 2.069526843737901e-09, "loss": 0.2975, "step": 9225 }, { "epoch": 2.9782477097542275, "grad_norm": 0.40474534606595985, "learning_rate": 2.0072917834310088e-09, "loss": 0.2212, "step": 9226 }, { "epoch": 2.9785705637838493, "grad_norm": 0.4045415089278613, "learning_rate": 1.946006708475334e-09, "loss": 0.2531, "step": 9227 }, { "epoch": 2.978893417813471, "grad_norm": 0.4238724966790232, "learning_rate": 1.8856716266346663e-09, "loss": 0.24, "step": 9228 }, { "epoch": 2.979216271843093, "grad_norm": 0.36269937134604996, "learning_rate": 1.8262865455537237e-09, "loss": 0.1995, "step": 9229 }, { "epoch": 2.979539125872715, "grad_norm": 0.4011246917907336, "learning_rate": 1.7678514727564876e-09, "loss": 0.2686, "step": 9230 }, { "epoch": 2.9798619799023367, "grad_norm": 0.4019648319501467, "learning_rate": 1.7103664156478681e-09, "loss": 0.1917, "step": 9231 }, { "epoch": 2.9801848339319585, "grad_norm": 0.31605411431500907, "learning_rate": 1.653831381508708e-09, "loss": 0.2267, "step": 9232 }, { "epoch": 2.9805076879615804, "grad_norm": 0.405580240096691, "learning_rate": 1.598246377504109e-09, "loss": 0.2397, "step": 9233 }, { "epoch": 2.9808305419912022, "grad_norm": 0.36256915916324944, "learning_rate": 1.543611410675938e-09, "loss": 0.2908, "step": 9234 }, { "epoch": 2.981153396020824, "grad_norm": 0.3748188278411586, "learning_rate": 1.4899264879461582e-09, "loss": 0.2473, "step": 9235 }, { "epoch": 2.981476250050446, "grad_norm": 0.3596843203960296, "learning_rate": 1.4371916161159958e-09, "loss": 0.2554, "step": 9236 }, { "epoch": 2.9817991040800678, "grad_norm": 0.3225428114365135, "learning_rate": 1.3854068018676058e-09, "loss": 0.2572, "step": 9237 }, { "epoch": 2.9821219581096896, "grad_norm": 0.4253308567973335, "learning_rate": 1.334572051762406e-09, "loss": 0.3089, "step": 9238 }, { "epoch": 2.9824448121393115, "grad_norm": 0.4036262830704453, "learning_rate": 1.2846873722410779e-09, "loss": 0.2609, "step": 9239 }, { "epoch": 2.9827676661689333, "grad_norm": 0.3573223944416433, "learning_rate": 1.2357527696227334e-09, "loss": 0.2713, "step": 9240 }, { "epoch": 2.983090520198555, "grad_norm": 0.3720939925283301, "learning_rate": 1.1877682501082455e-09, "loss": 0.2799, "step": 9241 }, { "epoch": 2.983413374228177, "grad_norm": 0.38382218806211604, "learning_rate": 1.140733819776918e-09, "loss": 0.238, "step": 9242 }, { "epoch": 2.983736228257799, "grad_norm": 0.411379786174343, "learning_rate": 1.0946494845881505e-09, "loss": 0.3125, "step": 9243 }, { "epoch": 2.9840590822874207, "grad_norm": 0.3488593425244811, "learning_rate": 1.0495152503806061e-09, "loss": 0.2686, "step": 9244 }, { "epoch": 2.9843819363170425, "grad_norm": 0.4044628148109259, "learning_rate": 1.0053311228730432e-09, "loss": 0.2908, "step": 9245 }, { "epoch": 2.9847047903466644, "grad_norm": 0.3040547278359496, "learning_rate": 9.620971076634844e-10, "loss": 0.1965, "step": 9246 }, { "epoch": 2.985027644376286, "grad_norm": 0.3221945766006144, "learning_rate": 9.198132102292145e-10, "loss": 0.2584, "step": 9247 }, { "epoch": 2.985350498405908, "grad_norm": 0.32485937453619634, "learning_rate": 8.784794359276149e-10, "loss": 0.2426, "step": 9248 }, { "epoch": 2.98567335243553, "grad_norm": 0.35044382353185055, "learning_rate": 8.380957899961628e-10, "loss": 0.3014, "step": 9249 }, { "epoch": 2.9859962064651517, "grad_norm": 0.38761728369051385, "learning_rate": 7.986622775507657e-10, "loss": 0.2082, "step": 9250 }, { "epoch": 2.9863190604947736, "grad_norm": 0.4027044719770335, "learning_rate": 7.601789035874273e-10, "loss": 0.2283, "step": 9251 }, { "epoch": 2.9866419145243954, "grad_norm": 0.40364709675638594, "learning_rate": 7.2264567298308e-10, "loss": 0.2181, "step": 9252 }, { "epoch": 2.9869647685540173, "grad_norm": 0.37825232493605176, "learning_rate": 6.860625904922535e-10, "loss": 0.2782, "step": 9253 }, { "epoch": 2.987287622583639, "grad_norm": 0.37268446216375845, "learning_rate": 6.504296607495741e-10, "loss": 0.2222, "step": 9254 }, { "epoch": 2.987610476613261, "grad_norm": 0.352673938328673, "learning_rate": 6.157468882705963e-10, "loss": 0.3019, "step": 9255 }, { "epoch": 2.9879333306428832, "grad_norm": 0.3094770588689358, "learning_rate": 5.820142774493053e-10, "loss": 0.271, "step": 9256 }, { "epoch": 2.988256184672505, "grad_norm": 0.4018461389929417, "learning_rate": 5.492318325597822e-10, "loss": 0.2535, "step": 9257 }, { "epoch": 2.988579038702127, "grad_norm": 0.43086157242236883, "learning_rate": 5.173995577545387e-10, "loss": 0.274, "step": 9258 }, { "epoch": 2.9889018927317488, "grad_norm": 0.33413956225273134, "learning_rate": 4.865174570678476e-10, "loss": 0.224, "step": 9259 }, { "epoch": 2.9892247467613706, "grad_norm": 0.381436211534585, "learning_rate": 4.565855344115799e-10, "loss": 0.2659, "step": 9260 }, { "epoch": 2.9895476007909925, "grad_norm": 0.34663391380103625, "learning_rate": 4.2760379357853484e-10, "loss": 0.2656, "step": 9261 }, { "epoch": 2.9898704548206143, "grad_norm": 0.4129241896323359, "learning_rate": 3.9957223824077517e-10, "loss": 0.2724, "step": 9262 }, { "epoch": 2.990193308850236, "grad_norm": 0.3819708795641599, "learning_rate": 3.7249087194962673e-10, "loss": 0.2895, "step": 9263 }, { "epoch": 2.990516162879858, "grad_norm": 0.3870439434154613, "learning_rate": 3.463596981365114e-10, "loss": 0.202, "step": 9264 }, { "epoch": 2.99083901690948, "grad_norm": 0.3655116020877851, "learning_rate": 3.2117872011128144e-10, "loss": 0.2097, "step": 9265 }, { "epoch": 2.9911618709391017, "grad_norm": 0.3876211565952942, "learning_rate": 2.9694794106555066e-10, "loss": 0.2787, "step": 9266 }, { "epoch": 2.9914847249687235, "grad_norm": 0.42426891832041413, "learning_rate": 2.7366736406853057e-10, "loss": 0.2207, "step": 9267 }, { "epoch": 2.9918075789983454, "grad_norm": 0.34051819384837617, "learning_rate": 2.5133699207036123e-10, "loss": 0.2697, "step": 9268 }, { "epoch": 2.992130433027967, "grad_norm": 0.3900718134224035, "learning_rate": 2.2995682790044603e-10, "loss": 0.2025, "step": 9269 }, { "epoch": 2.992453287057589, "grad_norm": 0.3811111099522623, "learning_rate": 2.09526874266619e-10, "loss": 0.2567, "step": 9270 }, { "epoch": 2.992776141087211, "grad_norm": 0.38769221486299965, "learning_rate": 1.9004713375764259e-10, "loss": 0.281, "step": 9271 }, { "epoch": 2.9930989951168328, "grad_norm": 0.4450157243964105, "learning_rate": 1.7151760884237533e-10, "loss": 0.3153, "step": 9272 }, { "epoch": 2.9934218491464546, "grad_norm": 0.38253177303672825, "learning_rate": 1.539383018681062e-10, "loss": 0.2374, "step": 9273 }, { "epoch": 2.9937447031760764, "grad_norm": 0.4600522810147365, "learning_rate": 1.3730921506138748e-10, "loss": 0.2484, "step": 9274 }, { "epoch": 2.9940675572056983, "grad_norm": 0.4032721215873999, "learning_rate": 1.2163035052970005e-10, "loss": 0.2468, "step": 9275 }, { "epoch": 2.99439041123532, "grad_norm": 0.4108615095015119, "learning_rate": 1.06901710259788e-10, "loss": 0.2424, "step": 9276 }, { "epoch": 2.994713265264942, "grad_norm": 0.3831445323112701, "learning_rate": 9.312329611765869e-11, "loss": 0.3001, "step": 9277 }, { "epoch": 2.995036119294564, "grad_norm": 0.413502187405372, "learning_rate": 8.029510984858268e-11, "loss": 0.2346, "step": 9278 }, { "epoch": 2.9953589733241857, "grad_norm": 0.3788972671444472, "learning_rate": 6.841715307792651e-11, "loss": 0.285, "step": 9279 }, { "epoch": 2.9956818273538075, "grad_norm": 0.3988186018276293, "learning_rate": 5.748942731031992e-11, "loss": 0.3124, "step": 9280 }, { "epoch": 2.99600468138343, "grad_norm": 0.3874704903893047, "learning_rate": 4.7511933931321247e-11, "loss": 0.2353, "step": 9281 }, { "epoch": 2.9963275354130516, "grad_norm": 0.5650157520592216, "learning_rate": 3.8484674204919414e-11, "loss": 0.2596, "step": 9282 }, { "epoch": 2.9966503894426735, "grad_norm": 0.3585368192630899, "learning_rate": 3.040764927436657e-11, "loss": 0.2582, "step": 9283 }, { "epoch": 2.9969732434722953, "grad_norm": 0.3725851301530421, "learning_rate": 2.3280860162178118e-11, "loss": 0.2849, "step": 9284 }, { "epoch": 2.997296097501917, "grad_norm": 0.4348929421003764, "learning_rate": 1.710430777346339e-11, "loss": 0.2464, "step": 9285 }, { "epoch": 2.997618951531539, "grad_norm": 0.42941353803555143, "learning_rate": 1.1877992888431611e-11, "loss": 0.2255, "step": 9286 }, { "epoch": 2.997941805561161, "grad_norm": 0.3546643219969446, "learning_rate": 7.60191617155126e-12, "loss": 0.2384, "step": 9287 }, { "epoch": 2.9982646595907827, "grad_norm": 0.408687494159154, "learning_rate": 4.276078162390728e-12, "loss": 0.2418, "step": 9288 }, { "epoch": 2.9985875136204045, "grad_norm": 0.4044799962886107, "learning_rate": 1.900479283944989e-12, "loss": 0.2216, "step": 9289 }, { "epoch": 2.9989103676500264, "grad_norm": 0.3744559916820863, "learning_rate": 4.751198359742581e-13, "loss": 0.3165, "step": 9290 }, { "epoch": 2.9992332216796482, "grad_norm": 0.33193961374566666, "learning_rate": 0.0, "loss": 0.2566, "step": 9291 } ], "logging_steps": 1.0, "max_steps": 9291, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.1233140673802934e+22, "train_batch_size": 2, "trial_name": null, "trial_params": null }