| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.6470588235294117, | |
| "eval_steps": 500, | |
| "global_step": 90000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014705882352941176, | |
| "grad_norm": 0.35731062293052673, | |
| "learning_rate": 4.975490196078432e-05, | |
| "loss": 0.5354, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.029411764705882353, | |
| "grad_norm": 2.1595866680145264, | |
| "learning_rate": 4.9509803921568634e-05, | |
| "loss": 0.1627, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.04411764705882353, | |
| "grad_norm": 0.14825384318828583, | |
| "learning_rate": 4.9264705882352944e-05, | |
| "loss": 0.101, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.058823529411764705, | |
| "grad_norm": 0.07121703773736954, | |
| "learning_rate": 4.901960784313725e-05, | |
| "loss": 0.0952, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.07352941176470588, | |
| "grad_norm": 0.06089532747864723, | |
| "learning_rate": 4.877450980392157e-05, | |
| "loss": 0.0881, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.08823529411764706, | |
| "grad_norm": 0.037034619599580765, | |
| "learning_rate": 4.8529411764705885e-05, | |
| "loss": 0.0717, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.10294117647058823, | |
| "grad_norm": 0.01999847963452339, | |
| "learning_rate": 4.82843137254902e-05, | |
| "loss": 0.0848, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.018650399520993233, | |
| "learning_rate": 4.803921568627452e-05, | |
| "loss": 0.068, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.1323529411764706, | |
| "grad_norm": 0.023023229092359543, | |
| "learning_rate": 4.7794117647058826e-05, | |
| "loss": 0.0626, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.14705882352941177, | |
| "grad_norm": 0.018715515732765198, | |
| "learning_rate": 4.7549019607843135e-05, | |
| "loss": 0.0555, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.16176470588235295, | |
| "grad_norm": 0.11842140555381775, | |
| "learning_rate": 4.730392156862745e-05, | |
| "loss": 0.0658, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.17647058823529413, | |
| "grad_norm": 0.04816881939768791, | |
| "learning_rate": 4.705882352941177e-05, | |
| "loss": 0.0648, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.19117647058823528, | |
| "grad_norm": 0.025240018963813782, | |
| "learning_rate": 4.681372549019608e-05, | |
| "loss": 0.0632, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.20588235294117646, | |
| "grad_norm": 0.007390766404569149, | |
| "learning_rate": 4.656862745098039e-05, | |
| "loss": 0.0705, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.22058823529411764, | |
| "grad_norm": 0.011664963327348232, | |
| "learning_rate": 4.632352941176471e-05, | |
| "loss": 0.0552, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 39.69175720214844, | |
| "learning_rate": 4.607843137254902e-05, | |
| "loss": 0.0686, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.007366931065917015, | |
| "learning_rate": 4.5833333333333334e-05, | |
| "loss": 0.061, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.2647058823529412, | |
| "grad_norm": 0.003396671498194337, | |
| "learning_rate": 4.558823529411765e-05, | |
| "loss": 0.0464, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.27941176470588236, | |
| "grad_norm": 0.045366521924734116, | |
| "learning_rate": 4.5343137254901966e-05, | |
| "loss": 0.053, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 1.137495756149292, | |
| "learning_rate": 4.5098039215686275e-05, | |
| "loss": 0.0484, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.3088235294117647, | |
| "grad_norm": 0.0062417215667665005, | |
| "learning_rate": 4.485294117647059e-05, | |
| "loss": 0.0564, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.3235294117647059, | |
| "grad_norm": 0.003990447614341974, | |
| "learning_rate": 4.460784313725491e-05, | |
| "loss": 0.0364, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.3382352941176471, | |
| "grad_norm": 0.11569799482822418, | |
| "learning_rate": 4.4362745098039216e-05, | |
| "loss": 0.0436, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 0.01159907691180706, | |
| "learning_rate": 4.411764705882353e-05, | |
| "loss": 0.0456, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.36764705882352944, | |
| "grad_norm": 10.926911354064941, | |
| "learning_rate": 4.387254901960784e-05, | |
| "loss": 0.0464, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.38235294117647056, | |
| "grad_norm": 0.03444543853402138, | |
| "learning_rate": 4.362745098039216e-05, | |
| "loss": 0.0512, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.39705882352941174, | |
| "grad_norm": 0.002395658055320382, | |
| "learning_rate": 4.3382352941176474e-05, | |
| "loss": 0.0586, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.4117647058823529, | |
| "grad_norm": 0.010224021971225739, | |
| "learning_rate": 4.313725490196079e-05, | |
| "loss": 0.0407, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.4264705882352941, | |
| "grad_norm": 0.010642035864293575, | |
| "learning_rate": 4.28921568627451e-05, | |
| "loss": 0.0536, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.4411764705882353, | |
| "grad_norm": 0.00970557238906622, | |
| "learning_rate": 4.2647058823529415e-05, | |
| "loss": 0.0526, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.45588235294117646, | |
| "grad_norm": 0.0030253385193645954, | |
| "learning_rate": 4.2401960784313724e-05, | |
| "loss": 0.0432, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.5423064827919006, | |
| "learning_rate": 4.215686274509804e-05, | |
| "loss": 0.0381, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.4852941176470588, | |
| "grad_norm": 7.454125881195068, | |
| "learning_rate": 4.1911764705882356e-05, | |
| "loss": 0.0397, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 3.87744402885437, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.0564, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.5147058823529411, | |
| "grad_norm": 33.39067840576172, | |
| "learning_rate": 4.142156862745099e-05, | |
| "loss": 0.0455, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.5294117647058824, | |
| "grad_norm": 0.008246080949902534, | |
| "learning_rate": 4.11764705882353e-05, | |
| "loss": 0.0515, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.5441176470588235, | |
| "grad_norm": 0.018734918907284737, | |
| "learning_rate": 4.0931372549019607e-05, | |
| "loss": 0.0453, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.5588235294117647, | |
| "grad_norm": 0.007373027969151735, | |
| "learning_rate": 4.068627450980392e-05, | |
| "loss": 0.0485, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.5735294117647058, | |
| "grad_norm": 0.0023267469368875027, | |
| "learning_rate": 4.044117647058824e-05, | |
| "loss": 0.0372, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.1311252862215042, | |
| "learning_rate": 4.0196078431372555e-05, | |
| "loss": 0.0442, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.6029411764705882, | |
| "grad_norm": 0.003710985416546464, | |
| "learning_rate": 3.9950980392156864e-05, | |
| "loss": 0.0585, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.6176470588235294, | |
| "grad_norm": 0.004861747846007347, | |
| "learning_rate": 3.970588235294117e-05, | |
| "loss": 0.0374, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.6323529411764706, | |
| "grad_norm": 0.004351571202278137, | |
| "learning_rate": 3.946078431372549e-05, | |
| "loss": 0.0265, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.6470588235294118, | |
| "grad_norm": 0.008851751685142517, | |
| "learning_rate": 3.9215686274509805e-05, | |
| "loss": 0.0314, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.6617647058823529, | |
| "grad_norm": 0.0046307104639709, | |
| "learning_rate": 3.897058823529412e-05, | |
| "loss": 0.0445, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.6764705882352942, | |
| "grad_norm": 0.002735880669206381, | |
| "learning_rate": 3.872549019607844e-05, | |
| "loss": 0.0433, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.6911764705882353, | |
| "grad_norm": 0.0218490082770586, | |
| "learning_rate": 3.8480392156862746e-05, | |
| "loss": 0.0444, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 2.371448278427124, | |
| "learning_rate": 3.8235294117647055e-05, | |
| "loss": 0.033, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.7205882352941176, | |
| "grad_norm": 0.005015780217945576, | |
| "learning_rate": 3.799019607843137e-05, | |
| "loss": 0.0422, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 1.3102850914001465, | |
| "learning_rate": 3.774509803921569e-05, | |
| "loss": 0.035, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.004522955510765314, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0303, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.7647058823529411, | |
| "grad_norm": 0.006077844649553299, | |
| "learning_rate": 3.725490196078432e-05, | |
| "loss": 0.0503, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.7794117647058824, | |
| "grad_norm": 0.0026446939446032047, | |
| "learning_rate": 3.700980392156863e-05, | |
| "loss": 0.0264, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.7941176470588235, | |
| "grad_norm": 0.004485867917537689, | |
| "learning_rate": 3.6764705882352945e-05, | |
| "loss": 0.0388, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.8088235294117647, | |
| "grad_norm": 0.0073866695165634155, | |
| "learning_rate": 3.6519607843137254e-05, | |
| "loss": 0.0283, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 0.012984287925064564, | |
| "learning_rate": 3.627450980392157e-05, | |
| "loss": 0.0393, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.8382352941176471, | |
| "grad_norm": 0.01751883700489998, | |
| "learning_rate": 3.6029411764705886e-05, | |
| "loss": 0.0388, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.8529411764705882, | |
| "grad_norm": 0.0035843336954712868, | |
| "learning_rate": 3.5784313725490195e-05, | |
| "loss": 0.0384, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.8676470588235294, | |
| "grad_norm": 0.0645672082901001, | |
| "learning_rate": 3.553921568627451e-05, | |
| "loss": 0.0378, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 0.008566264994442463, | |
| "learning_rate": 3.529411764705883e-05, | |
| "loss": 0.0315, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.8970588235294118, | |
| "grad_norm": 0.010571740567684174, | |
| "learning_rate": 3.5049019607843136e-05, | |
| "loss": 0.0324, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.9117647058823529, | |
| "grad_norm": 0.0022533361334353685, | |
| "learning_rate": 3.480392156862745e-05, | |
| "loss": 0.0292, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.9264705882352942, | |
| "grad_norm": 0.006164130289107561, | |
| "learning_rate": 3.455882352941177e-05, | |
| "loss": 0.0407, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.007435985840857029, | |
| "learning_rate": 3.431372549019608e-05, | |
| "loss": 0.0449, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.9558823529411765, | |
| "grad_norm": 0.003777585458010435, | |
| "learning_rate": 3.4068627450980394e-05, | |
| "loss": 0.0247, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.9705882352941176, | |
| "grad_norm": 0.005975374951958656, | |
| "learning_rate": 3.382352941176471e-05, | |
| "loss": 0.0309, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.9852941176470589, | |
| "grad_norm": 0.0016012012492865324, | |
| "learning_rate": 3.357843137254902e-05, | |
| "loss": 0.0294, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.020027656108140945, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.022, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.0147058823529411, | |
| "grad_norm": 10.916418075561523, | |
| "learning_rate": 3.308823529411765e-05, | |
| "loss": 0.0264, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.0294117647058822, | |
| "grad_norm": 0.013728056102991104, | |
| "learning_rate": 3.284313725490196e-05, | |
| "loss": 0.032, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.0441176470588236, | |
| "grad_norm": 0.02546643279492855, | |
| "learning_rate": 3.2598039215686276e-05, | |
| "loss": 0.0404, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 1.0588235294117647, | |
| "grad_norm": 0.004185052588582039, | |
| "learning_rate": 3.235294117647059e-05, | |
| "loss": 0.0306, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.0735294117647058, | |
| "grad_norm": 0.016653403639793396, | |
| "learning_rate": 3.210784313725491e-05, | |
| "loss": 0.0217, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 1.088235294117647, | |
| "grad_norm": 0.0028331661596894264, | |
| "learning_rate": 3.186274509803922e-05, | |
| "loss": 0.0265, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.1029411764705883, | |
| "grad_norm": 0.013931985944509506, | |
| "learning_rate": 3.161764705882353e-05, | |
| "loss": 0.0399, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 1.1176470588235294, | |
| "grad_norm": 0.001668413169682026, | |
| "learning_rate": 3.137254901960784e-05, | |
| "loss": 0.0202, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.1323529411764706, | |
| "grad_norm": 0.00567347789183259, | |
| "learning_rate": 3.112745098039216e-05, | |
| "loss": 0.0336, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 1.1470588235294117, | |
| "grad_norm": 0.003477458842098713, | |
| "learning_rate": 3.0882352941176475e-05, | |
| "loss": 0.0253, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.161764705882353, | |
| "grad_norm": 0.21819466352462769, | |
| "learning_rate": 3.063725490196079e-05, | |
| "loss": 0.0293, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.004708552733063698, | |
| "learning_rate": 3.0392156862745097e-05, | |
| "loss": 0.0268, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.1911764705882353, | |
| "grad_norm": 0.0018363581039011478, | |
| "learning_rate": 3.0147058823529413e-05, | |
| "loss": 0.0266, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 1.2058823529411764, | |
| "grad_norm": 0.005581580102443695, | |
| "learning_rate": 2.9901960784313725e-05, | |
| "loss": 0.0361, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.2205882352941178, | |
| "grad_norm": 0.0035322746261954308, | |
| "learning_rate": 2.965686274509804e-05, | |
| "loss": 0.0274, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 1.2352941176470589, | |
| "grad_norm": 0.0038708180654793978, | |
| "learning_rate": 2.9411764705882354e-05, | |
| "loss": 0.0241, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.0019606975838541985, | |
| "learning_rate": 2.916666666666667e-05, | |
| "loss": 0.0219, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 1.2647058823529411, | |
| "grad_norm": 0.00241417670622468, | |
| "learning_rate": 2.8921568627450986e-05, | |
| "loss": 0.0244, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.2794117647058822, | |
| "grad_norm": 0.003593308152630925, | |
| "learning_rate": 2.8676470588235295e-05, | |
| "loss": 0.027, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 1.2941176470588236, | |
| "grad_norm": 1.4003372192382812, | |
| "learning_rate": 2.8431372549019608e-05, | |
| "loss": 0.0392, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.3088235294117647, | |
| "grad_norm": 0.41921645402908325, | |
| "learning_rate": 2.8186274509803924e-05, | |
| "loss": 0.0299, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 1.3235294117647058, | |
| "grad_norm": 1.567896842956543, | |
| "learning_rate": 2.7941176470588236e-05, | |
| "loss": 0.0282, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.3382352941176472, | |
| "grad_norm": 0.003143745008856058, | |
| "learning_rate": 2.7696078431372552e-05, | |
| "loss": 0.0197, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 1.3529411764705883, | |
| "grad_norm": 0.0016428233357146382, | |
| "learning_rate": 2.7450980392156865e-05, | |
| "loss": 0.0344, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.3676470588235294, | |
| "grad_norm": 0.005276167765259743, | |
| "learning_rate": 2.7205882352941174e-05, | |
| "loss": 0.0312, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 1.3823529411764706, | |
| "grad_norm": 0.004461635369807482, | |
| "learning_rate": 2.696078431372549e-05, | |
| "loss": 0.025, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 1.3970588235294117, | |
| "grad_norm": 0.003238542238250375, | |
| "learning_rate": 2.6715686274509806e-05, | |
| "loss": 0.0236, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.0017602238804101944, | |
| "learning_rate": 2.647058823529412e-05, | |
| "loss": 0.0234, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.4264705882352942, | |
| "grad_norm": 0.002768098609521985, | |
| "learning_rate": 2.6225490196078435e-05, | |
| "loss": 0.026, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 1.4411764705882353, | |
| "grad_norm": 0.002579926745966077, | |
| "learning_rate": 2.5980392156862747e-05, | |
| "loss": 0.0167, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 1.4558823529411764, | |
| "grad_norm": 0.008888750337064266, | |
| "learning_rate": 2.5735294117647057e-05, | |
| "loss": 0.0288, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.0038515792693942785, | |
| "learning_rate": 2.5490196078431373e-05, | |
| "loss": 0.0252, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 1.4852941176470589, | |
| "grad_norm": 4.428643226623535, | |
| "learning_rate": 2.5245098039215685e-05, | |
| "loss": 0.016, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.002629584399983287, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0214, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.5147058823529411, | |
| "grad_norm": 0.004097859375178814, | |
| "learning_rate": 2.4754901960784317e-05, | |
| "loss": 0.0183, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 1.5294117647058822, | |
| "grad_norm": 0.00568019924685359, | |
| "learning_rate": 2.4509803921568626e-05, | |
| "loss": 0.0271, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 1.5441176470588234, | |
| "grad_norm": 0.0020534582436084747, | |
| "learning_rate": 2.4264705882352942e-05, | |
| "loss": 0.0281, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 1.5588235294117647, | |
| "grad_norm": 0.003300599753856659, | |
| "learning_rate": 2.401960784313726e-05, | |
| "loss": 0.0284, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 1.5735294117647058, | |
| "grad_norm": 0.001472037984058261, | |
| "learning_rate": 2.3774509803921568e-05, | |
| "loss": 0.0205, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 1.5882352941176472, | |
| "grad_norm": 0.002835978288203478, | |
| "learning_rate": 2.3529411764705884e-05, | |
| "loss": 0.0234, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.6029411764705883, | |
| "grad_norm": 0.003979724366217852, | |
| "learning_rate": 2.3284313725490196e-05, | |
| "loss": 0.0233, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 1.6176470588235294, | |
| "grad_norm": 0.0036057273391634226, | |
| "learning_rate": 2.303921568627451e-05, | |
| "loss": 0.0319, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 1.6323529411764706, | |
| "grad_norm": 0.0024822901468724012, | |
| "learning_rate": 2.2794117647058825e-05, | |
| "loss": 0.0211, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.008930359967052937, | |
| "learning_rate": 2.2549019607843138e-05, | |
| "loss": 0.0309, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 1.6617647058823528, | |
| "grad_norm": 1.3381928205490112, | |
| "learning_rate": 2.2303921568627454e-05, | |
| "loss": 0.019, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 1.6764705882352942, | |
| "grad_norm": 0.004481327719986439, | |
| "learning_rate": 2.2058823529411766e-05, | |
| "loss": 0.0228, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 1.6911764705882353, | |
| "grad_norm": 0.012597435154020786, | |
| "learning_rate": 2.181372549019608e-05, | |
| "loss": 0.0285, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 1.7058823529411766, | |
| "grad_norm": 0.0034783107694238424, | |
| "learning_rate": 2.1568627450980395e-05, | |
| "loss": 0.0165, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 1.7205882352941178, | |
| "grad_norm": 0.0032083301339298487, | |
| "learning_rate": 2.1323529411764707e-05, | |
| "loss": 0.0306, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 1.7352941176470589, | |
| "grad_norm": 0.0039000194519758224, | |
| "learning_rate": 2.107843137254902e-05, | |
| "loss": 0.0199, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.0034919639583677053, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.034, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 0.004928025882691145, | |
| "learning_rate": 2.058823529411765e-05, | |
| "loss": 0.0213, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 1.7794117647058822, | |
| "grad_norm": 0.003211489412933588, | |
| "learning_rate": 2.034313725490196e-05, | |
| "loss": 0.0206, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 1.7941176470588234, | |
| "grad_norm": 0.011239697225391865, | |
| "learning_rate": 2.0098039215686277e-05, | |
| "loss": 0.0201, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 1.8088235294117647, | |
| "grad_norm": 0.0024609589017927647, | |
| "learning_rate": 1.9852941176470586e-05, | |
| "loss": 0.0286, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 1.8235294117647058, | |
| "grad_norm": 0.0046806493774056435, | |
| "learning_rate": 1.9607843137254903e-05, | |
| "loss": 0.0216, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 1.8382352941176472, | |
| "grad_norm": 2.30717396736145, | |
| "learning_rate": 1.936274509803922e-05, | |
| "loss": 0.0246, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 1.8529411764705883, | |
| "grad_norm": 0.00428669573739171, | |
| "learning_rate": 1.9117647058823528e-05, | |
| "loss": 0.0166, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 1.8676470588235294, | |
| "grad_norm": 0.011403021402657032, | |
| "learning_rate": 1.8872549019607844e-05, | |
| "loss": 0.0272, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.0065813250839710236, | |
| "learning_rate": 1.862745098039216e-05, | |
| "loss": 0.0217, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 1.8970588235294117, | |
| "grad_norm": 0.0024323465768247843, | |
| "learning_rate": 1.8382352941176472e-05, | |
| "loss": 0.0176, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 1.9117647058823528, | |
| "grad_norm": 0.001616469700820744, | |
| "learning_rate": 1.8137254901960785e-05, | |
| "loss": 0.0169, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 1.9264705882352942, | |
| "grad_norm": 0.004322522785514593, | |
| "learning_rate": 1.7892156862745098e-05, | |
| "loss": 0.0216, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 1.9411764705882353, | |
| "grad_norm": 0.0024695000611245632, | |
| "learning_rate": 1.7647058823529414e-05, | |
| "loss": 0.032, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 1.9558823529411766, | |
| "grad_norm": 0.010675052180886269, | |
| "learning_rate": 1.7401960784313726e-05, | |
| "loss": 0.023, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 1.9705882352941178, | |
| "grad_norm": 0.49467232823371887, | |
| "learning_rate": 1.715686274509804e-05, | |
| "loss": 0.0169, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 1.9852941176470589, | |
| "grad_norm": 0.01075649168342352, | |
| "learning_rate": 1.6911764705882355e-05, | |
| "loss": 0.0177, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.026703685522079468, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0259, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 2.014705882352941, | |
| "grad_norm": 0.0023890878073871136, | |
| "learning_rate": 1.642156862745098e-05, | |
| "loss": 0.0128, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 2.0294117647058822, | |
| "grad_norm": 0.0015898487763479352, | |
| "learning_rate": 1.6176470588235296e-05, | |
| "loss": 0.0112, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.0441176470588234, | |
| "grad_norm": 0.0041648312471807, | |
| "learning_rate": 1.593137254901961e-05, | |
| "loss": 0.0155, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 2.0588235294117645, | |
| "grad_norm": 0.10951696336269379, | |
| "learning_rate": 1.568627450980392e-05, | |
| "loss": 0.0225, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 2.073529411764706, | |
| "grad_norm": 0.0021414640359580517, | |
| "learning_rate": 1.5441176470588237e-05, | |
| "loss": 0.0174, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 2.088235294117647, | |
| "grad_norm": 0.001627126126550138, | |
| "learning_rate": 1.5196078431372548e-05, | |
| "loss": 0.0154, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 2.1029411764705883, | |
| "grad_norm": 0.005821748171001673, | |
| "learning_rate": 1.4950980392156863e-05, | |
| "loss": 0.0255, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.0071876379661262035, | |
| "learning_rate": 1.4705882352941177e-05, | |
| "loss": 0.024, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.1323529411764706, | |
| "grad_norm": 0.00294076488353312, | |
| "learning_rate": 1.4460784313725493e-05, | |
| "loss": 0.012, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 2.1470588235294117, | |
| "grad_norm": 0.004260234069079161, | |
| "learning_rate": 1.4215686274509804e-05, | |
| "loss": 0.0202, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 2.161764705882353, | |
| "grad_norm": 0.005722737871110439, | |
| "learning_rate": 1.3970588235294118e-05, | |
| "loss": 0.0283, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 2.176470588235294, | |
| "grad_norm": 0.016162721440196037, | |
| "learning_rate": 1.3725490196078432e-05, | |
| "loss": 0.0115, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 2.1911764705882355, | |
| "grad_norm": 0.005048350431025028, | |
| "learning_rate": 1.3480392156862745e-05, | |
| "loss": 0.023, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 2.2058823529411766, | |
| "grad_norm": 0.004102902952581644, | |
| "learning_rate": 1.323529411764706e-05, | |
| "loss": 0.018, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.2205882352941178, | |
| "grad_norm": 0.0021536860149353743, | |
| "learning_rate": 1.2990196078431374e-05, | |
| "loss": 0.0169, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 2.235294117647059, | |
| "grad_norm": 0.001363063813187182, | |
| "learning_rate": 1.2745098039215686e-05, | |
| "loss": 0.0129, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.0025944672524929047, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0095, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 2.264705882352941, | |
| "grad_norm": 0.001098868204280734, | |
| "learning_rate": 1.2254901960784313e-05, | |
| "loss": 0.0183, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 2.2794117647058822, | |
| "grad_norm": 0.00421817135065794, | |
| "learning_rate": 1.200980392156863e-05, | |
| "loss": 0.0214, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 2.2941176470588234, | |
| "grad_norm": 0.0016445108922198415, | |
| "learning_rate": 1.1764705882352942e-05, | |
| "loss": 0.009, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 2.3088235294117645, | |
| "grad_norm": 0.002265740418806672, | |
| "learning_rate": 1.1519607843137254e-05, | |
| "loss": 0.0161, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 2.323529411764706, | |
| "grad_norm": 0.007099386304616928, | |
| "learning_rate": 1.1274509803921569e-05, | |
| "loss": 0.0143, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 2.338235294117647, | |
| "grad_norm": 0.003914414439350367, | |
| "learning_rate": 1.1029411764705883e-05, | |
| "loss": 0.0115, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.002513893647119403, | |
| "learning_rate": 1.0784313725490197e-05, | |
| "loss": 0.0136, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 2.3676470588235294, | |
| "grad_norm": 0.0027356306090950966, | |
| "learning_rate": 1.053921568627451e-05, | |
| "loss": 0.0247, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 2.3823529411764706, | |
| "grad_norm": 0.0817839726805687, | |
| "learning_rate": 1.0294117647058824e-05, | |
| "loss": 0.0169, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 2.3970588235294117, | |
| "grad_norm": 0.08163878321647644, | |
| "learning_rate": 1.0049019607843139e-05, | |
| "loss": 0.0114, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 2.411764705882353, | |
| "grad_norm": 0.002625884721055627, | |
| "learning_rate": 9.803921568627451e-06, | |
| "loss": 0.0147, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 2.426470588235294, | |
| "grad_norm": 0.004531237296760082, | |
| "learning_rate": 9.558823529411764e-06, | |
| "loss": 0.0106, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 2.4411764705882355, | |
| "grad_norm": 0.0032483581453561783, | |
| "learning_rate": 9.31372549019608e-06, | |
| "loss": 0.0183, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 2.4558823529411766, | |
| "grad_norm": 0.0050782193429768085, | |
| "learning_rate": 9.068627450980392e-06, | |
| "loss": 0.0168, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 2.4705882352941178, | |
| "grad_norm": 0.003839900717139244, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.0209, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 2.485294117647059, | |
| "grad_norm": 0.008910595439374447, | |
| "learning_rate": 8.57843137254902e-06, | |
| "loss": 0.0271, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.008313077501952648, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.0175, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 2.514705882352941, | |
| "grad_norm": 0.0029984668362885714, | |
| "learning_rate": 8.088235294117648e-06, | |
| "loss": 0.0084, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 2.5294117647058822, | |
| "grad_norm": 0.004524989053606987, | |
| "learning_rate": 7.84313725490196e-06, | |
| "loss": 0.0156, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 2.5441176470588234, | |
| "grad_norm": 0.0023315059952437878, | |
| "learning_rate": 7.598039215686274e-06, | |
| "loss": 0.0121, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 2.5588235294117645, | |
| "grad_norm": 0.0010077670449391007, | |
| "learning_rate": 7.3529411764705884e-06, | |
| "loss": 0.0113, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 2.5735294117647056, | |
| "grad_norm": 0.0012684785760939121, | |
| "learning_rate": 7.107843137254902e-06, | |
| "loss": 0.0204, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.004918810911476612, | |
| "learning_rate": 6.862745098039216e-06, | |
| "loss": 0.0059, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 2.6029411764705883, | |
| "grad_norm": 0.011489451862871647, | |
| "learning_rate": 6.61764705882353e-06, | |
| "loss": 0.0273, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 2.6176470588235294, | |
| "grad_norm": 0.0028730102349072695, | |
| "learning_rate": 6.372549019607843e-06, | |
| "loss": 0.0277, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 2.6323529411764706, | |
| "grad_norm": 0.005050596781075001, | |
| "learning_rate": 6.127450980392157e-06, | |
| "loss": 0.0148, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 2.6470588235294117, | |
| "grad_norm": 0.0028206182178109884, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.0121, | |
| "step": 90000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 102000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 5.579973324914688e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |