| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 31.372549019607842, | |
| "eval_steps": 500, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 7.629533767700195, | |
| "learning_rate": 1e-05, | |
| "loss": 4.321, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 14.44616985321045, | |
| "learning_rate": 2e-05, | |
| "loss": 4.2862, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 19.094482421875, | |
| "learning_rate": 3e-05, | |
| "loss": 4.3249, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 11.822478294372559, | |
| "learning_rate": 4e-05, | |
| "loss": 4.249, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.39215686274509803, | |
| "grad_norm": 16.548583984375, | |
| "learning_rate": 5e-05, | |
| "loss": 4.5841, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 7.463245391845703, | |
| "learning_rate": 6e-05, | |
| "loss": 4.3256, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5490196078431373, | |
| "grad_norm": 9.122880935668945, | |
| "learning_rate": 7e-05, | |
| "loss": 4.391, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 18.88115119934082, | |
| "learning_rate": 8e-05, | |
| "loss": 4.1049, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 8.104399681091309, | |
| "learning_rate": 9e-05, | |
| "loss": 4.1016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 8.780770301818848, | |
| "learning_rate": 0.0001, | |
| "loss": 4.2899, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8627450980392157, | |
| "grad_norm": 8.454495429992676, | |
| "learning_rate": 9.996324205109356e-05, | |
| "loss": 4.1025, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 9.97666072845459, | |
| "learning_rate": 9.992648410218711e-05, | |
| "loss": 4.1319, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0196078431372548, | |
| "grad_norm": 7.962587833404541, | |
| "learning_rate": 9.988972615328065e-05, | |
| "loss": 3.9871, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0980392156862746, | |
| "grad_norm": 8.972702026367188, | |
| "learning_rate": 9.98529682043742e-05, | |
| "loss": 3.8708, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 11.219219207763672, | |
| "learning_rate": 9.981621025546775e-05, | |
| "loss": 3.838, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2549019607843137, | |
| "grad_norm": 12.35175609588623, | |
| "learning_rate": 9.97794523065613e-05, | |
| "loss": 3.8339, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 8.549275398254395, | |
| "learning_rate": 9.974269435765484e-05, | |
| "loss": 3.7836, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 14.15541934967041, | |
| "learning_rate": 9.97059364087484e-05, | |
| "loss": 3.6838, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.4901960784313726, | |
| "grad_norm": 16.026613235473633, | |
| "learning_rate": 9.966917845984195e-05, | |
| "loss": 3.6949, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5686274509803921, | |
| "grad_norm": 10.782464981079102, | |
| "learning_rate": 9.963242051093549e-05, | |
| "loss": 3.636, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 10.165632247924805, | |
| "learning_rate": 9.959566256202904e-05, | |
| "loss": 3.4669, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.7254901960784315, | |
| "grad_norm": 13.020809173583984, | |
| "learning_rate": 9.95589046131226e-05, | |
| "loss": 3.5055, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.803921568627451, | |
| "grad_norm": 8.814240455627441, | |
| "learning_rate": 9.952214666421615e-05, | |
| "loss": 3.6938, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 18.580322265625, | |
| "learning_rate": 9.94853887153097e-05, | |
| "loss": 3.7736, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9607843137254903, | |
| "grad_norm": 22.766721725463867, | |
| "learning_rate": 9.944863076640323e-05, | |
| "loss": 3.4302, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0392156862745097, | |
| "grad_norm": 11.897246360778809, | |
| "learning_rate": 9.941187281749679e-05, | |
| "loss": 3.4821, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 12.075397491455078, | |
| "learning_rate": 9.937511486859034e-05, | |
| "loss": 3.1794, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.196078431372549, | |
| "grad_norm": 12.672398567199707, | |
| "learning_rate": 9.933835691968389e-05, | |
| "loss": 3.0642, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.2745098039215685, | |
| "grad_norm": 8.902212142944336, | |
| "learning_rate": 9.930159897077743e-05, | |
| "loss": 3.4314, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 10.514148712158203, | |
| "learning_rate": 9.926484102187098e-05, | |
| "loss": 3.298, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.431372549019608, | |
| "grad_norm": 12.432036399841309, | |
| "learning_rate": 9.922808307296453e-05, | |
| "loss": 3.2431, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.5098039215686274, | |
| "grad_norm": 11.593744277954102, | |
| "learning_rate": 9.91913251240581e-05, | |
| "loss": 3.048, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 11.159161567687988, | |
| "learning_rate": 9.915456717515163e-05, | |
| "loss": 3.1779, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 8.838184356689453, | |
| "learning_rate": 9.911780922624518e-05, | |
| "loss": 3.4323, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.7450980392156863, | |
| "grad_norm": 10.507558822631836, | |
| "learning_rate": 9.908105127733873e-05, | |
| "loss": 3.291, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 9.107258796691895, | |
| "learning_rate": 9.904429332843229e-05, | |
| "loss": 3.3132, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.9019607843137254, | |
| "grad_norm": 9.832110404968262, | |
| "learning_rate": 9.900753537952582e-05, | |
| "loss": 3.2163, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.980392156862745, | |
| "grad_norm": 8.222602844238281, | |
| "learning_rate": 9.897077743061938e-05, | |
| "loss": 3.478, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 8.719328880310059, | |
| "learning_rate": 9.893401948171293e-05, | |
| "loss": 3.0145, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.1372549019607843, | |
| "grad_norm": 11.51774787902832, | |
| "learning_rate": 9.889726153280648e-05, | |
| "loss": 3.0454, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.215686274509804, | |
| "grad_norm": 10.710402488708496, | |
| "learning_rate": 9.886050358390002e-05, | |
| "loss": 2.7929, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 13.097260475158691, | |
| "learning_rate": 9.882374563499357e-05, | |
| "loss": 2.9186, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.372549019607843, | |
| "grad_norm": 15.856924057006836, | |
| "learning_rate": 9.878698768608712e-05, | |
| "loss": 2.9282, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.450980392156863, | |
| "grad_norm": 11.797476768493652, | |
| "learning_rate": 9.875022973718067e-05, | |
| "loss": 2.891, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 9.920392990112305, | |
| "learning_rate": 9.871347178827422e-05, | |
| "loss": 2.8574, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.607843137254902, | |
| "grad_norm": 8.427005767822266, | |
| "learning_rate": 9.867671383936776e-05, | |
| "loss": 2.9527, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.686274509803922, | |
| "grad_norm": 12.288470268249512, | |
| "learning_rate": 9.863995589046132e-05, | |
| "loss": 2.7286, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 8.801328659057617, | |
| "learning_rate": 9.860319794155488e-05, | |
| "loss": 2.9276, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.843137254901961, | |
| "grad_norm": 8.852784156799316, | |
| "learning_rate": 9.856643999264841e-05, | |
| "loss": 2.9347, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.9215686274509802, | |
| "grad_norm": 19.973102569580078, | |
| "learning_rate": 9.852968204374196e-05, | |
| "loss": 3.0643, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 10.647859573364258, | |
| "learning_rate": 9.849292409483552e-05, | |
| "loss": 3.0444, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.078431372549019, | |
| "grad_norm": 17.03227424621582, | |
| "learning_rate": 9.845616614592907e-05, | |
| "loss": 2.6609, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.1568627450980395, | |
| "grad_norm": 11.255087852478027, | |
| "learning_rate": 9.84194081970226e-05, | |
| "loss": 2.5831, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 7.963006019592285, | |
| "learning_rate": 9.838265024811616e-05, | |
| "loss": 2.6449, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.313725490196078, | |
| "grad_norm": 8.0176420211792, | |
| "learning_rate": 9.834589229920971e-05, | |
| "loss": 2.6803, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.392156862745098, | |
| "grad_norm": 12.322718620300293, | |
| "learning_rate": 9.830913435030326e-05, | |
| "loss": 2.6856, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 13.681117057800293, | |
| "learning_rate": 9.827237640139681e-05, | |
| "loss": 2.6411, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.549019607843137, | |
| "grad_norm": 9.620941162109375, | |
| "learning_rate": 9.823561845249035e-05, | |
| "loss": 2.6757, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.627450980392156, | |
| "grad_norm": 8.198792457580566, | |
| "learning_rate": 9.81988605035839e-05, | |
| "loss": 2.5249, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 11.216117858886719, | |
| "learning_rate": 9.816210255467745e-05, | |
| "loss": 2.5542, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.784313725490196, | |
| "grad_norm": 13.534390449523926, | |
| "learning_rate": 9.8125344605771e-05, | |
| "loss": 2.5632, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.862745098039216, | |
| "grad_norm": 8.238523483276367, | |
| "learning_rate": 9.808858665686455e-05, | |
| "loss": 2.7125, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 7.748435974121094, | |
| "learning_rate": 9.80518287079581e-05, | |
| "loss": 2.5738, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.019607843137255, | |
| "grad_norm": 8.043440818786621, | |
| "learning_rate": 9.801507075905166e-05, | |
| "loss": 2.3469, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.098039215686274, | |
| "grad_norm": 14.42237377166748, | |
| "learning_rate": 9.797831281014521e-05, | |
| "loss": 2.2214, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 7.59706974029541, | |
| "learning_rate": 9.794155486123875e-05, | |
| "loss": 2.279, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.254901960784314, | |
| "grad_norm": 9.453704833984375, | |
| "learning_rate": 9.79047969123323e-05, | |
| "loss": 2.4156, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 8.243208885192871, | |
| "learning_rate": 9.786803896342585e-05, | |
| "loss": 2.2599, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 7.8015971183776855, | |
| "learning_rate": 9.78312810145194e-05, | |
| "loss": 2.1746, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.490196078431373, | |
| "grad_norm": 12.369669914245605, | |
| "learning_rate": 9.779452306561294e-05, | |
| "loss": 2.2099, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.568627450980392, | |
| "grad_norm": 7.153007507324219, | |
| "learning_rate": 9.775776511670649e-05, | |
| "loss": 2.4386, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 11.485706329345703, | |
| "learning_rate": 9.772100716780004e-05, | |
| "loss": 2.2379, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 5.7254901960784315, | |
| "grad_norm": 8.15987777709961, | |
| "learning_rate": 9.768424921889359e-05, | |
| "loss": 2.4642, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 5.803921568627451, | |
| "grad_norm": 8.703965187072754, | |
| "learning_rate": 9.764749126998713e-05, | |
| "loss": 2.1796, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 8.410348892211914, | |
| "learning_rate": 9.761073332108068e-05, | |
| "loss": 2.3248, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 5.96078431372549, | |
| "grad_norm": 7.395628929138184, | |
| "learning_rate": 9.757397537217425e-05, | |
| "loss": 2.3482, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.03921568627451, | |
| "grad_norm": 7.840580940246582, | |
| "learning_rate": 9.75372174232678e-05, | |
| "loss": 2.3581, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 7.0679779052734375, | |
| "learning_rate": 9.750045947436133e-05, | |
| "loss": 2.1624, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.196078431372549, | |
| "grad_norm": 8.131471633911133, | |
| "learning_rate": 9.746370152545489e-05, | |
| "loss": 1.9443, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.2745098039215685, | |
| "grad_norm": 7.165848255157471, | |
| "learning_rate": 9.742694357654844e-05, | |
| "loss": 1.8785, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 7.661879062652588, | |
| "learning_rate": 9.739018562764199e-05, | |
| "loss": 2.0509, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 6.431372549019608, | |
| "grad_norm": 10.607108116149902, | |
| "learning_rate": 9.735342767873553e-05, | |
| "loss": 1.994, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 6.509803921568627, | |
| "grad_norm": 7.981103420257568, | |
| "learning_rate": 9.731666972982908e-05, | |
| "loss": 1.8272, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 8.540278434753418, | |
| "learning_rate": 9.727991178092263e-05, | |
| "loss": 2.0337, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 6.572484493255615, | |
| "learning_rate": 9.724315383201618e-05, | |
| "loss": 1.8927, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 6.745098039215686, | |
| "grad_norm": 6.641257286071777, | |
| "learning_rate": 9.720639588310972e-05, | |
| "loss": 2.1526, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 8.230134010314941, | |
| "learning_rate": 9.716963793420327e-05, | |
| "loss": 2.1622, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 6.901960784313726, | |
| "grad_norm": 8.205769538879395, | |
| "learning_rate": 9.713287998529682e-05, | |
| "loss": 2.3578, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 6.980392156862745, | |
| "grad_norm": 8.910941123962402, | |
| "learning_rate": 9.709612203639037e-05, | |
| "loss": 2.1729, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 6.27609920501709, | |
| "learning_rate": 9.705936408748392e-05, | |
| "loss": 1.8849, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.137254901960785, | |
| "grad_norm": 8.897392272949219, | |
| "learning_rate": 9.702260613857747e-05, | |
| "loss": 1.7128, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 7.215686274509804, | |
| "grad_norm": 8.366013526916504, | |
| "learning_rate": 9.698584818967103e-05, | |
| "loss": 1.7241, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 6.907725811004639, | |
| "learning_rate": 9.694909024076458e-05, | |
| "loss": 1.704, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 7.372549019607844, | |
| "grad_norm": 7.610422611236572, | |
| "learning_rate": 9.691233229185812e-05, | |
| "loss": 1.7402, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 7.450980392156863, | |
| "grad_norm": 6.888455390930176, | |
| "learning_rate": 9.687557434295167e-05, | |
| "loss": 1.9821, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 10.415329933166504, | |
| "learning_rate": 9.683881639404522e-05, | |
| "loss": 1.7391, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 7.607843137254902, | |
| "grad_norm": 7.084939956665039, | |
| "learning_rate": 9.680205844513877e-05, | |
| "loss": 1.8511, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 7.686274509803922, | |
| "grad_norm": 6.506997108459473, | |
| "learning_rate": 9.676530049623231e-05, | |
| "loss": 1.7976, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 6.158507823944092, | |
| "learning_rate": 9.672854254732586e-05, | |
| "loss": 1.6912, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 7.8431372549019605, | |
| "grad_norm": 7.078491687774658, | |
| "learning_rate": 9.669178459841941e-05, | |
| "loss": 1.9965, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.921568627450981, | |
| "grad_norm": 8.607373237609863, | |
| "learning_rate": 9.665502664951296e-05, | |
| "loss": 1.7438, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.7270073890686035, | |
| "learning_rate": 9.661826870060651e-05, | |
| "loss": 1.9924, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 8.07843137254902, | |
| "grad_norm": 7.503419876098633, | |
| "learning_rate": 9.658151075170005e-05, | |
| "loss": 1.3828, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 8.156862745098039, | |
| "grad_norm": 4.802187442779541, | |
| "learning_rate": 9.65447528027936e-05, | |
| "loss": 1.5968, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 6.051253318786621, | |
| "learning_rate": 9.650799485388717e-05, | |
| "loss": 1.5977, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 8.313725490196079, | |
| "grad_norm": 8.783585548400879, | |
| "learning_rate": 9.64712369049807e-05, | |
| "loss": 1.5205, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 8.392156862745098, | |
| "grad_norm": 7.984306812286377, | |
| "learning_rate": 9.643447895607426e-05, | |
| "loss": 1.503, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 6.121412754058838, | |
| "learning_rate": 9.639772100716781e-05, | |
| "loss": 1.3881, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 8.549019607843137, | |
| "grad_norm": 11.138391494750977, | |
| "learning_rate": 9.636096305826136e-05, | |
| "loss": 1.4414, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 8.627450980392156, | |
| "grad_norm": 7.716047763824463, | |
| "learning_rate": 9.632420510935491e-05, | |
| "loss": 1.6433, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 7.658680438995361, | |
| "learning_rate": 9.628744716044845e-05, | |
| "loss": 1.6958, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 8.784313725490197, | |
| "grad_norm": 7.3405961990356445, | |
| "learning_rate": 9.6250689211542e-05, | |
| "loss": 1.7082, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 8.862745098039216, | |
| "grad_norm": 7.0780792236328125, | |
| "learning_rate": 9.621393126263555e-05, | |
| "loss": 1.5016, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 6.372091293334961, | |
| "learning_rate": 9.61771733137291e-05, | |
| "loss": 1.7571, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 9.019607843137255, | |
| "grad_norm": 6.285316467285156, | |
| "learning_rate": 9.614041536482264e-05, | |
| "loss": 1.493, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 9.098039215686274, | |
| "grad_norm": 5.926991939544678, | |
| "learning_rate": 9.610365741591619e-05, | |
| "loss": 1.3031, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 6.716485977172852, | |
| "learning_rate": 9.606689946700974e-05, | |
| "loss": 1.5205, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 9.254901960784313, | |
| "grad_norm": 9.255678176879883, | |
| "learning_rate": 9.60301415181033e-05, | |
| "loss": 1.3186, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 6.3390116691589355, | |
| "learning_rate": 9.599338356919685e-05, | |
| "loss": 1.2825, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 6.447065830230713, | |
| "learning_rate": 9.59566256202904e-05, | |
| "loss": 1.3772, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.490196078431373, | |
| "grad_norm": 5.734104633331299, | |
| "learning_rate": 9.591986767138395e-05, | |
| "loss": 1.2352, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 9.568627450980392, | |
| "grad_norm": 6.8585968017578125, | |
| "learning_rate": 9.58831097224775e-05, | |
| "loss": 1.29, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 6.205005645751953, | |
| "learning_rate": 9.584635177357104e-05, | |
| "loss": 1.4425, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 9.72549019607843, | |
| "grad_norm": 7.428943157196045, | |
| "learning_rate": 9.580959382466459e-05, | |
| "loss": 1.2955, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 9.803921568627452, | |
| "grad_norm": 6.727294921875, | |
| "learning_rate": 9.577283587575814e-05, | |
| "loss": 1.3502, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 6.761404991149902, | |
| "learning_rate": 9.573607792685169e-05, | |
| "loss": 1.5468, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 9.96078431372549, | |
| "grad_norm": 6.411635875701904, | |
| "learning_rate": 9.569931997794523e-05, | |
| "loss": 1.3705, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 10.03921568627451, | |
| "grad_norm": 6.24171257019043, | |
| "learning_rate": 9.566256202903878e-05, | |
| "loss": 1.3356, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 11.419039726257324, | |
| "learning_rate": 9.562580408013233e-05, | |
| "loss": 1.0625, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 10.196078431372548, | |
| "grad_norm": 5.620776653289795, | |
| "learning_rate": 9.558904613122588e-05, | |
| "loss": 1.0987, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 10.27450980392157, | |
| "grad_norm": 5.87687349319458, | |
| "learning_rate": 9.555228818231942e-05, | |
| "loss": 1.1131, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 5.40950345993042, | |
| "learning_rate": 9.551553023341297e-05, | |
| "loss": 1.2988, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 10.431372549019608, | |
| "grad_norm": 7.283133029937744, | |
| "learning_rate": 9.547877228450652e-05, | |
| "loss": 1.1024, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 10.509803921568627, | |
| "grad_norm": 6.536468029022217, | |
| "learning_rate": 9.544201433560009e-05, | |
| "loss": 1.1333, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 5.083970069885254, | |
| "learning_rate": 9.540525638669363e-05, | |
| "loss": 1.198, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 10.666666666666666, | |
| "grad_norm": 6.129286289215088, | |
| "learning_rate": 9.536849843778718e-05, | |
| "loss": 1.1499, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 10.745098039215687, | |
| "grad_norm": 6.173052787780762, | |
| "learning_rate": 9.533174048888073e-05, | |
| "loss": 1.0153, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 9.189874649047852, | |
| "learning_rate": 9.529498253997428e-05, | |
| "loss": 1.3891, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 10.901960784313726, | |
| "grad_norm": 4.8431172370910645, | |
| "learning_rate": 9.525822459106782e-05, | |
| "loss": 1.3164, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 10.980392156862745, | |
| "grad_norm": 7.321885108947754, | |
| "learning_rate": 9.522146664216137e-05, | |
| "loss": 1.2641, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 6.731063365936279, | |
| "learning_rate": 9.518470869325492e-05, | |
| "loss": 1.0551, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 11.137254901960784, | |
| "grad_norm": 5.393633842468262, | |
| "learning_rate": 9.514795074434847e-05, | |
| "loss": 0.933, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 11.215686274509803, | |
| "grad_norm": 6.351990699768066, | |
| "learning_rate": 9.511119279544201e-05, | |
| "loss": 1.1135, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 6.7596893310546875, | |
| "learning_rate": 9.507443484653556e-05, | |
| "loss": 0.8813, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 11.372549019607844, | |
| "grad_norm": 8.091069221496582, | |
| "learning_rate": 9.503767689762911e-05, | |
| "loss": 1.0027, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 11.450980392156863, | |
| "grad_norm": 6.078036308288574, | |
| "learning_rate": 9.500091894872266e-05, | |
| "loss": 0.8464, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 5.587483882904053, | |
| "learning_rate": 9.496416099981622e-05, | |
| "loss": 1.0221, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 11.607843137254902, | |
| "grad_norm": 6.806708812713623, | |
| "learning_rate": 9.492740305090977e-05, | |
| "loss": 1.0172, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 11.686274509803921, | |
| "grad_norm": 5.886943340301514, | |
| "learning_rate": 9.489064510200332e-05, | |
| "loss": 1.0526, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 5.225791931152344, | |
| "learning_rate": 9.485388715309687e-05, | |
| "loss": 1.0879, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.843137254901961, | |
| "grad_norm": 5.893291473388672, | |
| "learning_rate": 9.481712920419041e-05, | |
| "loss": 1.0346, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 11.92156862745098, | |
| "grad_norm": 5.409924507141113, | |
| "learning_rate": 9.478037125528396e-05, | |
| "loss": 1.0874, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 5.1957688331604, | |
| "learning_rate": 9.474361330637751e-05, | |
| "loss": 1.1563, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 12.07843137254902, | |
| "grad_norm": 4.920179843902588, | |
| "learning_rate": 9.470685535747106e-05, | |
| "loss": 0.8673, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 12.156862745098039, | |
| "grad_norm": 5.249741554260254, | |
| "learning_rate": 9.467009740856461e-05, | |
| "loss": 0.8907, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 12.235294117647058, | |
| "grad_norm": 5.800076961517334, | |
| "learning_rate": 9.463333945965815e-05, | |
| "loss": 0.8124, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 12.313725490196079, | |
| "grad_norm": 5.083131313323975, | |
| "learning_rate": 9.45965815107517e-05, | |
| "loss": 0.925, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 12.392156862745098, | |
| "grad_norm": 10.037300109863281, | |
| "learning_rate": 9.455982356184525e-05, | |
| "loss": 0.8389, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 12.470588235294118, | |
| "grad_norm": 6.173994541168213, | |
| "learning_rate": 9.45230656129388e-05, | |
| "loss": 0.9503, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 12.549019607843137, | |
| "grad_norm": 4.115769386291504, | |
| "learning_rate": 9.448630766403234e-05, | |
| "loss": 0.8627, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 12.627450980392156, | |
| "grad_norm": 5.032641410827637, | |
| "learning_rate": 9.44495497151259e-05, | |
| "loss": 0.9014, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 12.705882352941176, | |
| "grad_norm": 5.60946798324585, | |
| "learning_rate": 9.441279176621946e-05, | |
| "loss": 0.8115, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 12.784313725490197, | |
| "grad_norm": 5.839189529418945, | |
| "learning_rate": 9.4376033817313e-05, | |
| "loss": 0.9689, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 12.862745098039216, | |
| "grad_norm": 5.109472751617432, | |
| "learning_rate": 9.433927586840655e-05, | |
| "loss": 0.8387, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 5.745982646942139, | |
| "learning_rate": 9.43025179195001e-05, | |
| "loss": 0.853, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 13.019607843137255, | |
| "grad_norm": 4.644872665405273, | |
| "learning_rate": 9.426575997059365e-05, | |
| "loss": 0.7598, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 13.098039215686274, | |
| "grad_norm": 5.00312614440918, | |
| "learning_rate": 9.42290020216872e-05, | |
| "loss": 0.6555, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 13.176470588235293, | |
| "grad_norm": 3.4448249340057373, | |
| "learning_rate": 9.419224407278074e-05, | |
| "loss": 0.7488, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 13.254901960784313, | |
| "grad_norm": 5.198800563812256, | |
| "learning_rate": 9.415548612387429e-05, | |
| "loss": 0.7083, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 6.767204761505127, | |
| "learning_rate": 9.411872817496784e-05, | |
| "loss": 0.7393, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 13.411764705882353, | |
| "grad_norm": 4.484736919403076, | |
| "learning_rate": 9.40819702260614e-05, | |
| "loss": 0.7486, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 13.490196078431373, | |
| "grad_norm": 4.29071569442749, | |
| "learning_rate": 9.404521227715493e-05, | |
| "loss": 0.64, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 13.568627450980392, | |
| "grad_norm": 5.528765678405762, | |
| "learning_rate": 9.400845432824848e-05, | |
| "loss": 0.8711, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 13.647058823529411, | |
| "grad_norm": 6.199097156524658, | |
| "learning_rate": 9.397169637934203e-05, | |
| "loss": 0.792, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 13.72549019607843, | |
| "grad_norm": 6.095465183258057, | |
| "learning_rate": 9.393493843043559e-05, | |
| "loss": 0.6913, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 13.803921568627452, | |
| "grad_norm": 5.053860664367676, | |
| "learning_rate": 9.389818048152912e-05, | |
| "loss": 0.8404, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 13.882352941176471, | |
| "grad_norm": 5.084766864776611, | |
| "learning_rate": 9.386142253262269e-05, | |
| "loss": 0.8854, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 13.96078431372549, | |
| "grad_norm": 7.908563613891602, | |
| "learning_rate": 9.382466458371624e-05, | |
| "loss": 0.7045, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 14.03921568627451, | |
| "grad_norm": 3.9915947914123535, | |
| "learning_rate": 9.378790663480979e-05, | |
| "loss": 0.6747, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 14.117647058823529, | |
| "grad_norm": 4.792238235473633, | |
| "learning_rate": 9.375114868590333e-05, | |
| "loss": 0.5822, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 14.196078431372548, | |
| "grad_norm": 3.964909553527832, | |
| "learning_rate": 9.371439073699688e-05, | |
| "loss": 0.6341, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 14.27450980392157, | |
| "grad_norm": 5.188769817352295, | |
| "learning_rate": 9.367763278809043e-05, | |
| "loss": 0.7031, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 14.352941176470589, | |
| "grad_norm": 8.642464637756348, | |
| "learning_rate": 9.364087483918398e-05, | |
| "loss": 0.6679, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 14.431372549019608, | |
| "grad_norm": 4.989500999450684, | |
| "learning_rate": 9.360411689027752e-05, | |
| "loss": 0.7224, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 14.509803921568627, | |
| "grad_norm": 5.256617069244385, | |
| "learning_rate": 9.356735894137107e-05, | |
| "loss": 0.6798, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 14.588235294117647, | |
| "grad_norm": 6.925418376922607, | |
| "learning_rate": 9.353060099246462e-05, | |
| "loss": 0.6554, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 14.666666666666666, | |
| "grad_norm": 5.8858513832092285, | |
| "learning_rate": 9.349384304355817e-05, | |
| "loss": 0.6254, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 14.745098039215687, | |
| "grad_norm": 5.78135871887207, | |
| "learning_rate": 9.345708509465173e-05, | |
| "loss": 0.727, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 14.823529411764707, | |
| "grad_norm": 6.644104480743408, | |
| "learning_rate": 9.342032714574526e-05, | |
| "loss": 0.6857, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 14.901960784313726, | |
| "grad_norm": 4.504312992095947, | |
| "learning_rate": 9.338356919683882e-05, | |
| "loss": 0.643, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 14.980392156862745, | |
| "grad_norm": 4.649731636047363, | |
| "learning_rate": 9.334681124793238e-05, | |
| "loss": 0.6495, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 15.058823529411764, | |
| "grad_norm": 4.2058610916137695, | |
| "learning_rate": 9.331005329902592e-05, | |
| "loss": 0.5624, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 15.137254901960784, | |
| "grad_norm": 6.13231897354126, | |
| "learning_rate": 9.327329535011947e-05, | |
| "loss": 0.5113, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 15.215686274509803, | |
| "grad_norm": 5.431331634521484, | |
| "learning_rate": 9.323653740121302e-05, | |
| "loss": 0.5455, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 15.294117647058824, | |
| "grad_norm": 7.001118183135986, | |
| "learning_rate": 9.319977945230657e-05, | |
| "loss": 0.5578, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 15.372549019607844, | |
| "grad_norm": 5.866312503814697, | |
| "learning_rate": 9.316302150340011e-05, | |
| "loss": 0.5321, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 15.450980392156863, | |
| "grad_norm": 5.004396438598633, | |
| "learning_rate": 9.312626355449366e-05, | |
| "loss": 0.6243, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 15.529411764705882, | |
| "grad_norm": 8.219724655151367, | |
| "learning_rate": 9.308950560558721e-05, | |
| "loss": 0.5888, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 15.607843137254902, | |
| "grad_norm": 5.223458766937256, | |
| "learning_rate": 9.305274765668076e-05, | |
| "loss": 0.5239, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 15.686274509803921, | |
| "grad_norm": 3.830970048904419, | |
| "learning_rate": 9.301598970777432e-05, | |
| "loss": 0.6341, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 15.764705882352942, | |
| "grad_norm": 3.6862759590148926, | |
| "learning_rate": 9.297923175886785e-05, | |
| "loss": 0.5209, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 15.843137254901961, | |
| "grad_norm": 4.332780361175537, | |
| "learning_rate": 9.29424738099614e-05, | |
| "loss": 0.6024, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 15.92156862745098, | |
| "grad_norm": 3.963103771209717, | |
| "learning_rate": 9.290571586105496e-05, | |
| "loss": 0.7055, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 4.53104305267334, | |
| "learning_rate": 9.286895791214851e-05, | |
| "loss": 0.5927, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 16.07843137254902, | |
| "grad_norm": 4.432463645935059, | |
| "learning_rate": 9.283219996324205e-05, | |
| "loss": 0.3956, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 16.15686274509804, | |
| "grad_norm": 2.843749523162842, | |
| "learning_rate": 9.279544201433561e-05, | |
| "loss": 0.4537, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 16.235294117647058, | |
| "grad_norm": 3.6052606105804443, | |
| "learning_rate": 9.275868406542916e-05, | |
| "loss": 0.5233, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 16.313725490196077, | |
| "grad_norm": 8.1451416015625, | |
| "learning_rate": 9.272192611652271e-05, | |
| "loss": 0.4672, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 16.392156862745097, | |
| "grad_norm": 4.1013407707214355, | |
| "learning_rate": 9.268516816761625e-05, | |
| "loss": 0.4495, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 16.470588235294116, | |
| "grad_norm": 5.130369186401367, | |
| "learning_rate": 9.26484102187098e-05, | |
| "loss": 0.6151, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 16.54901960784314, | |
| "grad_norm": 3.6761317253112793, | |
| "learning_rate": 9.261165226980335e-05, | |
| "loss": 0.4775, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 16.627450980392158, | |
| "grad_norm": 3.270599126815796, | |
| "learning_rate": 9.25748943208969e-05, | |
| "loss": 0.4543, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 16.705882352941178, | |
| "grad_norm": 6.377615451812744, | |
| "learning_rate": 9.253813637199044e-05, | |
| "loss": 0.5052, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 16.784313725490197, | |
| "grad_norm": 4.0827741622924805, | |
| "learning_rate": 9.2501378423084e-05, | |
| "loss": 0.4645, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 16.862745098039216, | |
| "grad_norm": 4.460141181945801, | |
| "learning_rate": 9.246462047417755e-05, | |
| "loss": 0.5117, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 16.941176470588236, | |
| "grad_norm": 4.786052227020264, | |
| "learning_rate": 9.24278625252711e-05, | |
| "loss": 0.561, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 17.019607843137255, | |
| "grad_norm": 4.18758487701416, | |
| "learning_rate": 9.239110457636463e-05, | |
| "loss": 0.4759, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 17.098039215686274, | |
| "grad_norm": 4.064152240753174, | |
| "learning_rate": 9.235434662745819e-05, | |
| "loss": 0.4141, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 17.176470588235293, | |
| "grad_norm": 3.276078939437866, | |
| "learning_rate": 9.231758867855174e-05, | |
| "loss": 0.3797, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 17.254901960784313, | |
| "grad_norm": 3.8203907012939453, | |
| "learning_rate": 9.22808307296453e-05, | |
| "loss": 0.4376, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 17.333333333333332, | |
| "grad_norm": 3.6585357189178467, | |
| "learning_rate": 9.224407278073884e-05, | |
| "loss": 0.5028, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 17.41176470588235, | |
| "grad_norm": 3.880546808242798, | |
| "learning_rate": 9.220731483183239e-05, | |
| "loss": 0.3938, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 17.49019607843137, | |
| "grad_norm": 5.758749008178711, | |
| "learning_rate": 9.217055688292594e-05, | |
| "loss": 0.3834, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 17.568627450980394, | |
| "grad_norm": 3.563232183456421, | |
| "learning_rate": 9.21337989340195e-05, | |
| "loss": 0.4363, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 17.647058823529413, | |
| "grad_norm": 4.751742839813232, | |
| "learning_rate": 9.209704098511303e-05, | |
| "loss": 0.5009, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 17.725490196078432, | |
| "grad_norm": 3.618528127670288, | |
| "learning_rate": 9.206028303620658e-05, | |
| "loss": 0.4387, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 17.80392156862745, | |
| "grad_norm": 3.945882558822632, | |
| "learning_rate": 9.202352508730013e-05, | |
| "loss": 0.5031, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 17.88235294117647, | |
| "grad_norm": 4.49643087387085, | |
| "learning_rate": 9.198676713839369e-05, | |
| "loss": 0.4466, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 17.96078431372549, | |
| "grad_norm": 4.1370673179626465, | |
| "learning_rate": 9.195000918948722e-05, | |
| "loss": 0.4352, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 18.03921568627451, | |
| "grad_norm": 3.39939546585083, | |
| "learning_rate": 9.191325124058077e-05, | |
| "loss": 0.4518, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 18.11764705882353, | |
| "grad_norm": 3.816342353820801, | |
| "learning_rate": 9.187649329167433e-05, | |
| "loss": 0.3852, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 18.19607843137255, | |
| "grad_norm": 2.6915409564971924, | |
| "learning_rate": 9.183973534276788e-05, | |
| "loss": 0.3584, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 18.274509803921568, | |
| "grad_norm": 4.04006814956665, | |
| "learning_rate": 9.180297739386143e-05, | |
| "loss": 0.3567, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 18.352941176470587, | |
| "grad_norm": 3.8536486625671387, | |
| "learning_rate": 9.176621944495497e-05, | |
| "loss": 0.3251, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 18.431372549019606, | |
| "grad_norm": 3.982511281967163, | |
| "learning_rate": 9.172946149604853e-05, | |
| "loss": 0.4031, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 18.509803921568626, | |
| "grad_norm": 2.8618922233581543, | |
| "learning_rate": 9.169270354714208e-05, | |
| "loss": 0.3933, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 18.58823529411765, | |
| "grad_norm": 6.304449558258057, | |
| "learning_rate": 9.165594559823562e-05, | |
| "loss": 0.4121, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 18.666666666666668, | |
| "grad_norm": 5.2535905838012695, | |
| "learning_rate": 9.161918764932917e-05, | |
| "loss": 0.3762, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 18.745098039215687, | |
| "grad_norm": 4.310611724853516, | |
| "learning_rate": 9.158242970042272e-05, | |
| "loss": 0.3436, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 18.823529411764707, | |
| "grad_norm": 3.7275941371917725, | |
| "learning_rate": 9.154567175151627e-05, | |
| "loss": 0.3255, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 18.901960784313726, | |
| "grad_norm": 3.8429691791534424, | |
| "learning_rate": 9.150891380260981e-05, | |
| "loss": 0.4637, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 18.980392156862745, | |
| "grad_norm": 5.648855686187744, | |
| "learning_rate": 9.147215585370336e-05, | |
| "loss": 0.5983, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 19.058823529411764, | |
| "grad_norm": 4.382913589477539, | |
| "learning_rate": 9.143539790479692e-05, | |
| "loss": 0.34, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 19.137254901960784, | |
| "grad_norm": 3.4657950401306152, | |
| "learning_rate": 9.139863995589047e-05, | |
| "loss": 0.2972, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 19.215686274509803, | |
| "grad_norm": 2.9859068393707275, | |
| "learning_rate": 9.136188200698402e-05, | |
| "loss": 0.3526, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 19.294117647058822, | |
| "grad_norm": 3.8208978176116943, | |
| "learning_rate": 9.132512405807756e-05, | |
| "loss": 0.3774, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 19.372549019607842, | |
| "grad_norm": 3.611250877380371, | |
| "learning_rate": 9.128836610917111e-05, | |
| "loss": 0.3775, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 19.45098039215686, | |
| "grad_norm": 2.989877700805664, | |
| "learning_rate": 9.125160816026466e-05, | |
| "loss": 0.3333, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 19.529411764705884, | |
| "grad_norm": 3.469022750854492, | |
| "learning_rate": 9.121485021135821e-05, | |
| "loss": 0.3057, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 19.607843137254903, | |
| "grad_norm": 2.706902027130127, | |
| "learning_rate": 9.117809226245176e-05, | |
| "loss": 0.431, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 19.686274509803923, | |
| "grad_norm": 3.100156307220459, | |
| "learning_rate": 9.114133431354531e-05, | |
| "loss": 0.3619, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 19.764705882352942, | |
| "grad_norm": 4.185247898101807, | |
| "learning_rate": 9.110457636463886e-05, | |
| "loss": 0.3541, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 19.84313725490196, | |
| "grad_norm": 4.356285572052002, | |
| "learning_rate": 9.106781841573242e-05, | |
| "loss": 0.3376, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 19.92156862745098, | |
| "grad_norm": 3.447700262069702, | |
| "learning_rate": 9.103106046682595e-05, | |
| "loss": 0.3219, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 4.159237384796143, | |
| "learning_rate": 9.09943025179195e-05, | |
| "loss": 0.4266, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 20.07843137254902, | |
| "grad_norm": 3.4393558502197266, | |
| "learning_rate": 9.095754456901306e-05, | |
| "loss": 0.3077, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 20.15686274509804, | |
| "grad_norm": 3.7608890533447266, | |
| "learning_rate": 9.092078662010661e-05, | |
| "loss": 0.3565, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 20.235294117647058, | |
| "grad_norm": 2.8301854133605957, | |
| "learning_rate": 9.088402867120015e-05, | |
| "loss": 0.313, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 20.313725490196077, | |
| "grad_norm": 3.782179594039917, | |
| "learning_rate": 9.08472707222937e-05, | |
| "loss": 0.3104, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 20.392156862745097, | |
| "grad_norm": 2.997694253921509, | |
| "learning_rate": 9.081051277338725e-05, | |
| "loss": 0.2572, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 20.470588235294116, | |
| "grad_norm": 4.569226264953613, | |
| "learning_rate": 9.07737548244808e-05, | |
| "loss": 0.2745, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 20.54901960784314, | |
| "grad_norm": 3.0622193813323975, | |
| "learning_rate": 9.073699687557434e-05, | |
| "loss": 0.3277, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 20.627450980392158, | |
| "grad_norm": 7.896496295928955, | |
| "learning_rate": 9.070023892666789e-05, | |
| "loss": 0.3422, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 20.705882352941178, | |
| "grad_norm": 3.433051109313965, | |
| "learning_rate": 9.066348097776145e-05, | |
| "loss": 0.3259, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 20.784313725490197, | |
| "grad_norm": 4.0141215324401855, | |
| "learning_rate": 9.0626723028855e-05, | |
| "loss": 0.3942, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 20.862745098039216, | |
| "grad_norm": 3.386195421218872, | |
| "learning_rate": 9.058996507994854e-05, | |
| "loss": 0.2881, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 20.941176470588236, | |
| "grad_norm": 2.885312080383301, | |
| "learning_rate": 9.05532071310421e-05, | |
| "loss": 0.2974, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 21.019607843137255, | |
| "grad_norm": 7.563695907592773, | |
| "learning_rate": 9.051644918213565e-05, | |
| "loss": 0.2997, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 21.098039215686274, | |
| "grad_norm": 2.875091075897217, | |
| "learning_rate": 9.04796912332292e-05, | |
| "loss": 0.2349, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 21.176470588235293, | |
| "grad_norm": 8.387333869934082, | |
| "learning_rate": 9.044293328432273e-05, | |
| "loss": 0.2667, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 21.254901960784313, | |
| "grad_norm": 3.013108968734741, | |
| "learning_rate": 9.040617533541629e-05, | |
| "loss": 0.2615, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 21.333333333333332, | |
| "grad_norm": 3.5930674076080322, | |
| "learning_rate": 9.036941738650984e-05, | |
| "loss": 0.3212, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 21.41176470588235, | |
| "grad_norm": 3.1566312313079834, | |
| "learning_rate": 9.033265943760339e-05, | |
| "loss": 0.2714, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 21.49019607843137, | |
| "grad_norm": 4.430455207824707, | |
| "learning_rate": 9.029590148869693e-05, | |
| "loss": 0.3139, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 21.568627450980394, | |
| "grad_norm": 2.473768949508667, | |
| "learning_rate": 9.025914353979048e-05, | |
| "loss": 0.2804, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 21.647058823529413, | |
| "grad_norm": 3.700646162033081, | |
| "learning_rate": 9.022238559088403e-05, | |
| "loss": 0.322, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 21.725490196078432, | |
| "grad_norm": 2.997344970703125, | |
| "learning_rate": 9.018562764197758e-05, | |
| "loss": 0.2736, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 21.80392156862745, | |
| "grad_norm": 3.385653495788574, | |
| "learning_rate": 9.014886969307113e-05, | |
| "loss": 0.2946, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 21.88235294117647, | |
| "grad_norm": 3.961817741394043, | |
| "learning_rate": 9.011211174416468e-05, | |
| "loss": 0.295, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 21.96078431372549, | |
| "grad_norm": 6.599482536315918, | |
| "learning_rate": 9.007535379525823e-05, | |
| "loss": 0.2975, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 22.03921568627451, | |
| "grad_norm": 2.4624006748199463, | |
| "learning_rate": 9.003859584635179e-05, | |
| "loss": 0.2299, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 22.11764705882353, | |
| "grad_norm": 2.7834863662719727, | |
| "learning_rate": 9.000183789744532e-05, | |
| "loss": 0.2237, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 22.19607843137255, | |
| "grad_norm": 2.49113392829895, | |
| "learning_rate": 8.996507994853887e-05, | |
| "loss": 0.2631, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 22.274509803921568, | |
| "grad_norm": 4.437926292419434, | |
| "learning_rate": 8.992832199963243e-05, | |
| "loss": 0.227, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 22.352941176470587, | |
| "grad_norm": 6.664700508117676, | |
| "learning_rate": 8.989156405072598e-05, | |
| "loss": 0.3085, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 22.431372549019606, | |
| "grad_norm": 3.945110321044922, | |
| "learning_rate": 8.985480610181952e-05, | |
| "loss": 0.2354, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 22.509803921568626, | |
| "grad_norm": 2.9697070121765137, | |
| "learning_rate": 8.981804815291307e-05, | |
| "loss": 0.2374, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 22.58823529411765, | |
| "grad_norm": 2.9439990520477295, | |
| "learning_rate": 8.978129020400662e-05, | |
| "loss": 0.2507, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 22.666666666666668, | |
| "grad_norm": 2.947354555130005, | |
| "learning_rate": 8.974453225510017e-05, | |
| "loss": 0.2647, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 22.745098039215687, | |
| "grad_norm": 9.103282928466797, | |
| "learning_rate": 8.970777430619372e-05, | |
| "loss": 0.3106, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 22.823529411764707, | |
| "grad_norm": 2.5283734798431396, | |
| "learning_rate": 8.967101635728726e-05, | |
| "loss": 0.2715, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 22.901960784313726, | |
| "grad_norm": 3.052879810333252, | |
| "learning_rate": 8.963425840838081e-05, | |
| "loss": 0.2977, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 22.980392156862745, | |
| "grad_norm": 3.37917423248291, | |
| "learning_rate": 8.959750045947437e-05, | |
| "loss": 0.2878, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 23.058823529411764, | |
| "grad_norm": 3.028381109237671, | |
| "learning_rate": 8.956074251056791e-05, | |
| "loss": 0.2169, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 23.137254901960784, | |
| "grad_norm": 2.4643447399139404, | |
| "learning_rate": 8.952398456166146e-05, | |
| "loss": 0.2073, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 23.215686274509803, | |
| "grad_norm": 2.397473096847534, | |
| "learning_rate": 8.948722661275502e-05, | |
| "loss": 0.2108, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 23.294117647058822, | |
| "grad_norm": 6.173182010650635, | |
| "learning_rate": 8.945046866384857e-05, | |
| "loss": 0.2428, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 23.372549019607842, | |
| "grad_norm": 3.343395948410034, | |
| "learning_rate": 8.941371071494212e-05, | |
| "loss": 0.275, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 23.45098039215686, | |
| "grad_norm": 2.379011631011963, | |
| "learning_rate": 8.937695276603566e-05, | |
| "loss": 0.2336, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 23.529411764705884, | |
| "grad_norm": 7.347818374633789, | |
| "learning_rate": 8.934019481712921e-05, | |
| "loss": 0.2575, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 23.607843137254903, | |
| "grad_norm": 7.274477005004883, | |
| "learning_rate": 8.930343686822276e-05, | |
| "loss": 0.2616, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 23.686274509803923, | |
| "grad_norm": 4.475617408752441, | |
| "learning_rate": 8.926667891931631e-05, | |
| "loss": 0.2336, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 23.764705882352942, | |
| "grad_norm": 3.119966506958008, | |
| "learning_rate": 8.922992097040985e-05, | |
| "loss": 0.2705, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 23.84313725490196, | |
| "grad_norm": 2.663884401321411, | |
| "learning_rate": 8.91931630215034e-05, | |
| "loss": 0.2863, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 23.92156862745098, | |
| "grad_norm": 6.014930725097656, | |
| "learning_rate": 8.915640507259695e-05, | |
| "loss": 0.2437, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 2.6029369831085205, | |
| "learning_rate": 8.91196471236905e-05, | |
| "loss": 0.2844, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 24.07843137254902, | |
| "grad_norm": 2.7642641067504883, | |
| "learning_rate": 8.908288917478405e-05, | |
| "loss": 0.2014, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 24.15686274509804, | |
| "grad_norm": 8.297749519348145, | |
| "learning_rate": 8.90461312258776e-05, | |
| "loss": 0.2657, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 24.235294117647058, | |
| "grad_norm": 3.9905178546905518, | |
| "learning_rate": 8.900937327697116e-05, | |
| "loss": 0.2114, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 24.313725490196077, | |
| "grad_norm": 2.5204246044158936, | |
| "learning_rate": 8.897261532806471e-05, | |
| "loss": 0.1942, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 24.392156862745097, | |
| "grad_norm": 1.52804434299469, | |
| "learning_rate": 8.893585737915824e-05, | |
| "loss": 0.2155, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 24.470588235294116, | |
| "grad_norm": 3.6253719329833984, | |
| "learning_rate": 8.88990994302518e-05, | |
| "loss": 0.2201, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 24.54901960784314, | |
| "grad_norm": 7.122885227203369, | |
| "learning_rate": 8.886234148134535e-05, | |
| "loss": 0.2297, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 24.627450980392158, | |
| "grad_norm": 2.7436540126800537, | |
| "learning_rate": 8.88255835324389e-05, | |
| "loss": 0.2201, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 24.705882352941178, | |
| "grad_norm": 5.490830898284912, | |
| "learning_rate": 8.878882558353244e-05, | |
| "loss": 0.2075, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 24.784313725490197, | |
| "grad_norm": 2.3640992641448975, | |
| "learning_rate": 8.875206763462599e-05, | |
| "loss": 0.2283, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 24.862745098039216, | |
| "grad_norm": 3.1421284675598145, | |
| "learning_rate": 8.871530968571954e-05, | |
| "loss": 0.2722, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 24.941176470588236, | |
| "grad_norm": 2.805938720703125, | |
| "learning_rate": 8.867855173681309e-05, | |
| "loss": 0.2361, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 25.019607843137255, | |
| "grad_norm": 7.49440336227417, | |
| "learning_rate": 8.864179378790663e-05, | |
| "loss": 0.2847, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 25.098039215686274, | |
| "grad_norm": 2.238097667694092, | |
| "learning_rate": 8.860503583900018e-05, | |
| "loss": 0.2008, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 25.176470588235293, | |
| "grad_norm": 2.652937650680542, | |
| "learning_rate": 8.856827789009373e-05, | |
| "loss": 0.1826, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 25.254901960784313, | |
| "grad_norm": 2.6613991260528564, | |
| "learning_rate": 8.85315199411873e-05, | |
| "loss": 0.1676, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 25.333333333333332, | |
| "grad_norm": 2.0592703819274902, | |
| "learning_rate": 8.849476199228083e-05, | |
| "loss": 0.2177, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 25.41176470588235, | |
| "grad_norm": 3.4880011081695557, | |
| "learning_rate": 8.845800404337439e-05, | |
| "loss": 0.1935, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 25.49019607843137, | |
| "grad_norm": 2.736335039138794, | |
| "learning_rate": 8.842124609446794e-05, | |
| "loss": 0.1994, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 25.568627450980394, | |
| "grad_norm": 5.7319135665893555, | |
| "learning_rate": 8.838448814556149e-05, | |
| "loss": 0.2072, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 25.647058823529413, | |
| "grad_norm": 10.623271942138672, | |
| "learning_rate": 8.834773019665503e-05, | |
| "loss": 0.2004, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 25.725490196078432, | |
| "grad_norm": 2.4742937088012695, | |
| "learning_rate": 8.831097224774858e-05, | |
| "loss": 0.2465, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 25.80392156862745, | |
| "grad_norm": 2.440775156021118, | |
| "learning_rate": 8.827421429884213e-05, | |
| "loss": 0.2073, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 25.88235294117647, | |
| "grad_norm": 4.591070175170898, | |
| "learning_rate": 8.823745634993568e-05, | |
| "loss": 0.225, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 25.96078431372549, | |
| "grad_norm": 2.302111864089966, | |
| "learning_rate": 8.820069840102923e-05, | |
| "loss": 0.2245, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 26.03921568627451, | |
| "grad_norm": 2.730738401412964, | |
| "learning_rate": 8.816394045212277e-05, | |
| "loss": 0.2307, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 26.11764705882353, | |
| "grad_norm": 1.9027403593063354, | |
| "learning_rate": 8.812718250321632e-05, | |
| "loss": 0.1931, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 26.19607843137255, | |
| "grad_norm": 2.853452444076538, | |
| "learning_rate": 8.809042455430987e-05, | |
| "loss": 0.1913, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 26.274509803921568, | |
| "grad_norm": 2.136833667755127, | |
| "learning_rate": 8.805366660540342e-05, | |
| "loss": 0.1896, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 26.352941176470587, | |
| "grad_norm": 3.3222334384918213, | |
| "learning_rate": 8.801690865649697e-05, | |
| "loss": 0.1924, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 26.431372549019606, | |
| "grad_norm": 3.190403938293457, | |
| "learning_rate": 8.798015070759053e-05, | |
| "loss": 0.1569, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 26.509803921568626, | |
| "grad_norm": 3.4979772567749023, | |
| "learning_rate": 8.794339275868408e-05, | |
| "loss": 0.1813, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 26.58823529411765, | |
| "grad_norm": 3.0356762409210205, | |
| "learning_rate": 8.790663480977762e-05, | |
| "loss": 0.2141, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 2.5389366149902344, | |
| "learning_rate": 8.786987686087117e-05, | |
| "loss": 0.1986, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 26.745098039215687, | |
| "grad_norm": 2.2066240310668945, | |
| "learning_rate": 8.783311891196472e-05, | |
| "loss": 0.2011, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 26.823529411764707, | |
| "grad_norm": 2.6409451961517334, | |
| "learning_rate": 8.779636096305827e-05, | |
| "loss": 0.233, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 26.901960784313726, | |
| "grad_norm": 2.3644559383392334, | |
| "learning_rate": 8.775960301415182e-05, | |
| "loss": 0.1877, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 26.980392156862745, | |
| "grad_norm": 3.346972942352295, | |
| "learning_rate": 8.772284506524536e-05, | |
| "loss": 0.2201, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 27.058823529411764, | |
| "grad_norm": 3.081000328063965, | |
| "learning_rate": 8.768608711633891e-05, | |
| "loss": 0.1929, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 27.137254901960784, | |
| "grad_norm": 1.6890923976898193, | |
| "learning_rate": 8.764932916743246e-05, | |
| "loss": 0.1539, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 27.215686274509803, | |
| "grad_norm": 4.7821221351623535, | |
| "learning_rate": 8.761257121852601e-05, | |
| "loss": 0.1717, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 27.294117647058822, | |
| "grad_norm": 2.38714861869812, | |
| "learning_rate": 8.757581326961955e-05, | |
| "loss": 0.2534, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 27.372549019607842, | |
| "grad_norm": 2.4988088607788086, | |
| "learning_rate": 8.75390553207131e-05, | |
| "loss": 0.2302, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 27.45098039215686, | |
| "grad_norm": 2.1674258708953857, | |
| "learning_rate": 8.750229737180665e-05, | |
| "loss": 0.1788, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 27.529411764705884, | |
| "grad_norm": 3.270306348800659, | |
| "learning_rate": 8.746553942290022e-05, | |
| "loss": 0.1815, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 27.607843137254903, | |
| "grad_norm": 2.9274301528930664, | |
| "learning_rate": 8.742878147399376e-05, | |
| "loss": 0.1871, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 27.686274509803923, | |
| "grad_norm": 2.2478270530700684, | |
| "learning_rate": 8.739202352508731e-05, | |
| "loss": 0.1861, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 27.764705882352942, | |
| "grad_norm": 3.159546136856079, | |
| "learning_rate": 8.735526557618086e-05, | |
| "loss": 0.2163, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 27.84313725490196, | |
| "grad_norm": 4.743581771850586, | |
| "learning_rate": 8.731850762727441e-05, | |
| "loss": 0.1892, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 27.92156862745098, | |
| "grad_norm": 4.11615514755249, | |
| "learning_rate": 8.728174967836795e-05, | |
| "loss": 0.2062, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 3.3777382373809814, | |
| "learning_rate": 8.72449917294615e-05, | |
| "loss": 0.2048, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 28.07843137254902, | |
| "grad_norm": 12.32438850402832, | |
| "learning_rate": 8.720823378055505e-05, | |
| "loss": 0.1704, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 28.15686274509804, | |
| "grad_norm": 2.3112239837646484, | |
| "learning_rate": 8.71714758316486e-05, | |
| "loss": 0.1787, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 28.235294117647058, | |
| "grad_norm": 2.7134642601013184, | |
| "learning_rate": 8.713471788274214e-05, | |
| "loss": 0.1595, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 28.313725490196077, | |
| "grad_norm": 1.900732398033142, | |
| "learning_rate": 8.709795993383569e-05, | |
| "loss": 0.1614, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 28.392156862745097, | |
| "grad_norm": 2.7066845893859863, | |
| "learning_rate": 8.706120198492924e-05, | |
| "loss": 0.1571, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 28.470588235294116, | |
| "grad_norm": 2.3240630626678467, | |
| "learning_rate": 8.70244440360228e-05, | |
| "loss": 0.1988, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 28.54901960784314, | |
| "grad_norm": 5.567899227142334, | |
| "learning_rate": 8.698768608711633e-05, | |
| "loss": 0.1768, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 28.627450980392158, | |
| "grad_norm": 2.9699206352233887, | |
| "learning_rate": 8.69509281382099e-05, | |
| "loss": 0.1649, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 28.705882352941178, | |
| "grad_norm": 2.124846935272217, | |
| "learning_rate": 8.691417018930345e-05, | |
| "loss": 0.2035, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 28.784313725490197, | |
| "grad_norm": 2.9401068687438965, | |
| "learning_rate": 8.6877412240397e-05, | |
| "loss": 0.1728, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 28.862745098039216, | |
| "grad_norm": 2.0019986629486084, | |
| "learning_rate": 8.684065429149054e-05, | |
| "loss": 0.176, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 28.941176470588236, | |
| "grad_norm": 6.4335222244262695, | |
| "learning_rate": 8.680389634258409e-05, | |
| "loss": 0.1593, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 29.019607843137255, | |
| "grad_norm": 1.7808016538619995, | |
| "learning_rate": 8.676713839367764e-05, | |
| "loss": 0.1949, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 29.098039215686274, | |
| "grad_norm": 1.9336371421813965, | |
| "learning_rate": 8.673038044477119e-05, | |
| "loss": 0.1508, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 29.176470588235293, | |
| "grad_norm": 1.271824598312378, | |
| "learning_rate": 8.669362249586473e-05, | |
| "loss": 0.1448, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 29.254901960784313, | |
| "grad_norm": 2.432981252670288, | |
| "learning_rate": 8.665686454695828e-05, | |
| "loss": 0.1454, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 29.333333333333332, | |
| "grad_norm": 1.8998444080352783, | |
| "learning_rate": 8.662010659805183e-05, | |
| "loss": 0.1893, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 29.41176470588235, | |
| "grad_norm": 2.9416303634643555, | |
| "learning_rate": 8.658334864914538e-05, | |
| "loss": 0.1999, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 29.49019607843137, | |
| "grad_norm": 2.7823660373687744, | |
| "learning_rate": 8.654659070023893e-05, | |
| "loss": 0.1543, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 29.568627450980394, | |
| "grad_norm": 1.9040496349334717, | |
| "learning_rate": 8.650983275133247e-05, | |
| "loss": 0.1588, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 29.647058823529413, | |
| "grad_norm": 2.2860541343688965, | |
| "learning_rate": 8.647307480242602e-05, | |
| "loss": 0.1376, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 29.725490196078432, | |
| "grad_norm": 2.458211898803711, | |
| "learning_rate": 8.643631685351957e-05, | |
| "loss": 0.1502, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 29.80392156862745, | |
| "grad_norm": 4.291934967041016, | |
| "learning_rate": 8.639955890461313e-05, | |
| "loss": 0.1664, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 29.88235294117647, | |
| "grad_norm": 8.039182662963867, | |
| "learning_rate": 8.636280095570668e-05, | |
| "loss": 0.1673, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 29.96078431372549, | |
| "grad_norm": 1.9125957489013672, | |
| "learning_rate": 8.632604300680023e-05, | |
| "loss": 0.2028, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 30.03921568627451, | |
| "grad_norm": 1.9361660480499268, | |
| "learning_rate": 8.628928505789378e-05, | |
| "loss": 0.1725, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 30.11764705882353, | |
| "grad_norm": 2.263054132461548, | |
| "learning_rate": 8.625252710898732e-05, | |
| "loss": 0.1409, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 30.19607843137255, | |
| "grad_norm": 1.9042737483978271, | |
| "learning_rate": 8.621576916008087e-05, | |
| "loss": 0.1336, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 30.274509803921568, | |
| "grad_norm": 1.9479308128356934, | |
| "learning_rate": 8.617901121117442e-05, | |
| "loss": 0.1503, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 30.352941176470587, | |
| "grad_norm": 2.608462333679199, | |
| "learning_rate": 8.614225326226797e-05, | |
| "loss": 0.1654, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 30.431372549019606, | |
| "grad_norm": 2.01275372505188, | |
| "learning_rate": 8.610549531336152e-05, | |
| "loss": 0.1427, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 30.509803921568626, | |
| "grad_norm": 8.765314102172852, | |
| "learning_rate": 8.606873736445506e-05, | |
| "loss": 0.1878, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 30.58823529411765, | |
| "grad_norm": 2.0812177658081055, | |
| "learning_rate": 8.603197941554861e-05, | |
| "loss": 0.1411, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 30.666666666666668, | |
| "grad_norm": 2.890509843826294, | |
| "learning_rate": 8.599522146664216e-05, | |
| "loss": 0.1884, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 30.745098039215687, | |
| "grad_norm": 2.644294023513794, | |
| "learning_rate": 8.595846351773572e-05, | |
| "loss": 0.1868, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 30.823529411764707, | |
| "grad_norm": 1.564231276512146, | |
| "learning_rate": 8.592170556882925e-05, | |
| "loss": 0.1733, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 30.901960784313726, | |
| "grad_norm": 2.647084951400757, | |
| "learning_rate": 8.588494761992282e-05, | |
| "loss": 0.2895, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 30.980392156862745, | |
| "grad_norm": 1.3914289474487305, | |
| "learning_rate": 8.584818967101637e-05, | |
| "loss": 0.1602, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 31.058823529411764, | |
| "grad_norm": 26.367515563964844, | |
| "learning_rate": 8.581143172210992e-05, | |
| "loss": 0.4469, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 31.137254901960784, | |
| "grad_norm": 1.853583574295044, | |
| "learning_rate": 8.577467377320346e-05, | |
| "loss": 0.1277, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 31.215686274509803, | |
| "grad_norm": 2.2600510120391846, | |
| "learning_rate": 8.573791582429701e-05, | |
| "loss": 0.1729, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 31.294117647058822, | |
| "grad_norm": 1.9825807809829712, | |
| "learning_rate": 8.570115787539056e-05, | |
| "loss": 0.1597, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 31.372549019607842, | |
| "grad_norm": 1.8277662992477417, | |
| "learning_rate": 8.566439992648411e-05, | |
| "loss": 0.1419, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 27305, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 215, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 522593501184000.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |