diff --git "a/weights/best_model_metadata.json" "b/weights/best_model_metadata.json" --- "a/weights/best_model_metadata.json" +++ "b/weights/best_model_metadata.json" @@ -1,241 +1,256 @@ { - "epoch": 4, + "epoch": 5, "optimizer_state_dict": { "state": { "0": { - "step": "tensor(6260.)", - "exp_avg": "tensor([[ 7.7419e-05, -7.2244e-06, 3.0531e-05, ..., 2.5845e-05,\n 1.8728e-05, 3.9241e-06],\n [ 3.3904e-06, -1.4326e-05, 4.1395e-06, ..., 4.7042e-06,\n 1.7799e-05, -3.5841e-06],\n [-2.3233e-06, 2.1983e-06, 1.2519e-05, ..., -6.5935e-07,\n -8.1431e-07, -1.6264e-06],\n ...,\n [-3.6582e-06, 4.0445e-05, -2.2387e-05, ..., 4.7334e-05,\n -1.9004e-06, 2.3796e-05],\n [-5.6052e-45, -5.6052e-45, 5.6052e-45, ..., -5.6052e-45,\n 5.6052e-45, -5.6052e-45],\n [-3.9960e-05, -3.0237e-05, 2.3135e-05, ..., -2.7614e-05,\n 1.2621e-05, -4.3781e-06]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.2214e-08, 9.8888e-09, 5.6333e-09, ..., 6.1682e-09, 6.1113e-09,\n 6.5305e-09],\n [5.4207e-09, 5.4562e-09, 3.2772e-09, ..., 3.2730e-09, 3.7076e-09,\n 2.4386e-09],\n [1.1351e-08, 1.0384e-08, 1.3721e-08, ..., 8.7800e-09, 7.4439e-09,\n 6.3112e-09],\n ...,\n [1.1335e-08, 1.3139e-08, 1.9069e-08, ..., 8.3830e-09, 7.8853e-09,\n 8.1240e-09],\n [3.1389e-13, 1.6375e-12, 4.9763e-13, ..., 2.9173e-15, 1.5702e-12,\n 9.2530e-15],\n [1.0876e-08, 1.0017e-08, 1.0210e-08, ..., 8.8808e-09, 5.4898e-09,\n 4.7628e-09]], device='cuda:0')" + "step": "tensor(7512.)", + "exp_avg": "tensor([[-1.0491e-05, -1.8756e-05, 3.1406e-06, ..., -1.4494e-05,\n 6.2594e-07, 8.9240e-07],\n [-7.2855e-06, 1.5954e-06, 2.8693e-05, ..., 8.1217e-07,\n 4.8767e-06, -2.8365e-05],\n [ 1.1289e-05, -2.5604e-05, 7.9899e-06, ..., 7.2300e-06,\n 4.4164e-07, -2.1137e-05],\n ...,\n [ 6.0636e-06, -3.2096e-05, 7.3410e-06, ..., 7.2256e-06,\n -2.7950e-05, -1.6518e-05],\n [-5.6052e-45, -5.6052e-45, 5.6052e-45, ..., -5.6052e-45,\n 5.6052e-45, -5.6052e-45],\n [-1.4321e-05, -1.1473e-05, -6.2940e-07, ..., -1.8249e-06,\n -1.0515e-05, 1.8671e-05]], device='cuda:0')", + "exp_avg_sq": "tensor([[1.2631e-08, 1.0092e-08, 5.7972e-09, ..., 6.7559e-09, 6.8399e-09,\n 6.4820e-09],\n [5.1994e-09, 5.0326e-09, 2.8316e-09, ..., 2.9525e-09, 3.2975e-09,\n 2.1800e-09],\n [1.1561e-08, 9.9322e-09, 1.3502e-08, ..., 8.0087e-09, 6.8592e-09,\n 6.9060e-09],\n ...,\n [1.1178e-08, 1.2353e-08, 1.9119e-08, ..., 8.6691e-09, 7.8107e-09,\n 8.3348e-09],\n [8.9697e-14, 4.6793e-13, 1.4220e-13, ..., 8.3365e-16, 4.4870e-13,\n 2.6441e-15],\n [1.0137e-08, 1.0153e-08, 1.0214e-08, ..., 9.0691e-09, 5.4310e-09,\n 5.1230e-09]], device='cuda:0')" }, "1": { - "step": "tensor(6260.)", - "exp_avg": "tensor([ 1.5406e-03, -1.1476e-03, -2.6409e-04, ..., 2.9145e-04,\n 5.6052e-45, -1.4334e-03], device='cuda:0')", - "exp_avg_sq": "tensor([1.3140e-05, 7.1135e-06, 1.9650e-05, ..., 2.1019e-05, 1.0628e-08,\n 1.4784e-05], device='cuda:0')" + "step": "tensor(7512.)", + "exp_avg": "tensor([ 1.5959e-04, -2.5446e-04, 1.1861e-04, ..., 5.5200e-04,\n 5.6052e-45, -3.7160e-04], device='cuda:0')", + "exp_avg_sq": "tensor([1.3634e-05, 6.0766e-06, 1.9405e-05, ..., 2.0885e-05, 3.0371e-09,\n 1.5491e-05], device='cuda:0')" }, "2": { - "step": "tensor(6260.)", - "exp_avg": "tensor([[-1.1485e-06, -5.7111e-07, 2.8125e-06, ..., -1.6099e-06,\n 5.6052e-45, -6.3541e-06],\n [-5.3498e-06, 5.3070e-06, -2.1827e-06, ..., 7.7663e-06,\n -5.6052e-45, -6.0954e-06],\n [-4.6320e-06, 1.8319e-06, -1.0043e-06, ..., 6.1784e-06,\n -5.6052e-45, 5.3614e-06],\n ...,\n [-2.2621e-08, 2.2287e-09, -1.9373e-07, ..., 4.6302e-07,\n 5.6052e-45, 1.4883e-07],\n [-1.2227e-06, -9.5979e-09, 1.4337e-07, ..., -1.5761e-08,\n -5.6052e-45, 1.0289e-05],\n [-2.3983e-05, 8.2098e-06, -1.4137e-05, ..., -2.6232e-05,\n 5.6052e-45, -6.9302e-06]], device='cuda:0')", - "exp_avg_sq": "tensor([[2.3307e-09, 3.9215e-10, 2.4352e-09, ..., 9.7499e-09, 2.5026e-13,\n 1.1423e-09],\n [3.0932e-09, 1.1406e-09, 8.4448e-10, ..., 6.5127e-10, 1.3886e-11,\n 4.0199e-09],\n [8.5765e-10, 8.2466e-10, 8.1234e-10, ..., 9.9855e-10, 9.5692e-11,\n 2.0900e-09],\n ...,\n [4.1790e-11, 2.6058e-11, 3.9500e-10, ..., 2.8262e-10, 9.6092e-12,\n 2.8260e-10],\n [1.6500e-09, 2.2122e-11, 4.5411e-10, ..., 3.4966e-10, 1.9384e-14,\n 1.0797e-09],\n [4.3279e-09, 9.7757e-10, 1.0829e-08, ..., 2.3307e-09, 1.7174e-12,\n 3.1455e-09]], device='cuda:0')" + "step": "tensor(7512.)", + "exp_avg": "tensor([[ 1.5417e-06, -1.2857e-08, -5.3836e-06, ..., 1.6783e-05,\n 5.6052e-45, 1.6000e-05],\n [ 1.1963e-06, 3.4353e-06, 1.2080e-06, ..., 1.5608e-06,\n -5.6052e-45, -2.3203e-05],\n [ 2.2897e-06, 2.8499e-07, 7.4077e-07, ..., -1.2110e-05,\n -5.6052e-45, 1.5150e-06],\n ...,\n [-6.0717e-08, 4.3581e-07, 1.8418e-06, ..., 6.3973e-07,\n 5.6052e-45, 1.1569e-08],\n [-5.0018e-06, -7.1451e-07, -6.1910e-06, ..., -9.2834e-07,\n -5.6052e-45, -1.2894e-06],\n [-4.3011e-06, -8.1580e-07, 6.3227e-05, ..., 1.6049e-06,\n 5.6052e-45, -1.2083e-05]], device='cuda:0')", + "exp_avg_sq": "tensor([[1.9021e-09, 3.8362e-10, 2.3030e-09, ..., 7.8229e-09, 7.1515e-14,\n 8.8721e-10],\n [2.7390e-09, 1.1005e-09, 6.1551e-10, ..., 3.9103e-10, 3.9680e-12,\n 3.7034e-09],\n [6.9946e-10, 6.4883e-10, 6.2728e-10, ..., 8.6593e-10, 2.7345e-11,\n 1.6464e-09],\n ...,\n [2.7892e-11, 2.0384e-11, 2.5093e-10, ..., 1.0583e-10, 2.7459e-12,\n 1.3197e-10],\n [1.2578e-09, 1.7416e-11, 2.4336e-10, ..., 4.4390e-10, 5.5392e-15,\n 7.9846e-10],\n [3.9775e-09, 6.6087e-10, 8.4403e-09, ..., 2.4633e-09, 4.9076e-13,\n 2.4511e-09]], device='cuda:0')" }, "3": { - "step": "tensor(6260.)", - "exp_avg": "tensor([ 1.0477e-04, 1.1376e-04, 6.8580e-05, 1.1062e-04, -1.5102e-04,\n 1.8946e-05, 2.2321e-04, 4.4966e-04, -2.1635e-04, 3.6279e-04,\n -4.8620e-05, -1.1527e-06, 7.3181e-05, 9.5280e-05, 1.7739e-04,\n 1.6335e-04, -1.1370e-04, -1.1923e-05, 2.4393e-05, 1.0780e-04,\n -9.4480e-05, 2.5872e-04, 4.0348e-04, -1.9958e-04, -8.5265e-05,\n -1.9338e-04, 1.9970e-04, -8.4242e-05, 2.3576e-05, -3.3020e-05,\n -3.9298e-05, -7.7467e-06, -3.9194e-04, -9.2478e-05, 5.7959e-05,\n 5.9237e-05, 4.3912e-04, -1.0686e-04, 1.5071e-04, 8.6443e-06,\n 7.1097e-04, 1.8752e-04, 2.1190e-05, 4.4484e-05, 4.0833e-05,\n 9.2181e-05, 1.0533e-18, 7.5112e-05, 2.1147e-05, 1.4307e-04,\n 4.7949e-05, -1.8811e-05, 3.2247e-05, -2.0407e-06, -4.8476e-05,\n -2.7448e-04, -2.3936e-04, -9.3065e-05, 4.3918e-05, 8.5998e-05,\n 1.7864e-05, -1.4735e-04, -9.0984e-05, -2.1071e-06, 1.3226e-04,\n 5.5162e-05, -2.7692e-05, 1.8201e-04, 9.0112e-05, -4.0201e-04,\n 1.3602e-04, 5.6052e-45, 6.4871e-06, 2.7337e-04, -1.0230e-04,\n -1.5436e-04, -4.0674e-05, 5.9901e-05, -1.7824e-04, 1.3852e-04,\n 4.4658e-05, -1.3218e-04, -1.4014e-04, 1.7143e-05, -6.1319e-05,\n -1.4133e-04, -3.2296e-04, 3.2819e-04, -1.4037e-04, -3.6881e-04,\n 6.7872e-06, 2.7025e-04, -5.5208e-04, -6.8871e-05, 3.1423e-05,\n 1.5790e-04, 1.0097e-05, 1.3257e-04, -2.4127e-04, 3.0916e-05,\n 1.5010e-05, -2.2331e-04, -7.9740e-05, -2.9505e-06, -1.0984e-05,\n -1.8364e-06, -4.5352e-05, -1.1953e-04, 1.7912e-05, -2.8878e-04,\n 3.9155e-05, 1.4545e-04, -4.1241e-06, -7.3574e-05, 4.0110e-04,\n 5.3793e-05, 1.5396e-04, -2.4731e-04, -3.7000e-05, 1.9019e-04,\n -1.4076e-04, 1.2596e-04, -3.3846e-04, 2.6942e-04, -2.5929e-04,\n -2.7995e-04, 7.5528e-06, 2.9249e-05, -3.6848e-04, -1.2839e-04,\n -7.5530e-05, 1.4641e-04, 1.3239e-05, -7.3279e-05, -1.5624e-04,\n 2.0339e-04, 1.8814e-04, 8.7494e-05, -1.8627e-04, 3.8127e-04,\n 9.6811e-05, -2.5912e-05, 8.9033e-05, 2.4748e-05, 1.6964e-04,\n -1.7393e-04, -2.3904e-05, -1.5047e-04, 1.1144e-04, 2.8452e-05,\n 1.1765e-06, 4.0289e-05, 1.1107e-04, -6.5209e-05, 2.4792e-05,\n 5.6931e-13, 7.5780e-05, 1.4192e-04, 7.0784e-05, -4.5672e-04,\n 1.4378e-04, 7.7271e-06, -2.6640e-04, 8.8398e-05, -2.3331e-04,\n 1.5183e-05, 3.1811e-04, -1.2715e-05, 1.1393e-05, -2.7617e-04,\n -2.7581e-05, 8.1515e-05, 1.2959e-07, -4.6871e-04, 4.8529e-04,\n -3.3767e-04, -2.5262e-04, 1.7203e-04, -1.8694e-04, 1.1845e-04,\n 8.9294e-05, 2.0866e-04, 1.2850e-04, 8.5298e-06, 1.4894e-04,\n -9.5003e-05, 1.9936e-04, -1.3296e-04, -4.4874e-05, 2.0801e-04,\n 1.2607e-04, -1.7662e-06, 9.1341e-05, -1.8251e-04, -2.0454e-04,\n 4.9065e-05, -2.9016e-05, 2.0897e-04, 5.6052e-45, 9.3667e-05,\n -2.2721e-06, 2.3989e-04, 2.0541e-04, 4.6781e-06, 1.2744e-04,\n -4.4005e-05, -4.8953e-06, -2.6030e-04, 1.9106e-04, 9.8366e-05,\n 4.9292e-05, -1.3986e-04, 1.2331e-42, -3.2295e-05, -4.8253e-05,\n -1.3859e-04, 5.3263e-04, 6.0121e-05, -2.2526e-04, -8.2499e-05,\n 4.7832e-05, 2.7444e-05, -1.1168e-04, 1.5333e-04, -8.4024e-06,\n 3.1697e-04, -2.3630e-04, -2.1602e-04, -1.5837e-04, -5.0515e-04,\n -1.1856e-04, 4.3696e-05, 5.7240e-05, -1.5466e-04, -2.6130e-04,\n -2.7558e-05, -5.1042e-05, -5.5390e-05, -1.4133e-04, 1.0975e-05,\n -4.0222e-04, 1.2004e-04, -8.3381e-05, -1.3901e-04, -4.6265e-05,\n 1.5251e-04, 2.1482e-05, -3.0831e-04, 4.1425e-05, -9.0239e-05,\n 1.5997e-04, -1.0658e-04, 1.9397e-04, -8.7239e-05, -8.7222e-05,\n -1.9878e-04, 1.4545e-05, -4.9886e-05, 1.6795e-04, -1.6524e-05,\n -1.7034e-04, -4.8796e-04, 8.6639e-05, 1.0742e-04, 1.3301e-04,\n -7.4432e-05, -3.7230e-04, 8.4317e-05, -2.9158e-04, 1.5610e-04,\n -1.6772e-04, 8.4969e-05, -1.0371e-04, -3.1074e-06, 2.4064e-04,\n -1.6916e-04, 2.6740e-05, 1.1810e-05, 1.1866e-05, -1.3244e-05,\n -1.0962e-04, 6.9882e-05, -3.3924e-04, 1.7443e-04, -5.3820e-05,\n 6.3928e-05, 6.3257e-05, 9.3041e-05, 1.6266e-04, -3.9549e-04,\n -1.3156e-05, -2.1692e-04, -1.1161e-04, -6.8772e-05, 2.7330e-10,\n 1.4159e-04, 6.2701e-05, 3.7193e-04, 2.4729e-06, 1.1071e-04,\n 1.1206e-04, -6.4637e-05, 2.0143e-04, 3.8784e-06, -7.4142e-05,\n 6.1758e-05, -4.0520e-04, -8.6071e-05, 1.0703e-04, 2.0677e-05,\n -2.2336e-05, 2.3289e-05, -8.7161e-05, -1.1777e-04, -2.6742e-04,\n 4.0852e-05, -3.2601e-04, -1.7674e-04, -7.0380e-05, 2.6905e-04,\n -2.4768e-04, 5.2272e-05, 1.4922e-04, 1.9704e-04, -1.0745e-04,\n 9.4971e-05, 3.7686e-05, 1.2015e-04, -4.0044e-04, 2.5644e-04,\n 4.8441e-05, -1.9149e-04, -3.9381e-05, 5.9652e-05, 1.1766e-04,\n 3.9545e-05, 1.2883e-04, 1.3951e-04, 8.3888e-05, 1.5828e-05,\n 1.7781e-04, -1.2960e-04, 7.2166e-05, 1.9132e-06, -1.0379e-04,\n -2.5532e-04, 1.9610e-05, 5.7568e-06, -9.5458e-05, 2.9811e-04,\n -4.2334e-05, -1.7680e-04, 4.2791e-05, 2.5948e-04, 7.9276e-05,\n 2.5633e-05, 3.0468e-05, 2.0320e-04, 8.5785e-05, -1.1651e-04,\n -5.4273e-05, -7.3785e-04, 1.9157e-04, -3.3559e-04, -2.2174e-05,\n 6.2231e-05, 5.6052e-45, -1.7726e-05, -1.7570e-04, 2.0834e-04,\n -7.0150e-05, 1.2011e-04, 2.3193e-05, 2.5905e-04, 4.8156e-05,\n 9.4227e-05, -1.5386e-04, -3.2985e-04, -8.4540e-09, 1.3336e-04,\n 3.2558e-11, 7.8638e-05, 1.3845e-04, 2.0236e-04, 2.7482e-05,\n 1.9566e-04, 7.6982e-05, 7.3572e-05, 1.9124e-05, -4.4878e-04,\n -7.3165e-05, 4.0464e-05, 8.3604e-05, 2.8651e-05, -1.3938e-05,\n -7.2535e-05, -2.8282e-05, -3.0575e-04, 9.7689e-05, 1.9464e-04,\n 3.7703e-05, -6.0119e-05, -2.4153e-04, 8.2015e-05, 1.4069e-04,\n 8.2936e-05, 5.2972e-05, -3.1064e-05, 1.0269e-04, 2.1319e-05,\n 2.9697e-05, 5.1293e-05, -1.4735e-04, -1.1027e-04, -1.0017e-05,\n 3.7506e-05, 1.9099e-04, -2.7635e-04, -1.7746e-04, -1.0333e-05,\n -1.1964e-06, -7.3104e-05, -2.7289e-04, 1.2494e-05, -1.1267e-04,\n 9.3180e-05, 8.4389e-05, 1.2577e-04, -3.1900e-05, 6.9829e-05,\n 4.4992e-04, -2.2451e-05, -6.2085e-05, 1.0990e-04, 1.0630e-04,\n 2.3869e-04, -1.3656e-04, -8.4981e-05, -1.8713e-04, 4.7313e-05,\n 3.7676e-04, 1.2203e-04, 1.5593e-04, 7.6599e-05, -4.5713e-04,\n 3.9790e-04, 5.6052e-45, -1.1085e-04, 6.5030e-05, 2.4468e-04,\n -4.8410e-05, -1.5028e-04, 6.5386e-05, 8.2005e-05, 1.5086e-05,\n 2.5292e-04, 7.3439e-05, -6.3366e-05, 2.4668e-04, -1.3422e-04,\n -1.8001e-04, -6.6304e-05, -6.5024e-06, 6.7363e-05, -9.5922e-05,\n -3.2134e-05, 1.9623e-04, -3.5708e-05, -1.3757e-04, -1.3367e-04,\n -1.0616e-05, 4.4708e-05, 5.9554e-06, 4.1430e-05, 5.7943e-05,\n 3.9001e-05, -1.1295e-04, 5.3587e-05, 1.4498e-04, 1.8632e-04,\n -3.6238e-05, 1.3988e-04, 1.4748e-04, -2.5542e-04, -1.4115e-04,\n -7.5265e-05, 1.2344e-04, -1.2543e-04, 1.4823e-04, 3.9968e-04,\n 3.4148e-04, -8.2524e-05, 2.1892e-04, 2.9281e-04, 1.4586e-04,\n 5.6052e-45, -1.4999e-04, -7.8472e-06, -3.9977e-05, 1.2903e-04,\n 8.7892e-06, -5.1465e-05, 1.0537e-04, 5.8208e-05, 1.5826e-04,\n -1.4232e-04, -1.7625e-05, 8.3911e-05, -2.5242e-05, 3.7977e-05,\n -1.4690e-04, -4.8904e-05], device='cuda:0')", - "exp_avg_sq": "tensor([3.4901e-07, 4.9780e-07, 4.5792e-07, 4.8573e-07, 3.5745e-07, 4.6700e-07,\n 2.1051e-07, 5.0464e-07, 3.7209e-07, 2.4541e-07, 7.1930e-07, 3.6780e-07,\n 1.9960e-07, 3.7045e-07, 9.0267e-08, 4.4603e-07, 6.2253e-07, 5.7818e-07,\n 2.6456e-07, 3.0200e-07, 1.0456e-07, 3.3981e-07, 7.9043e-07, 2.8181e-07,\n 4.2366e-07, 4.3592e-07, 5.2335e-07, 5.1445e-07, 1.7712e-07, 1.8183e-07,\n 5.1255e-07, 6.2132e-07, 6.1939e-07, 7.1401e-07, 1.9657e-07, 3.3129e-07,\n 3.2808e-07, 6.1111e-07, 7.0171e-07, 3.2175e-07, 6.3079e-07, 7.1098e-07,\n 5.5448e-07, 2.0753e-07, 5.6254e-07, 5.8143e-07, 4.6728e-07, 2.4432e-07,\n 6.2282e-07, 3.0151e-07, 3.3118e-07, 2.1630e-07, 4.8689e-07, 3.5626e-07,\n 3.2663e-07, 5.7372e-07, 4.4594e-07, 3.6973e-07, 4.8599e-07, 5.2061e-07,\n 3.8244e-07, 5.1999e-07, 4.0964e-07, 5.2100e-07, 3.4740e-07, 1.8908e-07,\n 3.9346e-07, 4.1781e-07, 3.3507e-07, 2.3654e-07, 4.3065e-07, 2.2398e-15,\n 3.7701e-07, 4.1233e-07, 2.1751e-07, 3.8878e-07, 5.4075e-07, 5.4777e-07,\n 5.2886e-07, 4.0562e-07, 2.7422e-07, 4.5003e-07, 2.8292e-07, 4.2921e-07,\n 3.4534e-07, 2.1345e-07, 4.0643e-07, 7.4621e-07, 2.4236e-07, 5.3800e-07,\n 3.2723e-07, 5.9020e-07, 4.1647e-07, 6.1475e-07, 1.6163e-07, 6.0077e-07,\n 6.0399e-07, 6.7833e-07, 8.2945e-07, 3.5063e-07, 5.0820e-07, 4.1673e-07,\n 4.1482e-07, 7.0023e-07, 2.7133e-07, 3.6670e-07, 4.9629e-07, 3.1619e-07,\n 2.9640e-07, 2.3822e-07, 7.6810e-08, 4.8421e-07, 2.2558e-07, 5.3604e-07,\n 4.7887e-07, 5.1480e-07, 3.2878e-07, 6.5371e-07, 2.9507e-07, 2.9290e-07,\n 4.7715e-07, 4.1540e-07, 6.0668e-07, 4.6661e-07, 4.9506e-07, 2.3042e-07,\n 5.2414e-07, 3.8682e-07, 3.0927e-07, 2.1464e-07, 1.5432e-07, 2.3888e-07,\n 3.0761e-07, 1.9994e-07, 2.6545e-07, 3.9507e-07, 6.6907e-07, 4.2116e-07,\n 4.2793e-07, 5.7846e-07, 6.4881e-07, 4.4181e-07, 1.0540e-07, 7.9226e-07,\n 4.4806e-07, 4.6135e-07, 6.6322e-07, 3.3756e-07, 3.0772e-07, 6.5047e-07,\n 5.1120e-07, 4.5501e-07, 2.2361e-07, 7.9076e-07, 3.2080e-07, 1.0430e-09,\n 3.8042e-07, 4.0912e-07, 2.3389e-07, 2.9209e-07, 2.1960e-07, 3.2842e-07,\n 3.3549e-07, 4.0562e-07, 4.4690e-07, 2.8572e-07, 5.8275e-07, 5.0191e-07,\n 5.4766e-07, 2.8384e-07, 4.5149e-07, 2.3664e-07, 1.0455e-08, 6.0392e-07,\n 4.0145e-07, 2.8710e-07, 5.1689e-07, 4.0183e-07, 2.7466e-07, 1.9466e-07,\n 2.4865e-07, 2.5448e-07, 3.2309e-07, 1.9651e-07, 3.2357e-07, 8.9897e-07,\n 3.8397e-07, 4.2723e-07, 5.3729e-07, 4.8969e-07, 6.5440e-07, 5.0850e-07,\n 5.9515e-07, 5.0035e-07, 7.4258e-07, 5.0683e-07, 5.4641e-07, 2.0189e-07,\n 1.6596e-16, 6.6499e-07, 5.5023e-07, 3.1932e-07, 3.6105e-07, 6.1298e-07,\n 3.3548e-07, 3.3084e-07, 3.1938e-07, 6.1478e-07, 1.6463e-07, 1.2872e-07,\n 4.9612e-07, 5.8799e-07, 3.3898e-07, 4.0943e-07, 3.1607e-07, 4.3309e-07,\n 4.1719e-07, 1.7738e-07, 4.9472e-07, 5.5189e-07, 3.2298e-07, 2.5237e-07,\n 3.2492e-07, 4.5085e-07, 5.4742e-07, 5.8879e-07, 6.1076e-07, 8.3486e-07,\n 4.9658e-07, 2.4045e-07, 3.0293e-07, 4.7960e-07, 4.8984e-07, 4.0100e-07,\n 4.7781e-07, 2.6311e-07, 5.0894e-07, 3.5116e-07, 4.7168e-07, 8.5078e-07,\n 2.8361e-07, 5.2249e-07, 2.8813e-07, 3.2718e-07, 1.8626e-07, 2.5006e-07,\n 3.4277e-07, 3.4120e-07, 3.2144e-07, 5.0437e-07, 3.3093e-07, 5.0051e-07,\n 4.5540e-07, 6.1511e-07, 5.1284e-07, 4.3680e-07, 3.3738e-07, 3.0149e-07,\n 4.7970e-07, 2.5696e-07, 4.6504e-07, 3.9066e-07, 4.6030e-07, 3.8832e-07,\n 9.1776e-08, 2.8071e-07, 6.1818e-07, 4.4937e-07, 7.0595e-07, 4.4950e-07,\n 4.6531e-07, 4.7784e-07, 2.8059e-07, 7.4658e-07, 1.8031e-07, 2.5060e-07,\n 3.1726e-07, 6.5142e-07, 3.3772e-07, 1.8310e-07, 4.6568e-07, 4.4181e-07,\n 4.2281e-07, 3.6107e-07, 5.1441e-07, 5.3050e-07, 1.2671e-07, 4.3878e-07,\n 2.9601e-07, 5.5376e-07, 6.0354e-07, 5.0992e-07, 7.5289e-07, 2.2094e-07,\n 6.4931e-08, 5.4879e-07, 3.6594e-07, 3.7682e-07, 5.2444e-07, 3.6397e-07,\n 1.9318e-07, 2.9806e-07, 5.1638e-07, 6.3940e-07, 4.3617e-07, 3.0835e-07,\n 3.4528e-07, 5.7015e-07, 2.6914e-07, 4.9114e-07, 3.9302e-07, 2.3057e-07,\n 2.5393e-07, 3.7498e-07, 6.3723e-07, 4.4607e-07, 2.0030e-07, 5.8750e-07,\n 2.1584e-07, 4.1554e-07, 4.2400e-07, 3.4224e-07, 8.1866e-07, 2.9809e-07,\n 2.0921e-07, 4.7519e-07, 4.5442e-07, 2.4226e-07, 6.2201e-07, 4.0839e-07,\n 5.8200e-07, 6.4224e-07, 4.7172e-07, 2.5576e-07, 5.0983e-07, 5.0638e-07,\n 4.9579e-07, 2.4395e-07, 1.0644e-07, 4.3306e-07, 4.6720e-07, 3.9897e-07,\n 4.4289e-07, 8.2057e-09, 3.1663e-07, 9.4660e-07, 2.1463e-07, 1.7159e-07,\n 2.9124e-07, 5.2223e-07, 3.1503e-07, 5.6536e-07, 3.6594e-07, 4.0489e-07,\n 6.6517e-07, 2.4098e-07, 2.2393e-07, 4.5338e-07, 4.4010e-07, 3.3143e-07,\n 4.5810e-07, 2.1017e-07, 4.6123e-07, 6.2484e-07, 4.4140e-07, 4.5624e-07,\n 1.4758e-11, 5.0992e-07, 2.3912e-07, 5.1579e-07, 3.7319e-07, 5.0737e-07,\n 5.3290e-07, 3.6051e-07, 2.2441e-07, 3.5299e-07, 2.4764e-07, 3.4130e-07,\n 3.6462e-08, 2.5128e-07, 5.4420e-09, 4.0580e-07, 6.0589e-07, 5.6025e-07,\n 4.2791e-07, 5.6832e-07, 4.4755e-07, 1.6124e-07, 3.6226e-07, 6.3319e-07,\n 3.8873e-07, 4.2324e-07, 5.7647e-07, 5.2951e-07, 2.5311e-07, 2.2406e-07,\n 2.7735e-07, 4.2267e-07, 2.0156e-07, 6.0279e-07, 5.0002e-07, 2.6103e-07,\n 6.4814e-07, 3.5559e-07, 7.1142e-07, 4.5199e-07, 4.3067e-07, 7.7637e-07,\n 3.8712e-07, 4.8204e-07, 6.2978e-07, 3.1425e-07, 5.8953e-07, 6.5175e-07,\n 4.2345e-07, 2.8833e-07, 6.9242e-07, 3.0576e-07, 2.9983e-07, 2.4886e-07,\n 2.5190e-07, 3.7736e-07, 2.8962e-07, 2.9376e-07, 7.8339e-07, 3.6087e-07,\n 2.0723e-07, 4.6125e-07, 2.2941e-07, 7.7926e-07, 6.1360e-07, 5.5141e-07,\n 2.8900e-07, 3.6297e-07, 2.6125e-07, 2.0251e-07, 4.6401e-07, 1.6262e-07,\n 1.8248e-07, 3.6406e-07, 5.5654e-07, 4.4869e-07, 2.0911e-07, 2.9894e-07,\n 7.4050e-07, 3.9037e-07, 2.1039e-13, 3.3572e-07, 1.9306e-07, 6.3397e-07,\n 5.2783e-07, 1.8366e-07, 3.5176e-07, 4.9930e-07, 6.5063e-07, 4.7547e-07,\n 1.8812e-07, 3.3188e-07, 7.3594e-08, 4.3111e-07, 4.8595e-07, 2.3549e-07,\n 2.9108e-07, 5.8303e-07, 6.6009e-07, 1.8412e-07, 3.2794e-07, 5.4452e-07,\n 2.2323e-07, 5.0946e-07, 8.1826e-07, 2.7505e-07, 5.3600e-07, 4.5628e-07,\n 3.8038e-07, 2.9003e-07, 3.7763e-07, 4.8779e-07, 2.2477e-07, 3.1693e-07,\n 6.9090e-07, 3.7828e-07, 3.1496e-07, 3.8370e-07, 4.8291e-07, 5.3087e-07,\n 1.3978e-07, 5.1581e-07, 4.9655e-07, 6.9022e-07, 7.8373e-07, 4.8621e-07,\n 3.0570e-07, 6.3252e-07, 3.7424e-07, 5.7693e-18, 4.0560e-07, 7.3005e-07,\n 2.8521e-07, 5.1881e-07, 4.3237e-07, 5.1581e-07, 3.0775e-07, 5.5621e-07,\n 3.6184e-07, 9.2218e-07, 1.2871e-07, 4.3169e-07, 1.9843e-07, 1.0242e-07,\n 2.5739e-07, 4.5693e-07], device='cuda:0')" + "step": "tensor(7512.)", + "exp_avg": "tensor([-6.9714e-05, 1.3519e-04, 1.1274e-04, 1.8002e-04, 6.6106e-05,\n -7.8606e-05, -1.2633e-04, -2.5624e-04, 1.4824e-04, -4.3287e-05,\n -1.4767e-05, -2.1287e-04, -1.2944e-04, 1.0045e-04, -1.2051e-04,\n -2.2711e-05, 5.7637e-05, -2.3298e-04, 1.1542e-04, 1.2813e-04,\n -2.1763e-04, 6.7675e-05, -7.5142e-05, 5.2661e-05, 1.6310e-04,\n 3.5633e-04, -2.9645e-05, -9.6667e-05, 7.3332e-05, 6.2512e-05,\n 8.3744e-05, 5.5074e-04, -6.3867e-05, 2.1053e-04, -1.4129e-04,\n -4.4591e-05, 2.4502e-04, 1.2150e-04, -6.1692e-05, -1.0275e-04,\n -1.7773e-04, 4.4002e-05, 2.6397e-04, 1.8004e-05, -5.7789e-05,\n 1.3846e-05, 3.4827e-10, 1.4072e-04, 2.1824e-04, 2.1547e-04,\n -1.2002e-04, -8.3691e-05, -3.2442e-04, 9.6260e-05, -2.6402e-04,\n 2.7475e-04, -6.3193e-06, -7.9919e-05, -2.5383e-04, 6.4915e-05,\n 8.0956e-05, 1.6495e-04, 1.2132e-05, -1.0177e-04, 3.3316e-04,\n 3.6969e-05, 1.1297e-04, -1.6627e-04, 1.1285e-04, -3.4390e-05,\n -2.0929e-04, 5.6052e-45, 1.0640e-04, 4.2766e-05, 2.0711e-05,\n -1.9728e-04, 3.3022e-04, -5.6819e-05, -1.2980e-05, 2.7307e-04,\n -2.6605e-04, -2.6335e-05, 1.7724e-04, 1.4426e-04, -1.4610e-04,\n -7.1427e-05, -2.6415e-04, 8.4679e-08, 3.9154e-04, 3.5914e-04,\n 5.8498e-05, 2.3337e-04, 1.1317e-04, 3.5534e-04, 1.6972e-05,\n 8.1138e-05, -4.9024e-05, -5.2266e-05, -2.4156e-04, -1.5313e-04,\n -1.4220e-04, 2.4864e-05, -6.1122e-05, -1.6518e-04, 7.1210e-05,\n -1.4094e-04, 1.4973e-05, -2.2126e-04, 1.9487e-04, 1.0716e-04,\n 1.3105e-04, 9.7225e-05, 8.7451e-05, 1.3640e-04, 1.5161e-04,\n 4.9818e-05, -2.1077e-05, 2.4092e-04, 3.3867e-05, -1.7279e-04,\n -2.2476e-04, -1.0746e-04, 7.1516e-06, 9.7607e-05, -3.7104e-05,\n -1.8135e-04, -1.5506e-04, 3.1795e-05, -7.1982e-05, -7.1170e-05,\n -7.4031e-05, -2.2064e-04, 1.7014e-04, -2.4539e-05, 6.5915e-05,\n 1.1890e-04, 4.0588e-04, -1.1636e-04, 1.5736e-04, -2.1290e-04,\n -2.0132e-04, -1.6392e-04, 2.0796e-05, -2.5751e-04, 4.4426e-05,\n -7.2969e-05, -4.5633e-04, 1.0653e-04, -2.4307e-04, 1.6733e-04,\n 1.0203e-04, -1.7246e-05, 9.9892e-05, 2.5065e-04, 7.1177e-05,\n 1.6374e-07, 1.3576e-05, -1.3945e-05, 5.0143e-05, 1.6661e-04,\n -8.6322e-05, 1.4196e-04, -2.6786e-04, 1.0217e-05, -2.7483e-04,\n -4.9105e-06, -1.2862e-04, -1.0321e-04, -1.8297e-04, -1.0278e-04,\n -1.5641e-04, 1.5509e-05, -1.1228e-04, -1.1262e-05, 1.6463e-04,\n 1.7069e-04, -2.3565e-04, 1.1500e-04, -3.2799e-04, 8.9163e-05,\n -2.5360e-06, 8.8776e-06, 9.1944e-05, -2.2272e-04, 1.2651e-05,\n 1.5035e-04, 1.4162e-04, -1.4817e-04, -4.8330e-04, -1.4194e-04,\n -4.4695e-05, 5.2449e-05, 8.1479e-05, 6.0424e-05, -1.3454e-04,\n 1.5367e-04, -1.5542e-05, -2.2468e-05, 2.4486e-05, 1.4035e-04,\n -6.7703e-07, 2.3946e-04, -1.9327e-05, -1.2698e-04, -8.7427e-05,\n 2.6864e-04, 2.5876e-04, 7.6412e-06, -7.2347e-05, 9.6562e-05,\n 3.9476e-04, 7.7474e-05, 5.6052e-45, 3.3298e-05, 2.8513e-04,\n -4.7690e-05, -3.8084e-04, -3.3695e-05, 1.6989e-04, -1.3664e-04,\n 1.6858e-04, -4.8392e-06, -2.8184e-05, -8.5871e-05, 1.2791e-04,\n -5.4610e-05, 2.0779e-04, -7.6112e-05, 3.3158e-04, 3.3192e-04,\n 9.9192e-05, -6.8308e-05, -2.5454e-04, 9.5265e-05, 1.1344e-04,\n -5.6148e-05, -2.7322e-04, 4.2853e-07, -1.6939e-04, -3.3026e-04,\n 1.6522e-04, -2.1283e-04, 2.5097e-04, -1.3067e-04, -7.0704e-05,\n -3.6267e-05, -1.7516e-05, 3.7866e-05, -1.9762e-04, 9.4543e-06,\n 5.2238e-05, -4.9302e-06, 2.0742e-05, 2.7552e-04, 1.3011e-04,\n 5.6483e-05, 1.6350e-04, 1.3478e-04, 8.2139e-05, 1.3283e-04,\n -9.3219e-06, 3.1265e-05, -3.7300e-05, -5.6369e-05, 2.5202e-04,\n 1.0419e-04, -6.2360e-05, 1.3492e-04, -1.5450e-04, -2.4573e-05,\n 3.6344e-04, 1.2099e-04, 3.0914e-04, -9.5316e-05, 6.9173e-05,\n -4.6093e-05, -1.3407e-04, 3.4167e-04, -1.2435e-04, -8.9128e-05,\n -2.1129e-04, -1.7180e-04, 5.7479e-05, 1.1953e-04, 2.7053e-04,\n 4.6046e-05, -4.3853e-06, 8.8191e-05, -2.7273e-04, -1.4238e-05,\n 6.9996e-06, -5.1174e-05, -3.8938e-05, 1.1477e-04, 1.7287e-06,\n 3.5836e-04, 1.6345e-04, -1.8350e-05, 1.3530e-04, -1.4938e-04,\n -8.3312e-06, 3.7979e-05, -1.8585e-05, 2.4725e-04, -8.4710e-05,\n 3.0328e-04, -1.0087e-04, -1.4026e-04, -4.6320e-05, -3.2538e-05,\n -2.2627e-04, -7.1900e-05, -1.7510e-05, -7.6092e-05, -2.4006e-04,\n -1.6561e-04, 8.4916e-05, 3.1636e-05, 1.1141e-04, 2.3511e-05,\n 1.4724e-04, -1.9948e-04, -1.4235e-04, 1.9893e-07, -6.2315e-06,\n 8.3356e-05, -1.4316e-04, 3.1325e-05, -9.3570e-05, 7.2145e-05,\n -1.8538e-05, -1.3980e-05, 1.0821e-04, -9.2145e-05, 1.0637e-04,\n 6.1176e-06, -2.2610e-05, 1.4322e-04, -6.7234e-05, -1.4515e-04,\n -6.3198e-05, -4.3313e-05, 4.6493e-05, -6.3133e-05, 4.5620e-05,\n -3.6178e-06, -9.2456e-05, -2.3232e-05, 1.0229e-04, -1.0941e-04,\n -1.6821e-04, -1.7877e-04, 3.0735e-05, 6.9517e-05, -1.2001e-04,\n 9.2955e-05, 1.1934e-04, 3.1763e-04, 1.6053e-05, 1.7199e-04,\n -8.1573e-05, 2.0880e-04, 4.8977e-05, -3.0938e-04, -7.5462e-05,\n 2.4597e-04, 5.6052e-45, -1.5408e-04, -8.6137e-05, -1.3715e-05,\n 1.7441e-04, 2.5224e-04, 9.6214e-05, -5.2259e-05, -5.5476e-06,\n 1.2932e-04, 9.8128e-05, 3.0892e-05, 9.0144e-05, 6.4694e-05,\n -3.1881e-07, 2.6323e-04, 9.2618e-05, 1.5024e-04, -5.7530e-05,\n -8.8069e-05, -2.4589e-05, -1.6350e-04, 1.7903e-04, -4.5841e-05,\n -7.7514e-05, -1.7588e-04, -3.2243e-04, 1.0588e-04, -2.8530e-05,\n 6.2576e-06, -5.4933e-05, -6.6323e-05, -5.3146e-05, 2.5377e-04,\n -1.9290e-04, -1.6248e-04, 1.1579e-04, 1.9292e-05, -4.4180e-06,\n 1.1878e-04, 2.2192e-04, 1.6945e-04, 1.6897e-05, -1.2430e-04,\n -2.1568e-04, 1.3177e-04, 1.4874e-04, 2.0192e-04, 6.3095e-05,\n -1.8330e-04, 3.7707e-05, -1.0745e-04, 1.0460e-04, 2.8610e-05,\n -1.1536e-05, -1.1756e-04, -5.6247e-06, -4.2898e-07, -2.4053e-07,\n 2.1461e-04, -4.3640e-05, 1.0580e-04, 1.0940e-04, -8.5417e-05,\n 1.0562e-05, 9.6867e-05, 4.9086e-04, -8.4268e-06, 1.3946e-04,\n 1.6394e-04, 7.9345e-05, -1.2280e-04, -3.1077e-05, 6.9653e-05,\n 8.1315e-05, 2.7675e-04, 1.6817e-04, -2.1060e-05, 3.1326e-04,\n 1.7473e-04, -1.1184e-06, 1.8730e-04, 3.0076e-04, -4.9158e-04,\n 1.1461e-04, 2.4770e-04, 1.1471e-04, 2.5517e-04, -1.3379e-04,\n 9.4656e-05, 2.0992e-04, -9.2186e-05, 1.3382e-04, 4.2769e-05,\n 6.7000e-06, 8.9245e-05, -1.9426e-04, 1.0885e-04, -2.4979e-04,\n 3.0906e-05, -5.5381e-05, -1.5541e-04, -3.8700e-05, -4.3848e-05,\n 1.5576e-05, 1.4768e-06, 1.8818e-04, -6.6947e-04, 7.8634e-05,\n -2.2848e-04, -3.5399e-04, -1.2413e-04, 7.2649e-05, -2.1785e-04,\n -1.9180e-04, 8.2942e-05, -1.8548e-04, -9.4183e-05, 2.2088e-04,\n -1.8203e-04, -1.2546e-04, -3.4195e-05, 2.1179e-06, 1.2085e-04,\n -5.3214e-05, -1.2829e-04, -1.4882e-04, 4.5423e-05, -3.2520e-04,\n 2.4849e-05, -3.1402e-05, -3.4883e-05, 6.7005e-05, 5.5191e-06,\n -7.1970e-05, 6.1761e-05, -1.4480e-05, 7.2007e-05, -1.9176e-04,\n -1.7207e-04, -1.1276e-04, 8.3445e-05, 5.0981e-05, -7.1939e-05,\n 4.2634e-05, 1.0695e-04], device='cuda:0')", + "exp_avg_sq": "tensor([2.3484e-07, 3.4523e-07, 2.8980e-07, 4.0010e-07, 2.3056e-07, 3.5879e-07,\n 1.7751e-07, 3.7942e-07, 2.5364e-07, 2.1221e-07, 4.8565e-07, 3.2009e-07,\n 1.3705e-07, 2.9571e-07, 9.1714e-08, 3.3057e-07, 3.9179e-07, 4.2216e-07,\n 2.4649e-07, 2.0673e-07, 1.2739e-07, 2.9929e-07, 5.5926e-07, 2.2805e-07,\n 3.4529e-07, 3.3547e-07, 4.0639e-07, 4.2706e-07, 1.4161e-07, 1.5150e-07,\n 3.5740e-07, 4.7627e-07, 4.4767e-07, 5.4512e-07, 1.7502e-07, 2.8510e-07,\n 2.6220e-07, 4.8846e-07, 4.8495e-07, 2.9050e-07, 5.0480e-07, 5.5083e-07,\n 4.2450e-07, 1.7542e-07, 4.0138e-07, 4.3543e-07, 1.3353e-07, 2.2788e-07,\n 4.9204e-07, 2.5495e-07, 2.6947e-07, 1.9157e-07, 4.1023e-07, 3.0045e-07,\n 2.7419e-07, 4.6800e-07, 3.7042e-07, 3.3994e-07, 4.2730e-07, 4.0988e-07,\n 2.9470e-07, 4.1016e-07, 3.5046e-07, 3.9769e-07, 2.8443e-07, 1.6581e-07,\n 3.1908e-07, 2.8107e-07, 2.7161e-07, 2.0186e-07, 3.5756e-07, 6.4004e-16,\n 3.0868e-07, 3.0567e-07, 2.1645e-07, 3.4055e-07, 4.2281e-07, 3.9383e-07,\n 4.4618e-07, 3.3713e-07, 2.4529e-07, 2.8206e-07, 2.6426e-07, 3.5254e-07,\n 2.7049e-07, 1.8680e-07, 3.1552e-07, 5.7880e-07, 2.3224e-07, 4.0148e-07,\n 2.6297e-07, 4.5406e-07, 2.6326e-07, 3.9296e-07, 1.3638e-07, 4.9253e-07,\n 4.8436e-07, 4.9697e-07, 6.2203e-07, 2.6642e-07, 4.1787e-07, 3.1733e-07,\n 3.3737e-07, 3.7164e-07, 2.0734e-07, 2.5327e-07, 3.8486e-07, 2.3550e-07,\n 2.2459e-07, 2.0226e-07, 7.3097e-08, 4.2133e-07, 1.9766e-07, 4.2085e-07,\n 3.4488e-07, 3.9416e-07, 3.0812e-07, 4.9441e-07, 2.4069e-07, 2.2782e-07,\n 3.7473e-07, 3.1177e-07, 5.4191e-07, 3.6062e-07, 4.4504e-07, 1.6198e-07,\n 4.8120e-07, 2.6326e-07, 2.6445e-07, 1.6025e-07, 1.5332e-07, 2.0230e-07,\n 2.4999e-07, 1.6102e-07, 2.0636e-07, 3.1739e-07, 5.2207e-07, 3.4685e-07,\n 3.0204e-07, 4.6772e-07, 4.8843e-07, 3.6181e-07, 1.1956e-07, 5.8326e-07,\n 3.1692e-07, 3.7530e-07, 4.8054e-07, 2.6455e-07, 2.6143e-07, 4.8949e-07,\n 3.9701e-07, 3.8237e-07, 1.5956e-07, 5.3925e-07, 2.4722e-07, 4.3870e-09,\n 3.3467e-07, 3.3158e-07, 1.7585e-07, 2.3443e-07, 2.1144e-07, 2.5169e-07,\n 2.7383e-07, 2.7468e-07, 3.5168e-07, 2.5856e-07, 5.1630e-07, 4.1051e-07,\n 4.8289e-07, 2.4133e-07, 3.8907e-07, 1.9557e-07, 9.6363e-09, 4.0477e-07,\n 3.2402e-07, 2.4647e-07, 4.0122e-07, 3.3062e-07, 2.4780e-07, 1.7546e-07,\n 1.9458e-07, 2.4670e-07, 2.4928e-07, 1.5525e-07, 3.4435e-07, 5.3281e-07,\n 3.4440e-07, 3.3053e-07, 4.0465e-07, 3.9065e-07, 4.4158e-07, 4.4990e-07,\n 4.5687e-07, 3.8015e-07, 5.5183e-07, 3.7464e-07, 4.4103e-07, 1.8558e-07,\n 8.7023e-09, 5.1698e-07, 4.2859e-07, 2.7684e-07, 3.2792e-07, 4.5686e-07,\n 2.3889e-07, 2.9437e-07, 1.9066e-07, 5.1168e-07, 1.4372e-07, 1.0973e-07,\n 3.8068e-07, 3.4339e-07, 9.6866e-08, 3.0658e-07, 2.1661e-07, 3.2068e-07,\n 3.2389e-07, 1.5573e-07, 3.8772e-07, 3.9329e-07, 2.7666e-07, 2.0948e-07,\n 2.3067e-07, 3.4269e-07, 4.2738e-07, 4.6542e-07, 5.1523e-07, 4.0939e-07,\n 4.1210e-07, 2.2780e-07, 2.3717e-07, 3.3180e-07, 3.8317e-07, 3.3546e-07,\n 3.5373e-07, 2.2268e-07, 3.8207e-07, 2.7498e-07, 3.6383e-07, 4.9230e-07,\n 2.1341e-07, 4.2006e-07, 2.5287e-07, 2.5450e-07, 1.4961e-07, 2.2705e-07,\n 3.2021e-07, 2.7226e-07, 2.3898e-07, 3.6475e-07, 2.6125e-07, 4.0450e-07,\n 3.7709e-07, 4.4893e-07, 4.6062e-07, 3.4971e-07, 2.7557e-07, 2.4387e-07,\n 3.8852e-07, 2.3231e-07, 3.8897e-07, 3.5328e-07, 3.6806e-07, 3.2510e-07,\n 1.0128e-07, 2.2267e-07, 4.3328e-07, 3.6603e-07, 5.4180e-07, 3.8802e-07,\n 3.8050e-07, 3.8042e-07, 2.5781e-07, 5.7180e-07, 1.5410e-07, 1.8744e-07,\n 2.1202e-07, 4.5640e-07, 2.9099e-07, 1.6002e-07, 4.0175e-07, 3.0082e-07,\n 3.1421e-07, 3.1711e-07, 3.5100e-07, 4.2643e-07, 1.6595e-07, 3.4165e-07,\n 2.6454e-07, 4.7596e-07, 4.1875e-07, 3.7586e-07, 5.2030e-07, 2.0891e-07,\n 1.9862e-08, 4.4982e-07, 2.8343e-07, 3.0457e-07, 4.4159e-07, 2.9207e-07,\n 1.6169e-07, 2.5554e-07, 4.4094e-07, 4.8096e-07, 3.1480e-07, 2.6101e-07,\n 2.7344e-07, 4.3554e-07, 2.2474e-07, 4.1634e-07, 3.0046e-07, 1.9219e-07,\n 1.7587e-07, 3.1722e-07, 5.0927e-07, 3.5550e-07, 1.7958e-07, 4.2465e-07,\n 1.3277e-07, 3.2181e-07, 2.9845e-07, 2.5959e-07, 5.6416e-07, 2.3495e-07,\n 1.9253e-07, 3.8576e-07, 3.2380e-07, 2.0339e-07, 5.0646e-07, 3.0522e-07,\n 4.4877e-07, 4.6803e-07, 3.4373e-07, 2.1413e-07, 4.0431e-07, 3.6821e-07,\n 4.1843e-07, 2.2614e-07, 1.0396e-07, 3.3722e-07, 3.3883e-07, 3.2893e-07,\n 3.0596e-07, 3.7871e-08, 2.2876e-07, 4.2140e-07, 1.6679e-07, 1.4155e-07,\n 2.3280e-07, 3.9643e-07, 2.6393e-07, 3.1798e-07, 3.0924e-07, 3.2052e-07,\n 4.6076e-07, 1.9674e-07, 2.2476e-07, 3.5502e-07, 3.4182e-07, 2.6134e-07,\n 3.4035e-07, 1.5177e-07, 3.5356e-07, 4.6295e-07, 3.8389e-07, 3.0402e-07,\n 4.2172e-12, 3.9691e-07, 2.1018e-07, 4.3938e-07, 3.0089e-07, 4.2108e-07,\n 4.8021e-07, 2.2645e-07, 2.0944e-07, 2.9945e-07, 1.8960e-07, 2.6683e-07,\n 2.0557e-08, 2.2037e-07, 1.5605e-09, 3.1800e-07, 4.6110e-07, 4.2018e-07,\n 3.8691e-07, 4.1593e-07, 3.7150e-07, 1.4180e-07, 2.7798e-07, 4.5462e-07,\n 2.9128e-07, 3.0500e-07, 4.1424e-07, 4.6996e-07, 1.5187e-07, 1.9815e-07,\n 2.2803e-07, 2.9763e-07, 2.1735e-07, 4.3030e-07, 4.7170e-07, 1.7109e-07,\n 5.4666e-07, 2.7116e-07, 5.2917e-07, 3.6981e-07, 3.4845e-07, 5.2574e-07,\n 3.1740e-07, 3.6286e-07, 4.3006e-07, 2.4068e-07, 3.5580e-07, 5.0134e-07,\n 3.3937e-07, 2.2718e-07, 4.6558e-07, 2.7290e-07, 2.5284e-07, 2.1528e-07,\n 1.3643e-07, 3.1536e-07, 2.4498e-07, 1.9125e-07, 5.5960e-07, 2.9818e-07,\n 1.6598e-07, 3.1352e-07, 1.9465e-07, 5.8720e-07, 4.2048e-07, 4.0816e-07,\n 2.3905e-07, 3.1174e-07, 2.2317e-07, 1.7815e-07, 3.5637e-07, 1.4328e-07,\n 8.9335e-08, 2.4072e-07, 4.7636e-07, 3.7252e-07, 1.8333e-07, 2.0136e-07,\n 5.4898e-07, 2.8562e-07, 1.3597e-08, 3.8573e-07, 2.3715e-07, 4.9317e-07,\n 4.1030e-07, 1.5620e-07, 2.9919e-07, 4.1633e-07, 4.8111e-07, 3.8638e-07,\n 1.7510e-07, 2.7930e-07, 9.7520e-08, 3.0281e-07, 3.8524e-07, 2.1892e-07,\n 2.6979e-07, 4.3323e-07, 5.0907e-07, 1.4743e-07, 2.3813e-07, 4.1465e-07,\n 2.1017e-07, 3.2698e-07, 6.1126e-07, 1.8241e-07, 3.9687e-07, 3.8048e-07,\n 3.2325e-07, 2.4119e-07, 2.9082e-07, 4.1918e-07, 2.0284e-07, 2.9801e-07,\n 5.8869e-07, 2.8416e-07, 2.5520e-07, 3.2856e-07, 4.0016e-07, 4.0616e-07,\n 1.1497e-07, 3.8738e-07, 3.9613e-07, 5.0616e-07, 5.4972e-07, 3.6569e-07,\n 2.5772e-07, 4.6209e-07, 3.1162e-07, 5.6838e-10, 3.0791e-07, 5.4768e-07,\n 2.1253e-07, 3.7951e-07, 3.9375e-07, 3.6480e-07, 2.4928e-07, 4.0563e-07,\n 3.3601e-07, 4.4763e-07, 1.3043e-07, 3.1942e-07, 1.6386e-07, 1.1815e-07,\n 2.3013e-07, 3.5062e-07], device='cuda:0')" }, "4": { - "step": "tensor(6260.)", - "exp_avg": "tensor([[ 1.8051e-06, 7.7717e-06, -1.1762e-05, ..., -1.7779e-06,\n -1.6077e-06, -7.2762e-06],\n [ 3.8044e-06, -8.7437e-06, -2.4373e-05, ..., 1.3913e-06,\n 9.0766e-08, -6.5949e-06],\n [ 6.6412e-06, -1.3297e-05, 4.6889e-05, ..., 2.4836e-07,\n 1.4580e-05, 6.9022e-06],\n ...,\n [ 8.5893e-05, 2.2444e-06, -3.4216e-05, ..., -1.5663e-06,\n 1.0801e-05, -4.8861e-06],\n [-5.1267e-06, -1.1652e-05, -2.3581e-05, ..., -1.0031e-07,\n 8.5159e-07, 1.2613e-05],\n [ 4.8896e-05, -1.6472e-05, -3.4303e-05, ..., 9.9761e-07,\n -1.0900e-05, -1.6162e-05]], device='cuda:0')", - "exp_avg_sq": "tensor([[4.1308e-09, 1.6604e-09, 1.8253e-09, ..., 9.7495e-11, 6.8655e-10,\n 4.2496e-09],\n [8.3453e-09, 5.4105e-09, 6.6297e-09, ..., 1.1778e-10, 1.0796e-09,\n 8.1224e-09],\n [5.3479e-09, 4.2595e-09, 3.1662e-09, ..., 2.6425e-10, 1.9064e-09,\n 4.0530e-09],\n ...,\n [3.1641e-08, 3.7727e-09, 7.7409e-09, ..., 1.6950e-10, 3.6582e-09,\n 7.5386e-09],\n [7.7039e-09, 3.5320e-09, 6.5373e-09, ..., 1.2939e-10, 6.4699e-09,\n 9.9044e-09],\n [9.4282e-09, 5.2435e-09, 9.9640e-09, ..., 2.3744e-10, 1.0114e-09,\n 1.0140e-08]], device='cuda:0')" + "step": "tensor(7512.)", + "exp_avg": "tensor([[ 2.0178e-05, -8.1219e-06, -6.5636e-06, ..., 1.3333e-06,\n -8.3569e-06, 9.3330e-07],\n [-7.9213e-05, 1.0441e-05, -8.3200e-07, ..., -2.4622e-06,\n -2.2825e-06, 1.0565e-05],\n [ 5.3407e-05, -5.9471e-05, -4.6203e-06, ..., -6.0558e-07,\n 9.5238e-06, 3.0464e-06],\n ...,\n [-1.7153e-04, -2.8900e-05, -7.3160e-06, ..., 3.1479e-06,\n 1.4445e-05, 1.1287e-05],\n [ 5.9715e-05, 7.2422e-06, -1.5574e-05, ..., -2.9913e-07,\n 1.4758e-05, 5.6405e-05],\n [ 8.7183e-05, 8.7101e-06, 6.5730e-06, ..., -2.8533e-06,\n -1.4575e-06, -2.5082e-05]], device='cuda:0')", + "exp_avg_sq": "tensor([[3.5039e-09, 1.3377e-09, 1.2632e-09, ..., 8.0407e-11, 5.6991e-10,\n 3.4550e-09],\n [6.6532e-09, 4.2483e-09, 4.9993e-09, ..., 9.6561e-11, 9.3976e-10,\n 6.7537e-09],\n [4.5652e-09, 3.3875e-09, 2.2508e-09, ..., 2.4130e-10, 1.5463e-09,\n 3.3175e-09],\n ...,\n [2.9841e-08, 2.9000e-09, 5.7207e-09, ..., 1.3412e-10, 3.1712e-09,\n 6.0828e-09],\n [6.2803e-09, 2.7406e-09, 4.9615e-09, ..., 1.1379e-10, 5.2813e-09,\n 8.5720e-09],\n [8.1238e-09, 4.0145e-09, 8.3124e-09, ..., 2.0752e-10, 8.4866e-10,\n 9.0206e-09]], device='cuda:0')" }, "5": { - "step": "tensor(5008.)", - "exp_avg": "tensor([[ 2.9052e-07, -2.2928e-08, -4.3634e-08, ..., -6.0430e-07,\n 0.0000e+00, 9.5430e-07],\n [-2.2505e-07, 4.2085e-07, 5.5051e-07, ..., 1.4380e-06,\n 0.0000e+00, -1.6246e-08],\n [-1.1071e-06, -1.2538e-07, 1.2875e-07, ..., -3.5246e-06,\n 0.0000e+00, -2.3430e-06],\n ...,\n [ 7.2477e-07, 7.6364e-07, 3.4231e-07, ..., -9.2964e-08,\n 0.0000e+00, -4.8257e-07],\n [-8.8239e-08, 1.4731e-07, 1.9609e-06, ..., 1.0204e-06,\n 0.0000e+00, 1.1838e-06],\n [-8.0134e-07, -1.5626e-06, -5.1318e-07, ..., -3.0431e-06,\n 0.0000e+00, 1.0494e-06]], device='cuda:0')", - "exp_avg_sq": "tensor([[4.0832e-11, 7.7793e-12, 1.9464e-11, ..., 8.1092e-12, 0.0000e+00,\n 1.2444e-11],\n [4.0658e-11, 8.6062e-12, 2.6192e-11, ..., 3.9069e-11, 0.0000e+00,\n 1.8913e-11],\n [2.4880e-11, 1.2028e-11, 2.6414e-11, ..., 7.8947e-11, 0.0000e+00,\n 1.1963e-10],\n ...,\n [4.8356e-11, 1.9637e-11, 1.0225e-11, ..., 4.4803e-12, 0.0000e+00,\n 8.1128e-12],\n [1.7608e-12, 4.0500e-12, 1.1602e-10, ..., 1.0861e-10, 0.0000e+00,\n 2.0399e-11],\n [1.0155e-11, 1.2418e-11, 2.5896e-11, ..., 3.7060e-10, 0.0000e+00,\n 4.8629e-11]], device='cuda:0')" + "step": "tensor(6260.)", + "exp_avg": "tensor([[ 2.8242e-07, -3.8365e-06, -7.8467e-07, ..., 5.5767e-07,\n 0.0000e+00, 8.6096e-08],\n [ 1.1404e-06, -5.9890e-07, -2.3639e-09, ..., 5.0988e-07,\n 0.0000e+00, -5.7501e-07],\n [ 6.3628e-07, 2.2809e-06, 1.7754e-07, ..., 2.4754e-06,\n 0.0000e+00, 4.4769e-07],\n ...,\n [ 1.8124e-08, -3.1286e-07, 2.3418e-08, ..., 1.3328e-07,\n 0.0000e+00, 1.1637e-07],\n [-1.3784e-08, 1.4707e-07, -7.1477e-07, ..., -5.2046e-07,\n 0.0000e+00, -3.3927e-07],\n [ 1.1686e-07, 1.2986e-06, -4.8887e-07, ..., 3.2425e-06,\n 0.0000e+00, -5.5866e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[3.1801e-11, 7.5085e-12, 1.7052e-11, ..., 5.8299e-12, 0.0000e+00,\n 1.1730e-11],\n [2.9533e-11, 1.0374e-11, 1.7128e-11, ..., 2.9414e-11, 0.0000e+00,\n 1.3305e-11],\n [2.1325e-11, 9.6851e-12, 1.8477e-11, ..., 6.2010e-11, 0.0000e+00,\n 9.3464e-11],\n ...,\n [4.8922e-11, 1.2593e-11, 8.1611e-12, ..., 2.7990e-12, 0.0000e+00,\n 4.2076e-12],\n [1.0561e-12, 2.3769e-12, 8.6782e-11, ..., 8.9796e-11, 0.0000e+00,\n 1.9322e-11],\n [6.8887e-12, 1.1987e-11, 1.5949e-11, ..., 2.9406e-10, 0.0000e+00,\n 3.2394e-11]], device='cuda:0')" }, "6": { - "step": "tensor(5008.)", - "exp_avg": "tensor([-2.9050e-05, 3.7621e-06, -5.0623e-05, ..., 3.0815e-06,\n -2.6227e-05, 5.0914e-06], device='cuda:0')", - "exp_avg_sq": "tensor([4.7790e-09, 3.2526e-09, 4.1045e-09, ..., 2.5641e-09, 3.0099e-09,\n 2.8913e-09], device='cuda:0')" + "step": "tensor(6260.)", + "exp_avg": "tensor([-4.3102e-05, 1.6343e-05, 1.5130e-05, ..., 1.3290e-05,\n 1.9349e-05, 5.4179e-06], device='cuda:0')", + "exp_avg_sq": "tensor([3.7326e-09, 2.3368e-09, 3.3308e-09, ..., 1.8181e-09, 2.3493e-09,\n 2.2518e-09], device='cuda:0')" }, "7": { - "step": "tensor(5008.)", - "exp_avg": "tensor([[ 2.7085e-07, -4.0476e-07, 5.9547e-07, ..., 5.3408e-07,\n -1.3915e-06, -1.4484e-06],\n [ 5.7098e-08, 1.6263e-06, -1.3432e-06, ..., -8.1884e-07,\n -9.3186e-07, 4.7864e-07],\n [-3.6681e-07, 1.1444e-06, 5.7836e-07, ..., 5.7105e-07,\n -5.0189e-07, 1.1652e-06],\n ...,\n [-9.8853e-07, -3.5932e-06, -2.1081e-06, ..., -1.0942e-06,\n -2.0264e-06, 8.8557e-07],\n [-1.1547e-06, 1.0930e-06, -1.8229e-07, ..., -6.1655e-07,\n 1.6886e-06, 1.7918e-07],\n [ 2.8115e-06, -1.4956e-06, 2.8647e-07, ..., 4.3480e-06,\n 2.1652e-07, 1.1223e-06]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.4066e-11, 6.0948e-12, 7.8347e-12, ..., 7.1691e-12, 6.5819e-12,\n 7.8594e-12],\n [1.8706e-11, 1.1371e-11, 1.3297e-11, ..., 1.5869e-11, 1.2163e-11,\n 1.2811e-11],\n [1.8208e-11, 1.2180e-11, 1.1863e-11, ..., 1.2269e-11, 1.0056e-11,\n 1.6568e-11],\n ...,\n [1.7239e-11, 2.1737e-11, 2.0287e-11, ..., 1.3141e-11, 1.5983e-11,\n 1.4509e-11],\n [1.5653e-11, 1.1913e-11, 1.5217e-11, ..., 9.6066e-12, 1.0947e-11,\n 1.5791e-11],\n [2.1784e-11, 1.2656e-11, 1.6639e-11, ..., 5.4514e-11, 1.2477e-11,\n 1.0659e-11]], device='cuda:0')" + "step": "tensor(6260.)", + "exp_avg": "tensor([[-5.4170e-07, -1.1130e-07, -5.8549e-07, ..., -4.7177e-07,\n -6.2891e-08, -5.0514e-07],\n [-3.6957e-07, 2.1145e-06, -3.4050e-07, ..., 3.2896e-06,\n -3.9562e-07, 1.1057e-06],\n [-1.3433e-06, -1.7163e-06, -7.0609e-07, ..., 9.2374e-07,\n 6.0960e-07, -1.2995e-07],\n ...,\n [ 9.1756e-07, -1.1193e-06, 6.7311e-07, ..., 7.5302e-07,\n 2.5145e-06, -1.2271e-06],\n [-8.4692e-07, 2.2536e-07, -1.1305e-06, ..., -5.0420e-07,\n -4.3851e-07, -2.6687e-07],\n [-4.1283e-07, 1.6473e-06, -6.0153e-07, ..., -1.5008e-08,\n 1.1367e-06, -4.8076e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[9.2673e-12, 3.9961e-12, 5.1676e-12, ..., 3.9849e-12, 4.5910e-12,\n 5.2355e-12],\n [1.2835e-11, 7.5428e-12, 8.9072e-12, ..., 1.0853e-11, 9.3711e-12,\n 9.1924e-12],\n [1.2509e-11, 7.7540e-12, 7.9735e-12, ..., 7.5534e-12, 7.9635e-12,\n 1.2532e-11],\n ...,\n [1.0976e-11, 1.5140e-11, 1.4545e-11, ..., 8.6860e-12, 1.3039e-11,\n 1.0256e-11],\n [1.0723e-11, 7.3437e-12, 1.0388e-11, ..., 5.7471e-12, 8.2322e-12,\n 1.1442e-11],\n [1.5633e-11, 8.5535e-12, 1.1718e-11, ..., 4.1321e-11, 9.0040e-12,\n 6.4627e-12]], device='cuda:0')" }, "32": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([2.3813e-07], device='cuda:0')" + "exp_avg_sq": "tensor([6.8048e-08], device='cuda:0')" }, "33": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, -5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([2.1677e-09, 1.0537e-08, 3.2088e-09], device='cuda:0')" + "exp_avg_sq": "tensor([6.1945e-10, 3.0111e-09, 9.1694e-10], device='cuda:0')" }, "34": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45],\n device='cuda:0')", - "exp_avg_sq": "tensor([1.2936e-05, 1.1728e-07, 1.5841e-07, 1.6438e-07, 1.8847e-07, 2.1650e-07,\n 1.7212e-07, 1.2747e-07, 1.6100e-07, 1.4307e-07], device='cuda:0')" + "exp_avg_sq": "tensor([3.6966e-06, 3.3514e-08, 4.5266e-08, 4.6973e-08, 5.3856e-08, 6.1865e-08,\n 4.9186e-08, 3.6424e-08, 4.6007e-08, 4.0884e-08], device='cuda:0')" }, "36": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [-5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.9978e-14, 9.0025e-15, 1.3375e-14, ..., 5.8468e-14, 0.0000e+00,\n 1.9733e-14],\n [3.5963e-15, 8.6944e-15, 2.6151e-14, ..., 2.8911e-14, 0.0000e+00,\n 1.1606e-14],\n [3.2274e-12, 7.6360e-13, 4.2429e-11, ..., 1.3603e-11, 0.0000e+00,\n 3.7956e-11],\n ...,\n [8.9948e-13, 3.7408e-14, 7.0843e-12, ..., 3.9534e-12, 0.0000e+00,\n 1.0637e-11],\n [2.4244e-14, 8.7240e-15, 2.5640e-15, ..., 9.5197e-14, 0.0000e+00,\n 4.9087e-14],\n [8.1793e-16, 1.5347e-14, 2.0049e-15, ..., 5.1406e-15, 0.0000e+00,\n 2.0940e-14]], device='cuda:0')" + "exp_avg_sq": "tensor([[5.7089e-15, 2.5725e-15, 3.8221e-15, ..., 1.6708e-14, 0.0000e+00,\n 5.6389e-15],\n [1.0277e-15, 2.4845e-15, 7.4727e-15, ..., 8.2616e-15, 0.0000e+00,\n 3.3166e-15],\n [9.2226e-13, 2.1821e-13, 1.2124e-11, ..., 3.8871e-12, 0.0000e+00,\n 1.0846e-11],\n ...,\n [2.5703e-13, 1.0690e-14, 2.0244e-12, ..., 1.1297e-12, 0.0000e+00,\n 3.0396e-12],\n [6.9280e-15, 2.4930e-15, 7.3268e-16, ..., 2.7203e-14, 0.0000e+00,\n 1.4027e-14],\n [2.3373e-16, 4.3856e-15, 5.7293e-16, ..., 1.4690e-15, 0.0000e+00,\n 5.9838e-15]], device='cuda:0')" }, "37": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([3.5479e-11, 1.2245e-11, 4.6173e-08, 6.1268e-11, 2.2287e-08, 7.6758e-09,\n 1.4685e-09, 1.6852e-09, 4.2586e-08, 1.5583e-10, 2.1651e-08, 2.6345e-09,\n 2.3354e-11, 4.8711e-10, 3.4389e-12, 2.4702e-10, 3.5240e-09, 3.9909e-09,\n 2.5803e-10, 3.0989e-10, 3.7796e-09, 1.7438e-10, 2.6284e-09, 1.5656e-09,\n 1.3232e-09, 1.3183e-08, 2.5059e-09, 2.1820e-09, 1.6605e-09, 3.3095e-08,\n 1.2159e-08, 7.5240e-09, 8.7648e-08, 1.4254e-10, 1.1743e-09, 2.6267e-10,\n 3.6516e-09, 1.6229e-10, 1.4740e-08, 3.8932e-10, 1.4597e-08, 2.4892e-10,\n 7.7537e-10, 1.3820e-08, 2.2944e-11, 8.3743e-10, 1.2955e-09, 1.2917e-09,\n 2.6382e-10, 9.3042e-09, 3.2149e-09, 3.4474e-09, 2.3094e-08, 1.5104e-09,\n 2.5661e-10, 1.8493e-09, 3.6877e-13, 9.0062e-10, 3.9670e-09, 3.8889e-08,\n 4.7947e-09, 4.2538e-10, 1.3559e-11, 1.4481e-09, 1.6455e-09, 2.1766e-08,\n 1.9861e-10, 1.4583e-09, 4.7860e-10, 3.3452e-09, 2.0477e-09, 1.7861e-09,\n 5.5110e-10, 8.1518e-09, 1.1459e-08, 6.4706e-09, 7.8509e-09, 1.8956e-09,\n 5.6690e-10, 1.5675e-11, 1.9414e-09, 7.5427e-10, 1.9399e-10, 1.0924e-09,\n 1.5380e-09, 4.8332e-12, 1.6051e-08, 2.5757e-09, 4.6993e-10, 2.1004e-11,\n 1.5059e-10, 1.5371e-09, 8.7864e-09, 1.3755e-08, 1.2888e-08, 9.7868e-10,\n 1.1540e-08, 4.8378e-09, 3.3328e-09, 4.8035e-10, 1.4150e-09, 5.9385e-11,\n 2.5289e-10, 1.5162e-08, 6.0929e-09, 1.0643e-08, 3.4683e-09, 3.3259e-11,\n 1.4267e-08, 1.1920e-09, 2.0685e-09, 2.8603e-10, 3.0297e-09, 1.4728e-09,\n 1.3634e-09, 6.4464e-08, 4.7078e-11, 2.1756e-09, 4.1909e-10, 2.0143e-08,\n 2.0516e-08, 2.7997e-09, 6.4070e-09, 2.5240e-10, 3.1640e-10, 5.5276e-09,\n 2.9859e-09, 3.3730e-09, 4.5159e-10, 1.0579e-08, 8.3490e-12, 2.3854e-10,\n 5.7509e-08, 5.2433e-09, 9.8784e-11, 1.1378e-09, 3.5632e-09, 1.4181e-09,\n 2.3336e-08, 4.4948e-08, 1.0506e-09, 3.2503e-09, 2.7599e-10, 4.5542e-11,\n 5.0297e-10, 2.2788e-09, 2.3975e-09, 4.5208e-09, 1.7035e-10, 9.6365e-09,\n 8.6242e-11, 2.3726e-10, 7.9157e-09, 3.5301e-09, 2.9803e-09, 1.7216e-08,\n 2.4856e-10, 1.3209e-08, 1.1434e-10, 6.1727e-10, 1.4812e-11, 1.0650e-08,\n 9.8839e-12, 8.0410e-10, 1.3713e-08, 7.7832e-09, 3.6659e-10, 4.8140e-09,\n 4.0976e-11, 2.3992e-10, 1.0284e-09, 5.7059e-09, 1.3543e-08, 4.7687e-09,\n 2.8803e-09, 9.8009e-10, 1.4175e-08, 1.7174e-08, 4.1154e-11, 3.1486e-12,\n 3.1273e-11, 5.1175e-09, 4.1237e-10, 1.5600e-10, 5.3597e-10, 2.0011e-08,\n 1.2457e-09, 2.4551e-11, 4.4612e-10, 7.3899e-10, 8.7226e-09, 5.1564e-11,\n 8.1586e-09, 1.6295e-08, 4.8456e-09, 1.3196e-09, 3.9527e-10, 4.4637e-09,\n 1.2073e-09, 5.0518e-09, 1.3496e-09, 8.1166e-11, 4.2251e-08, 8.3211e-11,\n 2.8962e-09, 5.4242e-12, 3.2657e-09, 1.0863e-08, 2.8558e-11, 7.4243e-10,\n 1.2487e-08, 7.5474e-10, 5.7565e-09, 3.8747e-08, 3.5719e-08, 3.9376e-09,\n 5.8776e-11, 1.7933e-09, 1.0446e-08, 2.1473e-11, 2.4653e-09, 4.3438e-08,\n 1.3479e-09, 2.3601e-09, 7.8143e-09, 1.0053e-09, 3.8463e-10, 3.5114e-09,\n 2.1156e-09, 1.7640e-09, 9.9896e-10, 4.1275e-09, 1.8535e-09, 9.8731e-10,\n 4.2872e-09, 6.7889e-10, 1.3748e-09, 1.7219e-09, 4.1866e-09, 7.3752e-11,\n 2.0725e-08, 7.1628e-10, 1.9448e-10, 7.0694e-09, 2.4255e-09, 3.6291e-11,\n 3.7791e-09, 2.5139e-08, 1.8705e-08, 2.7413e-08, 9.2042e-10, 3.8629e-09,\n 2.9373e-09, 1.2797e-08, 1.0136e-10, 1.2750e-11], device='cuda:0')" + "exp_avg_sq": "tensor([1.0138e-11, 3.4991e-12, 1.3194e-08, 1.7508e-11, 6.3688e-09, 2.1934e-09,\n 4.1965e-10, 4.8155e-10, 1.2169e-08, 4.4529e-11, 6.1869e-09, 7.5282e-10,\n 6.6736e-12, 1.3920e-10, 9.8269e-13, 7.0587e-11, 1.0070e-09, 1.1404e-09,\n 7.3733e-11, 8.8553e-11, 1.0800e-09, 4.9829e-11, 7.5108e-10, 4.4740e-10,\n 3.7813e-10, 3.7672e-09, 7.1607e-10, 6.2351e-10, 4.7449e-10, 9.4572e-09,\n 3.4745e-09, 2.1500e-09, 2.5046e-08, 4.0732e-11, 3.3558e-10, 7.5059e-11,\n 1.0435e-09, 4.6376e-11, 4.2120e-09, 1.1125e-10, 4.1711e-09, 7.1131e-11,\n 2.2157e-10, 3.9492e-09, 6.5565e-12, 2.3930e-10, 3.7021e-10, 3.6912e-10,\n 7.5388e-11, 2.6587e-09, 9.1869e-10, 9.8514e-10, 6.5994e-09, 4.3160e-10,\n 7.3330e-11, 5.2844e-10, 1.0538e-13, 2.5736e-10, 1.1336e-09, 1.1113e-08,\n 1.3701e-09, 1.2156e-10, 3.8747e-12, 4.1382e-10, 4.7023e-10, 6.2197e-09,\n 5.6755e-11, 4.1672e-10, 1.3676e-10, 9.5592e-10, 5.8516e-10, 5.1039e-10,\n 1.5748e-10, 2.3294e-09, 3.2745e-09, 1.8490e-09, 2.2435e-09, 5.4168e-10,\n 1.6200e-10, 4.4793e-12, 5.5477e-10, 2.1554e-10, 5.5434e-11, 3.1215e-10,\n 4.3950e-10, 1.3811e-12, 4.5868e-09, 7.3603e-10, 1.3429e-10, 6.0020e-12,\n 4.3033e-11, 4.3923e-10, 2.5108e-09, 3.9307e-09, 3.6828e-09, 2.7967e-10,\n 3.2976e-09, 1.3824e-09, 9.5238e-10, 1.3727e-10, 4.0434e-10, 1.6970e-11,\n 7.2265e-11, 4.3328e-09, 1.7411e-09, 3.0412e-09, 9.9108e-10, 9.5041e-12,\n 4.0770e-09, 3.4062e-10, 5.9109e-10, 8.1736e-11, 8.6577e-10, 4.2085e-10,\n 3.8960e-10, 1.8421e-08, 1.3453e-11, 6.2169e-10, 1.1976e-10, 5.7561e-09,\n 5.8625e-09, 8.0003e-10, 1.8309e-09, 7.2126e-11, 9.0414e-11, 1.5796e-09,\n 8.5325e-10, 9.6386e-10, 1.2905e-10, 3.0231e-09, 2.3858e-12, 6.8163e-11,\n 1.6434e-08, 1.4983e-09, 2.8228e-11, 3.2514e-10, 1.0182e-09, 4.0524e-10,\n 6.6684e-09, 1.2844e-08, 3.0021e-10, 9.2880e-10, 7.8865e-11, 1.3014e-11,\n 1.4373e-10, 6.5120e-10, 6.8510e-10, 1.2918e-09, 4.8678e-11, 2.7537e-09,\n 2.4644e-11, 6.7799e-11, 2.2620e-09, 1.0087e-09, 8.5165e-10, 4.9197e-09,\n 7.1029e-11, 3.7746e-09, 3.2673e-11, 1.7639e-10, 4.2325e-12, 3.0432e-09,\n 2.8244e-12, 2.2978e-10, 3.9185e-09, 2.2241e-09, 1.0475e-10, 1.3757e-09,\n 1.1709e-11, 6.8560e-11, 2.9387e-10, 1.6305e-09, 3.8700e-09, 1.3627e-09,\n 8.2307e-10, 2.8007e-10, 4.0507e-09, 4.9077e-09, 1.1760e-11, 8.9973e-13,\n 8.9364e-12, 1.4624e-09, 1.1784e-10, 4.4579e-11, 1.5316e-10, 5.7182e-09,\n 3.5598e-10, 7.0157e-12, 1.2748e-10, 2.1117e-10, 2.4926e-09, 1.4735e-11,\n 2.3314e-09, 4.6564e-09, 1.3847e-09, 3.7710e-10, 1.1295e-10, 1.2755e-09,\n 3.4499e-10, 1.4436e-09, 3.8566e-10, 2.3194e-11, 1.2074e-08, 2.3778e-11,\n 8.2762e-10, 1.5500e-12, 9.3321e-10, 3.1043e-09, 8.1607e-12, 2.1216e-10,\n 3.5683e-09, 2.1567e-10, 1.6450e-09, 1.1072e-08, 1.0207e-08, 1.1252e-09,\n 1.6796e-11, 5.1245e-10, 2.9849e-09, 6.1360e-12, 7.0447e-10, 1.2413e-08,\n 3.8518e-10, 6.7442e-10, 2.2330e-09, 2.8726e-10, 1.0991e-10, 1.0034e-09,\n 6.0455e-10, 5.0406e-10, 2.8546e-10, 1.1795e-09, 5.2964e-10, 2.8213e-10,\n 1.2251e-09, 1.9400e-10, 3.9287e-10, 4.9206e-10, 1.1964e-09, 2.1075e-11,\n 5.9225e-09, 2.0468e-10, 5.5574e-11, 2.0201e-09, 6.9311e-10, 1.0370e-11,\n 1.0799e-09, 7.1837e-09, 5.3452e-09, 7.8336e-09, 2.6302e-10, 1.1038e-09,\n 8.3934e-10, 3.6569e-09, 2.8965e-11, 3.6433e-12], device='cuda:0')" }, "38": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([7.9704e-14, 9.1411e-17, 8.8704e-11, 3.0303e-13, 4.7183e-11, 1.4851e-11,\n 2.5882e-12, 3.7791e-12, 9.8010e-11, 3.8767e-13, 7.7471e-11, 4.5133e-12,\n 1.4122e-13, 6.0863e-13, 1.1052e-14, 8.9470e-13, 7.8248e-12, 1.0980e-11,\n 9.3614e-13, 6.7659e-13, 6.1462e-12, 4.2859e-13, 3.4815e-12, 2.6922e-12,\n 2.4777e-12, 1.6412e-11, 3.8433e-12, 3.9055e-12, 9.5015e-12, 9.6053e-11,\n 1.5007e-11, 1.3207e-11, 1.6564e-10, 2.0921e-13, 2.0429e-12, 9.9529e-13,\n 4.6437e-12, 4.8835e-13, 3.7353e-11, 5.4679e-13, 4.1141e-11, 8.4792e-13,\n 2.2833e-12, 1.8357e-11, 2.3008e-14, 9.2193e-13, 2.0377e-12, 1.8546e-12,\n 7.2091e-13, 2.2153e-11, 4.2467e-12, 9.0624e-12, 6.2485e-11, 2.1398e-12,\n 4.1960e-13, 2.7433e-12, 2.2177e-14, 1.1577e-12, 5.5717e-12, 1.1307e-10,\n 7.8786e-12, 6.6648e-13, 1.3349e-16, 3.2397e-12, 2.9583e-12, 6.0295e-11,\n 5.0269e-13, 2.0251e-12, 9.7442e-13, 4.3214e-12, 3.4468e-12, 3.2460e-12,\n 4.7888e-13, 1.6356e-11, 1.9750e-11, 9.8523e-12, 2.1741e-11, 2.8953e-12,\n 1.4936e-12, 4.0735e-15, 3.4282e-12, 1.0889e-12, 3.7526e-13, 1.3706e-12,\n 2.2512e-12, 1.4358e-14, 2.4454e-11, 6.2712e-12, 1.0598e-12, 3.7810e-15,\n 3.6536e-13, 2.5772e-12, 1.9775e-11, 3.7701e-11, 1.5567e-11, 1.2174e-12,\n 1.6158e-11, 7.1007e-12, 4.7434e-12, 2.1109e-12, 1.9639e-12, 8.5630e-14,\n 4.2130e-13, 3.0330e-11, 1.7723e-11, 2.0543e-11, 5.0156e-12, 1.1215e-13,\n 5.1175e-11, 2.1793e-12, 4.6328e-12, 6.0478e-13, 4.8523e-12, 1.8456e-12,\n 2.5458e-12, 1.6603e-10, 1.1213e-13, 2.2022e-12, 2.1938e-12, 6.3428e-11,\n 3.5417e-11, 3.8519e-12, 8.1148e-12, 2.1524e-12, 9.0250e-13, 1.1855e-11,\n 6.0894e-12, 7.9426e-12, 8.7641e-13, 3.4487e-11, 1.3871e-13, 4.7587e-13,\n 1.5612e-10, 4.7644e-12, 1.4011e-13, 1.5556e-12, 1.2188e-11, 2.0245e-12,\n 6.6321e-11, 1.4456e-10, 2.0891e-12, 6.9279e-12, 1.0518e-12, 1.9212e-13,\n 8.5141e-13, 3.4865e-12, 4.7304e-12, 8.4764e-12, 6.3999e-13, 3.4260e-11,\n 1.3430e-13, 2.7672e-13, 1.6942e-11, 7.0065e-12, 6.5180e-12, 3.0545e-11,\n 3.6096e-13, 1.5320e-11, 1.6713e-13, 9.2611e-13, 5.6083e-14, 1.7613e-11,\n 3.8010e-14, 1.8334e-12, 2.3534e-11, 1.2222e-11, 4.5365e-13, 1.0047e-11,\n 1.6078e-13, 8.0152e-13, 9.5758e-13, 3.3375e-11, 3.3462e-11, 1.4176e-11,\n 8.3792e-12, 2.2743e-12, 2.7943e-11, 4.4034e-11, 1.7528e-13, 2.6871e-15,\n 6.5686e-14, 1.4017e-11, 1.2706e-12, 2.8380e-13, 1.8892e-12, 2.6017e-11,\n 2.0613e-12, 5.1715e-14, 8.7449e-13, 1.3215e-12, 1.9529e-11, 3.2710e-13,\n 1.5227e-11, 5.9264e-11, 7.0989e-12, 3.8427e-12, 7.0962e-13, 7.6742e-12,\n 1.7045e-12, 9.5399e-12, 2.1977e-12, 5.8818e-13, 6.8430e-11, 2.8363e-13,\n 4.4418e-12, 1.0469e-14, 7.4854e-12, 2.4871e-11, 9.7962e-14, 8.2577e-13,\n 2.2010e-11, 1.0252e-12, 9.9002e-12, 7.1685e-11, 5.7302e-11, 7.4050e-12,\n 9.8317e-14, 2.9014e-12, 1.7102e-11, 1.8916e-14, 4.3630e-12, 7.4485e-11,\n 3.2052e-12, 4.3423e-12, 1.1139e-11, 2.3035e-12, 8.9871e-13, 5.7505e-12,\n 2.2569e-12, 4.0470e-12, 1.9029e-12, 1.1356e-11, 2.7720e-12, 1.8108e-12,\n 6.5694e-12, 1.5012e-12, 1.7322e-12, 2.4776e-12, 1.4612e-11, 1.5713e-13,\n 5.7337e-11, 1.1523e-12, 1.8873e-12, 7.5873e-12, 3.3990e-12, 2.7042e-13,\n 8.0982e-12, 7.8071e-11, 3.0536e-11, 9.5873e-11, 1.5020e-12, 7.9711e-12,\n 4.7331e-12, 2.2150e-11, 1.7520e-13, 4.7577e-14], device='cuda:0')" + "exp_avg_sq": "tensor([2.2776e-14, 2.6122e-17, 2.5348e-11, 8.6592e-14, 1.3483e-11, 4.2438e-12,\n 7.3959e-13, 1.0799e-12, 2.8007e-11, 1.1078e-13, 2.2138e-11, 1.2897e-12,\n 4.0355e-14, 1.7392e-13, 3.1581e-15, 2.5567e-13, 2.2360e-12, 3.1377e-12,\n 2.6751e-13, 1.9334e-13, 1.7563e-12, 1.2247e-13, 9.9486e-13, 7.6931e-13,\n 7.0803e-13, 4.6900e-12, 1.0983e-12, 1.1160e-12, 2.7151e-12, 2.7448e-11,\n 4.2885e-12, 3.7740e-12, 4.7332e-11, 5.9783e-14, 5.8378e-13, 2.8441e-13,\n 1.3270e-12, 1.3955e-13, 1.0674e-11, 1.5625e-13, 1.1756e-11, 2.4230e-13,\n 6.5248e-13, 5.2458e-12, 6.5748e-15, 2.6345e-13, 5.8230e-13, 5.2996e-13,\n 2.0601e-13, 6.3303e-12, 1.2135e-12, 2.5896e-12, 1.7856e-11, 6.1148e-13,\n 1.1990e-13, 7.8393e-13, 6.3374e-15, 3.3083e-13, 1.5922e-12, 3.2309e-11,\n 2.2514e-12, 1.9045e-13, 3.8147e-17, 9.2577e-13, 8.4535e-13, 1.7230e-11,\n 1.4365e-13, 5.7869e-13, 2.7845e-13, 1.2349e-12, 9.8494e-13, 9.2758e-13,\n 1.3684e-13, 4.6739e-12, 5.6438e-12, 2.8154e-12, 6.2126e-12, 8.2735e-13,\n 4.2680e-13, 1.1640e-15, 9.7964e-13, 3.1117e-13, 1.0723e-13, 3.9165e-13,\n 6.4330e-13, 4.1030e-15, 6.9880e-12, 1.7920e-12, 3.0285e-13, 1.0804e-15,\n 1.0441e-13, 7.3645e-13, 5.6510e-12, 1.0773e-11, 4.4483e-12, 3.4787e-13,\n 4.6173e-12, 2.0291e-12, 1.3555e-12, 6.0319e-13, 5.6119e-13, 2.4469e-14,\n 1.2039e-13, 8.6671e-12, 5.0646e-12, 5.8702e-12, 1.4332e-12, 3.2048e-14,\n 1.4624e-11, 6.2275e-13, 1.3239e-12, 1.7282e-13, 1.3866e-12, 5.2740e-13,\n 7.2749e-13, 4.7444e-11, 3.2041e-14, 6.2930e-13, 6.2689e-13, 1.8125e-11,\n 1.0121e-11, 1.1007e-12, 2.3189e-12, 6.1506e-13, 2.5790e-13, 3.3877e-12,\n 1.7401e-12, 2.2696e-12, 2.5044e-13, 9.8549e-12, 3.9639e-14, 1.3598e-13,\n 4.4613e-11, 1.3615e-12, 4.0038e-14, 4.4453e-13, 3.4828e-12, 5.7852e-13,\n 1.8952e-11, 4.1310e-11, 5.9698e-13, 1.9797e-12, 3.0056e-13, 5.4900e-14,\n 2.4330e-13, 9.9630e-13, 1.3517e-12, 2.4222e-12, 1.8288e-13, 9.7902e-12,\n 3.8376e-14, 7.9075e-14, 4.8414e-12, 2.0021e-12, 1.8626e-12, 8.7284e-12,\n 1.0315e-13, 4.3778e-12, 4.7760e-14, 2.6464e-13, 1.6026e-14, 5.0331e-12,\n 1.0862e-14, 5.2391e-13, 6.7251e-12, 3.4925e-12, 1.2963e-13, 2.8710e-12,\n 4.5945e-14, 2.2904e-13, 2.7364e-13, 9.5371e-12, 9.5620e-12, 4.0510e-12,\n 2.3944e-12, 6.4990e-13, 7.9849e-12, 1.2583e-11, 5.0088e-14, 7.6785e-16,\n 1.8770e-14, 4.0054e-12, 3.6308e-13, 8.1098e-14, 5.3984e-13, 7.4347e-12,\n 5.8903e-13, 1.4778e-14, 2.4989e-13, 3.7763e-13, 5.5806e-12, 9.3472e-14,\n 4.3512e-12, 1.6935e-11, 2.0286e-12, 1.0981e-12, 2.0278e-13, 2.1930e-12,\n 4.8707e-13, 2.7261e-12, 6.2802e-13, 1.6808e-13, 1.9554e-11, 8.1049e-14,\n 1.2693e-12, 2.9915e-15, 2.1390e-12, 7.1069e-12, 2.7994e-14, 2.3597e-13,\n 6.2894e-12, 2.9295e-13, 2.8290e-12, 2.0485e-11, 1.6374e-11, 2.1160e-12,\n 2.8095e-14, 8.2910e-13, 4.8871e-12, 5.4054e-15, 1.2468e-12, 2.1285e-11,\n 9.1590e-13, 1.2408e-12, 3.1831e-12, 6.5823e-13, 2.5681e-13, 1.6433e-12,\n 6.4492e-13, 1.1565e-12, 5.4376e-13, 3.2450e-12, 7.9213e-13, 5.1746e-13,\n 1.8773e-12, 4.2897e-13, 4.9498e-13, 7.0800e-13, 4.1755e-12, 4.4901e-14,\n 1.6384e-11, 3.2928e-13, 5.3931e-13, 2.1681e-12, 9.7129e-13, 7.7275e-14,\n 2.3141e-12, 2.2309e-11, 8.7258e-12, 2.7396e-11, 4.2921e-13, 2.2778e-12,\n 1.3525e-12, 6.3295e-12, 5.0064e-14, 1.3595e-14], device='cuda:0')" }, "39": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([1.6247e-13, 6.2031e-15, 1.2697e-10, 2.7348e-13, 5.7894e-11, 1.9428e-11,\n 4.6037e-12, 4.9469e-12, 1.1874e-10, 8.2611e-13, 6.2295e-11, 7.5230e-12,\n 2.2212e-13, 1.4894e-12, 4.4692e-14, 1.1311e-12, 1.0582e-11, 1.1909e-11,\n 8.3862e-13, 1.0809e-12, 1.0865e-11, 5.9651e-13, 7.6405e-12, 4.7399e-12,\n 3.9390e-12, 3.5885e-11, 7.5869e-12, 5.8530e-12, 6.3753e-12, 9.1851e-11,\n 3.1557e-11, 2.0942e-11, 2.3561e-10, 3.6821e-13, 3.5264e-12, 1.0358e-12,\n 9.4137e-12, 8.8770e-13, 3.9976e-11, 1.3051e-12, 3.7652e-11, 9.3495e-13,\n 2.5902e-12, 4.0672e-11, 1.6321e-14, 1.8929e-12, 3.8018e-12, 4.0204e-12,\n 7.8766e-13, 2.5730e-11, 9.5770e-12, 7.9764e-12, 6.0478e-11, 4.2259e-12,\n 8.7117e-13, 5.6561e-12, 5.4368e-14, 2.4078e-12, 1.0680e-11, 1.0578e-10,\n 1.2747e-11, 1.2617e-12, 1.2465e-14, 4.7263e-12, 5.1025e-12, 5.5920e-11,\n 6.8172e-13, 4.1648e-12, 1.5258e-12, 9.6215e-12, 6.2051e-12, 5.4793e-12,\n 1.4389e-12, 2.3395e-11, 3.2969e-11, 1.7767e-11, 2.2130e-11, 5.6316e-12,\n 1.9406e-12, 3.7025e-14, 5.7369e-12, 1.9835e-12, 7.6573e-13, 3.4124e-12,\n 3.8937e-12, 6.3891e-14, 4.3504e-11, 7.7361e-12, 2.1372e-12, 4.0352e-15,\n 4.7077e-13, 4.3164e-12, 2.2542e-11, 3.5727e-11, 3.2180e-11, 2.2522e-12,\n 3.0717e-11, 1.4153e-11, 8.6717e-12, 2.3922e-12, 4.2334e-12, 2.0625e-13,\n 8.6434e-13, 3.9743e-11, 1.7889e-11, 2.5054e-11, 1.0350e-11, 1.9718e-13,\n 3.5427e-11, 3.5804e-12, 6.2758e-12, 1.0801e-12, 7.7568e-12, 3.4230e-12,\n 3.7203e-12, 1.7368e-10, 2.5120e-13, 5.8762e-12, 1.9473e-12, 5.7286e-11,\n 5.4959e-11, 7.9737e-12, 1.8000e-11, 2.2323e-12, 1.0703e-12, 1.6426e-11,\n 7.4623e-12, 9.5658e-12, 1.5736e-12, 2.9771e-11, 1.8784e-13, 8.0659e-13,\n 1.5430e-10, 1.3904e-11, 3.0663e-13, 3.5631e-12, 1.0593e-11, 3.4054e-12,\n 6.1111e-11, 1.1934e-10, 3.4170e-12, 9.8687e-12, 8.2126e-13, 2.3470e-13,\n 1.6033e-12, 6.5393e-12, 6.1147e-12, 1.2605e-11, 8.1294e-13, 2.7744e-11,\n 2.8873e-13, 5.8635e-13, 2.0707e-11, 8.7667e-12, 8.5469e-12, 4.7211e-11,\n 7.7112e-13, 3.5071e-11, 3.6136e-13, 1.7112e-12, 2.8352e-13, 3.0442e-11,\n 1.3817e-13, 2.4282e-12, 3.8697e-11, 2.0192e-11, 1.1284e-12, 1.3100e-11,\n 2.1970e-13, 8.2566e-13, 2.5768e-12, 1.7505e-11, 3.4713e-11, 1.4803e-11,\n 7.7263e-12, 3.1823e-12, 4.0023e-11, 4.8944e-11, 4.9264e-13, 9.5902e-15,\n 1.0875e-13, 1.3781e-11, 1.4290e-12, 5.5292e-13, 1.6273e-12, 5.4572e-11,\n 3.4432e-12, 1.0774e-13, 1.5787e-12, 2.2260e-12, 2.3734e-11, 4.3554e-13,\n 2.0820e-11, 4.5715e-11, 1.3124e-11, 4.0293e-12, 1.2340e-12, 1.0899e-11,\n 3.2635e-12, 1.4484e-11, 4.0354e-12, 6.5851e-13, 1.1324e-10, 4.7519e-13,\n 7.8001e-12, 1.3719e-14, 9.3705e-12, 2.8435e-11, 3.3619e-13, 2.4197e-12,\n 3.5289e-11, 1.8121e-12, 1.6451e-11, 1.0148e-10, 9.5363e-11, 1.0583e-11,\n 2.4643e-13, 4.8550e-12, 2.7828e-11, 7.1413e-14, 5.9944e-12, 1.1865e-10,\n 3.9309e-12, 6.6650e-12, 2.2877e-11, 2.4278e-12, 1.3540e-12, 9.9616e-12,\n 6.0236e-12, 5.1833e-12, 3.2359e-12, 1.1182e-11, 4.3313e-12, 2.9331e-12,\n 1.2286e-11, 2.1692e-12, 3.5090e-12, 5.1169e-12, 1.0535e-11, 3.4787e-13,\n 5.5445e-11, 2.1555e-12, 1.4558e-12, 2.0433e-11, 6.6568e-12, 3.6057e-13,\n 1.1403e-11, 6.3919e-11, 5.4355e-11, 7.7766e-11, 2.8420e-12, 9.9083e-12,\n 6.9704e-12, 3.4935e-11, 4.2927e-13, 4.6440e-14], device='cuda:0')" + "exp_avg_sq": "tensor([4.6426e-14, 1.7726e-15, 3.6282e-11, 7.8150e-14, 1.6544e-11, 5.5517e-12,\n 1.3155e-12, 1.4136e-12, 3.3932e-11, 2.3607e-13, 1.7801e-11, 2.1498e-12,\n 6.3473e-14, 4.2560e-13, 1.2771e-14, 3.2322e-13, 3.0238e-12, 3.4030e-12,\n 2.3964e-13, 3.0887e-13, 3.1049e-12, 1.7046e-13, 2.1833e-12, 1.3545e-12,\n 1.1256e-12, 1.0254e-11, 2.1680e-12, 1.6725e-12, 1.8218e-12, 2.6247e-11,\n 9.0178e-12, 5.9843e-12, 6.7328e-11, 1.0522e-13, 1.0077e-12, 2.9600e-13,\n 2.6901e-12, 2.5367e-13, 1.1423e-11, 3.7295e-13, 1.0759e-11, 2.6717e-13,\n 7.4018e-13, 1.1622e-11, 4.6638e-15, 5.4090e-13, 1.0864e-12, 1.1489e-12,\n 2.2508e-13, 7.3525e-12, 2.7367e-12, 2.2793e-12, 1.7282e-11, 1.2076e-12,\n 2.4894e-13, 1.6163e-12, 1.5536e-14, 6.8804e-13, 3.0519e-12, 3.0228e-11,\n 3.6426e-12, 3.6053e-13, 3.5620e-15, 1.3506e-12, 1.4581e-12, 1.5980e-11,\n 1.9481e-13, 1.1901e-12, 4.3600e-13, 2.7494e-12, 1.7732e-12, 1.5658e-12,\n 4.1118e-13, 6.6854e-12, 9.4212e-12, 5.0771e-12, 6.3238e-12, 1.6093e-12,\n 5.5454e-13, 1.0580e-14, 1.6394e-12, 5.6680e-13, 2.1881e-13, 9.7512e-13,\n 1.1126e-12, 1.8257e-14, 1.2432e-11, 2.2107e-12, 6.1071e-13, 1.1531e-15,\n 1.3453e-13, 1.2334e-12, 6.4415e-12, 1.0209e-11, 9.1958e-12, 6.4360e-13,\n 8.7775e-12, 4.0444e-12, 2.4780e-12, 6.8358e-13, 1.2097e-12, 5.8938e-14,\n 2.4699e-13, 1.1357e-11, 5.1119e-12, 7.1594e-12, 2.9577e-12, 5.6345e-14,\n 1.0124e-11, 1.0231e-12, 1.7934e-12, 3.0864e-13, 2.2166e-12, 9.7815e-13,\n 1.0631e-12, 4.9631e-11, 7.1781e-14, 1.6792e-12, 5.5645e-13, 1.6370e-11,\n 1.5705e-11, 2.2785e-12, 5.1437e-12, 6.3790e-13, 3.0583e-13, 4.6939e-12,\n 2.1324e-12, 2.7335e-12, 4.4966e-13, 8.5073e-12, 5.3678e-14, 2.3049e-13,\n 4.4093e-11, 3.9732e-12, 8.7623e-14, 1.0182e-12, 3.0269e-12, 9.7313e-13,\n 1.7463e-11, 3.4102e-11, 9.7643e-13, 2.8201e-12, 2.3468e-13, 6.7068e-14,\n 4.5814e-13, 1.8687e-12, 1.7473e-12, 3.6021e-12, 2.3230e-13, 7.9281e-12,\n 8.2507e-14, 1.6755e-13, 5.9171e-12, 2.5052e-12, 2.4423e-12, 1.3491e-11,\n 2.2035e-13, 1.0022e-11, 1.0326e-13, 4.8900e-13, 8.1019e-14, 8.6990e-12,\n 3.9483e-14, 6.9388e-13, 1.1058e-11, 5.7701e-12, 3.2244e-13, 3.7434e-12,\n 6.2781e-14, 2.3594e-13, 7.3635e-13, 5.0022e-12, 9.9194e-12, 4.2300e-12,\n 2.2078e-12, 9.0936e-13, 1.1437e-11, 1.3986e-11, 1.4078e-13, 2.7405e-15,\n 3.1075e-14, 3.9382e-12, 4.0834e-13, 1.5800e-13, 4.6503e-13, 1.5594e-11,\n 9.8394e-13, 3.0787e-14, 4.5112e-13, 6.3609e-13, 6.7823e-12, 1.2446e-13,\n 5.9496e-12, 1.3064e-11, 3.7503e-12, 1.1514e-12, 3.5261e-13, 3.1144e-12,\n 9.3256e-13, 4.1391e-12, 1.1532e-12, 1.8818e-13, 3.2360e-11, 1.3579e-13,\n 2.2290e-12, 3.9203e-15, 2.6777e-12, 8.1255e-12, 9.6069e-14, 6.9145e-13,\n 1.0084e-11, 5.1783e-13, 4.7011e-12, 2.8998e-11, 2.7251e-11, 3.0242e-12,\n 7.0419e-14, 1.3874e-12, 7.9520e-12, 2.0407e-14, 1.7130e-12, 3.3906e-11,\n 1.1233e-12, 1.9046e-12, 6.5374e-12, 6.9376e-13, 3.8692e-13, 2.8466e-12,\n 1.7213e-12, 1.4812e-12, 9.2469e-13, 3.1953e-12, 1.2377e-12, 8.3817e-13,\n 3.5108e-12, 6.1986e-13, 1.0027e-12, 1.4622e-12, 3.0104e-12, 9.9407e-14,\n 1.5844e-11, 6.1596e-13, 4.1600e-13, 5.8388e-12, 1.9022e-12, 1.0304e-13,\n 3.2586e-12, 1.8265e-11, 1.5532e-11, 2.2222e-11, 8.1213e-13, 2.8314e-12,\n 1.9918e-12, 9.9828e-12, 1.2267e-13, 1.3271e-14], device='cuda:0')" }, "40": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[-5.6052e-45, 5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [-5.6052e-45, -5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [-5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[3.5706e-18, 1.4614e-16, 1.0016e-14, ..., 4.1279e-15, 0.0000e+00,\n 4.1635e-14],\n [9.5113e-15, 1.1034e-15, 9.9037e-14, ..., 6.0940e-14, 0.0000e+00,\n 2.6467e-14],\n [3.0439e-12, 4.6959e-14, 4.0767e-11, ..., 6.3829e-12, 0.0000e+00,\n 2.9291e-11],\n ...,\n [7.0622e-13, 1.9660e-13, 1.2457e-11, ..., 7.0788e-12, 0.0000e+00,\n 6.3716e-12],\n [4.0711e-15, 8.9847e-15, 1.5071e-13, ..., 1.3067e-14, 0.0000e+00,\n 7.2166e-14],\n [7.5040e-15, 9.8285e-14, 6.6878e-13, ..., 7.8228e-14, 0.0000e+00,\n 6.5772e-13]], device='cuda:0')" + "exp_avg_sq": "tensor([[1.0203e-18, 4.1760e-17, 2.8622e-15, ..., 1.1796e-15, 0.0000e+00,\n 1.1898e-14],\n [2.7179e-15, 3.1532e-16, 2.8301e-14, ..., 1.7414e-14, 0.0000e+00,\n 7.5632e-15],\n [8.6981e-13, 1.3419e-14, 1.1650e-11, ..., 1.8240e-12, 0.0000e+00,\n 8.3701e-12],\n ...,\n [2.0181e-13, 5.6179e-14, 3.5596e-12, ..., 2.0228e-12, 0.0000e+00,\n 1.8207e-12],\n [1.1633e-15, 2.5675e-15, 4.3066e-14, ..., 3.7341e-15, 0.0000e+00,\n 2.0622e-14],\n [2.1443e-15, 2.8086e-14, 1.9111e-13, ..., 2.2354e-14, 0.0000e+00,\n 1.8795e-13]], device='cuda:0')" }, "41": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([3.3835e-13, 7.2243e-12, 1.3722e-08, 7.9997e-12, 3.0142e-09, 4.3618e-09,\n 2.2646e-09, 2.9187e-10, 4.6284e-08, 1.3038e-09, 1.0715e-08, 9.6087e-09,\n 2.8775e-11, 8.9983e-10, 1.8683e-11, 3.7350e-11, 4.3476e-09, 2.4361e-09,\n 5.2058e-10, 1.3215e-10, 7.6710e-10, 7.6424e-11, 4.8852e-09, 2.3178e-09,\n 6.1803e-09, 8.3029e-09, 2.6610e-09, 3.8604e-09, 1.5112e-09, 2.1558e-08,\n 2.2770e-08, 7.8243e-10, 9.1130e-08, 8.0198e-10, 9.6481e-10, 8.7110e-11,\n 3.3302e-09, 7.0704e-11, 2.1733e-09, 7.6949e-10, 7.4301e-09, 7.0436e-11,\n 1.7769e-11, 4.5184e-08, 1.1286e-11, 2.0330e-09, 4.7965e-09, 3.8565e-09,\n 4.3637e-11, 2.5268e-09, 3.7897e-09, 4.9238e-10, 1.6053e-08, 1.0532e-09,\n 5.2142e-10, 1.0781e-08, 3.4522e-10, 9.8438e-10, 1.9934e-08, 1.9615e-08,\n 4.1991e-09, 5.5180e-09, 1.3263e-10, 2.6657e-09, 3.2738e-09, 1.8291e-08,\n 1.1067e-10, 4.8104e-09, 1.0136e-09, 2.3180e-09, 8.4550e-10, 9.0208e-10,\n 2.4175e-09, 2.5930e-08, 7.8374e-09, 2.2920e-08, 4.8593e-09, 6.9132e-10,\n 7.2894e-10, 3.3316e-11, 2.2499e-09, 3.2246e-10, 2.0753e-10, 5.8701e-09,\n 1.7047e-09, 1.7269e-10, 3.4212e-08, 3.4423e-09, 9.8817e-10, 8.1816e-12,\n 6.1984e-11, 2.1920e-09, 3.8552e-09, 1.0613e-09, 3.6857e-08, 1.4084e-10,\n 3.4405e-08, 2.3799e-08, 4.1368e-09, 2.7359e-10, 9.2966e-10, 5.5858e-10,\n 6.1777e-11, 1.7443e-08, 2.0307e-10, 2.8901e-09, 2.2910e-09, 3.8629e-10,\n 2.5195e-09, 1.1412e-08, 2.9406e-09, 1.9768e-10, 1.2261e-08, 1.2533e-09,\n 9.6230e-10, 3.6824e-09, 3.3300e-10, 2.8474e-09, 3.5378e-10, 9.6683e-09,\n 2.6582e-08, 3.3735e-09, 3.0178e-09, 3.9328e-11, 6.8490e-10, 3.8755e-08,\n 2.8445e-10, 2.1269e-09, 8.0531e-10, 2.5489e-09, 2.8260e-11, 8.0991e-10,\n 3.7994e-08, 1.7236e-09, 7.6030e-11, 3.2178e-09, 2.0448e-09, 3.1537e-10,\n 1.0529e-08, 1.8702e-08, 3.3728e-10, 9.3636e-10, 1.3631e-08, 2.2796e-11,\n 4.9270e-10, 2.9789e-09, 9.5323e-10, 1.1217e-08, 2.0318e-10, 6.5663e-09,\n 6.0961e-10, 3.7041e-11, 9.9037e-09, 2.6570e-09, 6.8212e-10, 9.6470e-09,\n 3.4870e-09, 1.3253e-08, 1.9312e-11, 1.7053e-09, 7.5054e-11, 9.1497e-09,\n 8.3724e-12, 1.4651e-08, 1.9437e-09, 9.7388e-09, 9.9612e-10, 3.4513e-09,\n 7.0512e-11, 2.4928e-12, 8.3391e-10, 3.1616e-09, 2.8635e-09, 3.5596e-09,\n 1.3242e-09, 1.1196e-09, 1.5780e-08, 3.5556e-09, 2.9469e-11, 5.2370e-12,\n 3.3983e-10, 3.6113e-09, 4.2674e-11, 3.9182e-10, 2.0199e-10, 4.4425e-08,\n 4.2148e-09, 6.0427e-11, 8.3467e-10, 9.9403e-10, 1.3943e-08, 3.9887e-11,\n 5.9162e-09, 1.5078e-08, 9.8644e-09, 1.1682e-10, 1.3915e-10, 4.3525e-10,\n 4.3243e-09, 1.2200e-08, 6.2447e-09, 7.3650e-11, 2.7204e-08, 3.3920e-12,\n 1.8069e-09, 1.2202e-11, 1.0444e-09, 8.8971e-09, 2.4941e-10, 2.8105e-10,\n 1.6745e-08, 5.2622e-10, 6.5089e-09, 3.5855e-08, 1.2062e-07, 8.9336e-10,\n 1.4769e-10, 5.5306e-09, 2.7387e-09, 3.6584e-12, 2.1476e-09, 4.9546e-08,\n 6.9506e-10, 4.5427e-09, 4.0996e-09, 5.1299e-10, 2.8697e-10, 6.8062e-09,\n 8.4046e-09, 4.2396e-10, 8.1681e-09, 1.2358e-09, 3.9323e-09, 5.2638e-09,\n 4.6854e-09, 2.3825e-09, 1.3652e-09, 2.6547e-09, 8.0153e-10, 1.5439e-10,\n 1.3811e-08, 2.8913e-09, 8.9446e-11, 1.1491e-08, 1.1797e-09, 2.3663e-11,\n 1.0791e-08, 1.3129e-08, 3.3048e-08, 1.0195e-08, 3.1693e-10, 7.8598e-09,\n 7.1589e-10, 1.0500e-08, 1.2463e-10, 8.9954e-10], device='cuda:0')" + "exp_avg_sq": "tensor([9.6686e-14, 2.0644e-12, 3.9213e-09, 2.2860e-12, 8.6134e-10, 1.2464e-09,\n 6.4714e-10, 8.3403e-11, 1.3226e-08, 3.7258e-10, 3.0619e-09, 2.7458e-09,\n 8.2228e-12, 2.5714e-10, 5.3387e-12, 1.0673e-11, 1.2424e-09, 6.9613e-10,\n 1.4876e-10, 3.7763e-11, 2.1920e-10, 2.1839e-11, 1.3960e-09, 6.6234e-10,\n 1.7661e-09, 2.3726e-09, 7.6039e-10, 1.1031e-09, 4.3182e-10, 6.1604e-09,\n 6.5066e-09, 2.2359e-10, 2.6041e-08, 2.2917e-10, 2.7570e-10, 2.4892e-11,\n 9.5163e-10, 2.0204e-11, 6.2104e-10, 2.1989e-10, 2.1232e-09, 2.0128e-11,\n 5.0777e-12, 1.2912e-08, 3.2251e-12, 5.8095e-10, 1.3706e-09, 1.1020e-09,\n 1.2470e-11, 7.2206e-10, 1.0829e-09, 1.4070e-10, 4.5873e-09, 3.0097e-10,\n 1.4900e-10, 3.0809e-09, 9.8650e-11, 2.8129e-10, 5.6962e-09, 5.6050e-09,\n 1.1999e-09, 1.5768e-09, 3.7900e-11, 7.6173e-10, 9.3552e-10, 5.2268e-09,\n 3.1624e-11, 1.3746e-09, 2.8964e-10, 6.6238e-10, 2.4161e-10, 2.5778e-10,\n 6.9083e-10, 7.4096e-09, 2.2396e-09, 6.5497e-09, 1.3886e-09, 1.9755e-10,\n 2.0830e-10, 9.5202e-12, 6.4292e-10, 9.2145e-11, 5.9303e-11, 1.6774e-09,\n 4.8712e-10, 4.9347e-11, 9.7764e-09, 9.8365e-10, 2.8238e-10, 2.3380e-12,\n 1.7713e-11, 6.2639e-10, 1.1017e-09, 3.0328e-10, 1.0532e-08, 4.0247e-11,\n 9.8316e-09, 6.8007e-09, 1.1821e-09, 7.8181e-11, 2.6566e-10, 1.5962e-10,\n 1.7653e-11, 4.9844e-09, 5.8029e-11, 8.2587e-10, 6.5467e-10, 1.1039e-10,\n 7.1996e-10, 3.2611e-09, 8.4030e-10, 5.6488e-11, 3.5036e-09, 3.5815e-10,\n 2.7498e-10, 1.0523e-09, 9.5157e-11, 8.1365e-10, 1.0110e-10, 2.7628e-09,\n 7.5959e-09, 9.6401e-10, 8.6236e-10, 1.1238e-11, 1.9572e-10, 1.1075e-08,\n 8.1285e-11, 6.0776e-10, 2.3012e-10, 7.2836e-10, 8.0755e-12, 2.3144e-10,\n 1.0857e-08, 4.9252e-10, 2.1726e-11, 9.1951e-10, 5.8431e-10, 9.0121e-11,\n 3.0086e-09, 5.3442e-09, 9.6381e-11, 2.6757e-10, 3.8952e-09, 6.5142e-12,\n 1.4079e-10, 8.5123e-10, 2.7239e-10, 3.2054e-09, 5.8060e-11, 1.8764e-09,\n 1.7420e-10, 1.0585e-11, 2.8301e-09, 7.5925e-10, 1.9492e-10, 2.7567e-09,\n 9.9643e-10, 3.7872e-09, 5.5186e-12, 4.8732e-10, 2.1447e-11, 2.6146e-09,\n 2.3925e-12, 4.1867e-09, 5.5544e-10, 2.7829e-09, 2.8465e-10, 9.8623e-10,\n 2.0149e-11, 7.1232e-13, 2.3830e-10, 9.0346e-10, 8.1826e-10, 1.0172e-09,\n 3.7841e-10, 3.1992e-10, 4.5092e-09, 1.0160e-09, 8.4210e-12, 1.4965e-12,\n 9.7111e-11, 1.0320e-09, 1.2194e-11, 1.1196e-10, 5.7720e-11, 1.2695e-08,\n 1.2044e-09, 1.7268e-11, 2.3851e-10, 2.8405e-10, 3.9842e-09, 1.1398e-11,\n 1.6906e-09, 4.3087e-09, 2.8188e-09, 3.3381e-11, 3.9763e-11, 1.2438e-10,\n 1.2357e-09, 3.4862e-09, 1.7845e-09, 2.1046e-11, 7.7737e-09, 9.6928e-13,\n 5.1632e-10, 3.4867e-12, 2.9846e-10, 2.5424e-09, 7.1271e-11, 8.0312e-11,\n 4.7849e-09, 1.5037e-10, 1.8600e-09, 1.0246e-08, 3.4469e-08, 2.5528e-10,\n 4.2203e-11, 1.5804e-09, 7.8262e-10, 1.0454e-12, 6.1368e-10, 1.4158e-08,\n 1.9862e-10, 1.2981e-09, 1.1715e-09, 1.4659e-10, 8.2005e-11, 1.9449e-09,\n 2.4017e-09, 1.2115e-10, 2.3341e-09, 3.5315e-10, 1.1237e-09, 1.5042e-09,\n 1.3389e-09, 6.8083e-10, 3.9012e-10, 7.5860e-10, 2.2904e-10, 4.4118e-11,\n 3.9466e-09, 8.2621e-10, 2.5560e-11, 3.2836e-09, 3.3711e-10, 6.7619e-12,\n 3.0837e-09, 3.7517e-09, 9.4437e-09, 2.9134e-09, 9.0564e-11, 2.2460e-09,\n 2.0457e-10, 3.0004e-09, 3.5615e-11, 2.5705e-10], device='cuda:0')" }, "42": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([1.1144e-14, 1.5780e-14, 2.3904e-11, 6.7419e-14, 5.3374e-12, 7.3714e-12,\n 3.8141e-12, 4.7729e-13, 8.3453e-11, 2.8020e-12, 1.7114e-11, 1.5570e-11,\n 8.5791e-14, 1.2713e-12, 5.8064e-14, 9.8766e-14, 8.2010e-12, 1.0174e-11,\n 1.3602e-12, 4.4066e-13, 1.8171e-12, 2.0218e-13, 1.2712e-11, 4.7530e-12,\n 2.2136e-11, 1.2961e-11, 5.3757e-12, 1.1652e-11, 5.8513e-12, 3.5687e-11,\n 4.2077e-11, 1.9233e-12, 1.8372e-10, 8.9440e-13, 1.6030e-12, 1.3001e-13,\n 5.2682e-12, 3.7647e-13, 2.9749e-12, 1.6588e-12, 9.7710e-12, 8.9178e-13,\n 2.3599e-14, 1.5089e-10, 4.6241e-14, 3.5810e-12, 1.3189e-11, 1.5266e-11,\n 1.3295e-13, 4.0147e-12, 7.3643e-12, 6.1669e-13, 2.8043e-11, 1.6976e-12,\n 1.1155e-12, 1.6749e-11, 1.2393e-12, 1.3295e-12, 4.7233e-11, 2.6043e-11,\n 6.1707e-12, 1.5095e-11, 3.5608e-13, 4.2468e-12, 8.3914e-12, 4.0014e-11,\n 3.2634e-13, 7.0845e-12, 1.8767e-12, 2.5925e-12, 1.3799e-12, 1.9873e-12,\n 3.9385e-12, 6.2733e-11, 1.0854e-11, 4.9619e-11, 9.6626e-12, 1.2148e-12,\n 1.4368e-12, 3.5516e-14, 2.7548e-12, 4.8309e-13, 6.3459e-13, 2.3101e-11,\n 2.2156e-12, 9.3900e-13, 9.1007e-11, 1.2032e-11, 2.4887e-12, 4.9114e-14,\n 3.0020e-13, 3.5795e-12, 5.0175e-12, 1.6990e-12, 9.8227e-11, 3.3366e-13,\n 8.7830e-11, 6.0927e-11, 7.7274e-12, 1.0875e-12, 1.7983e-12, 1.1551e-12,\n 1.9002e-13, 3.2457e-11, 7.2612e-13, 4.0040e-12, 2.6859e-12, 9.9456e-13,\n 2.5324e-12, 2.6372e-11, 6.5774e-12, 2.6338e-13, 1.8775e-11, 1.5775e-12,\n 2.4182e-12, 6.1742e-12, 1.2237e-12, 3.1897e-12, 1.3208e-12, 1.3634e-11,\n 5.0544e-11, 1.0006e-11, 3.3348e-12, 9.0311e-14, 2.0913e-12, 7.8762e-11,\n 5.5483e-13, 5.8995e-12, 1.8475e-12, 4.4877e-12, 1.6183e-13, 1.0824e-12,\n 7.1972e-11, 2.7938e-12, 1.9521e-13, 6.5242e-12, 5.5148e-12, 6.9905e-13,\n 1.6890e-11, 2.9662e-11, 6.7764e-13, 1.4140e-12, 2.6057e-11, 1.5221e-13,\n 1.2812e-12, 5.8168e-12, 1.3351e-12, 3.6409e-11, 8.5493e-13, 2.1694e-11,\n 1.1959e-12, 2.2944e-14, 2.0341e-11, 5.0605e-12, 1.2748e-12, 1.0012e-11,\n 5.5102e-12, 1.6991e-11, 2.8451e-14, 3.6085e-12, 8.8608e-13, 1.9916e-11,\n 4.6902e-14, 4.7316e-11, 3.3909e-12, 1.6679e-11, 2.4141e-12, 7.4987e-12,\n 1.6169e-13, 7.2581e-15, 1.1657e-12, 7.8604e-12, 3.6049e-12, 4.9561e-12,\n 2.9625e-12, 2.4753e-12, 3.4780e-11, 6.6265e-12, 1.0419e-13, 1.2524e-14,\n 5.5686e-13, 9.5801e-12, 3.9779e-14, 7.5909e-13, 5.0792e-13, 7.6930e-11,\n 5.3510e-12, 3.2950e-13, 3.7250e-12, 2.6453e-12, 4.2818e-11, 2.9560e-13,\n 8.8627e-12, 3.2578e-11, 1.7285e-11, 3.5031e-13, 1.3469e-13, 8.3069e-13,\n 6.7884e-12, 2.8940e-11, 1.4643e-11, 4.1385e-13, 4.6435e-11, 7.8776e-14,\n 1.9941e-12, 2.3080e-14, 1.3317e-12, 1.8892e-11, 1.0221e-12, 4.3824e-13,\n 5.0826e-11, 5.4774e-13, 8.6526e-12, 7.3214e-11, 3.4870e-10, 1.3132e-12,\n 2.5958e-13, 9.8291e-12, 3.9841e-12, 3.3104e-15, 2.1298e-12, 8.8707e-11,\n 1.8258e-12, 1.4199e-11, 7.4123e-12, 8.3640e-13, 7.4883e-13, 1.4232e-11,\n 1.4536e-11, 6.5740e-13, 1.0378e-11, 2.5494e-12, 8.1695e-12, 1.0883e-11,\n 7.4965e-12, 4.3875e-12, 2.2249e-12, 5.2209e-12, 1.2320e-12, 3.0936e-13,\n 2.8448e-11, 7.2805e-12, 8.7219e-13, 1.5963e-11, 1.2195e-12, 1.5670e-13,\n 2.9385e-11, 3.1586e-11, 7.7952e-11, 1.6647e-11, 5.1266e-13, 1.6946e-11,\n 1.0805e-12, 2.3883e-11, 3.2270e-13, 3.8248e-12], device='cuda:0')" + "exp_avg_sq": "tensor([3.1845e-15, 4.5093e-15, 6.8307e-12, 1.9266e-14, 1.5252e-12, 2.1064e-12,\n 1.0899e-12, 1.3639e-13, 2.3847e-11, 8.0070e-13, 4.8903e-12, 4.4494e-12,\n 2.4515e-14, 3.6329e-13, 1.6592e-14, 2.8223e-14, 2.3435e-12, 2.9074e-12,\n 3.8869e-13, 1.2592e-13, 5.1926e-13, 5.7773e-14, 3.6326e-12, 1.3582e-12,\n 6.3254e-12, 3.7038e-12, 1.5361e-12, 3.3298e-12, 1.6721e-12, 1.0198e-11,\n 1.2024e-11, 5.4959e-13, 5.2500e-11, 2.5558e-13, 4.5808e-13, 3.7151e-14,\n 1.5054e-12, 1.0758e-13, 8.5011e-13, 4.7402e-13, 2.7921e-12, 2.5483e-13,\n 6.7435e-15, 4.3117e-11, 1.3214e-14, 1.0233e-12, 3.7690e-12, 4.3625e-12,\n 3.7991e-14, 1.1472e-12, 2.1044e-12, 1.7622e-13, 8.0134e-12, 4.8510e-13,\n 3.1876e-13, 4.7861e-12, 3.5415e-13, 3.7993e-13, 1.3497e-11, 7.4420e-12,\n 1.7633e-12, 4.3137e-12, 1.0175e-13, 1.2136e-12, 2.3979e-12, 1.1434e-11,\n 9.3253e-14, 2.0244e-12, 5.3627e-13, 7.4082e-13, 3.9432e-13, 5.6787e-13,\n 1.1255e-12, 1.7926e-11, 3.1016e-12, 1.4179e-11, 2.7612e-12, 3.4715e-13,\n 4.1058e-13, 1.0149e-14, 7.8722e-13, 1.3805e-13, 1.8134e-13, 6.6012e-12,\n 6.3311e-13, 2.6833e-13, 2.6006e-11, 3.4382e-12, 7.1117e-13, 1.4035e-14,\n 8.5786e-14, 1.0229e-12, 1.4338e-12, 4.8550e-13, 2.8069e-11, 9.5346e-14,\n 2.5098e-11, 1.7410e-11, 2.2082e-12, 3.1075e-13, 5.1388e-13, 3.3007e-13,\n 5.4301e-14, 9.2747e-12, 2.0750e-13, 1.1442e-12, 7.6751e-13, 2.8420e-13,\n 7.2366e-13, 7.5359e-12, 1.8795e-12, 7.5262e-14, 5.3650e-12, 4.5078e-13,\n 6.9102e-13, 1.7643e-12, 3.4969e-13, 9.1149e-13, 3.7744e-13, 3.8960e-12,\n 1.4443e-11, 2.8592e-12, 9.5294e-13, 2.5807e-14, 5.9761e-13, 2.2507e-11,\n 1.5855e-13, 1.6858e-12, 5.2794e-13, 1.2824e-12, 4.6244e-14, 3.0930e-13,\n 2.0566e-11, 7.9835e-13, 5.5784e-14, 1.8643e-12, 1.5759e-12, 1.9976e-13,\n 4.8265e-12, 8.4760e-12, 1.9364e-13, 4.0406e-13, 7.4461e-12, 4.3495e-14,\n 3.6611e-13, 1.6622e-12, 3.8152e-13, 1.0404e-11, 2.4430e-13, 6.1991e-12,\n 3.4174e-13, 6.5564e-15, 5.8126e-12, 1.4461e-12, 3.6429e-13, 2.8609e-12,\n 1.5746e-12, 4.8553e-12, 8.1301e-15, 1.0311e-12, 2.5321e-13, 5.6911e-12,\n 1.3403e-14, 1.3521e-11, 9.6898e-13, 4.7662e-12, 6.8985e-13, 2.1428e-12,\n 4.6203e-14, 2.0741e-15, 3.3311e-13, 2.2462e-12, 1.0301e-12, 1.4162e-12,\n 8.4655e-13, 7.0734e-13, 9.9386e-12, 1.8936e-12, 2.9773e-14, 3.5789e-15,\n 1.5913e-13, 2.7376e-12, 1.1367e-14, 2.1692e-13, 1.4514e-13, 2.1984e-11,\n 1.5291e-12, 9.4157e-14, 1.0645e-12, 7.5590e-13, 1.2236e-11, 8.4471e-14,\n 2.5326e-12, 9.3093e-12, 4.9393e-12, 1.0010e-13, 3.8489e-14, 2.3738e-13,\n 1.9398e-12, 8.2697e-12, 4.1843e-12, 1.1826e-13, 1.3269e-11, 2.2511e-14,\n 5.6984e-13, 6.5952e-15, 3.8054e-13, 5.3986e-12, 2.9206e-13, 1.2523e-13,\n 1.4524e-11, 1.5652e-13, 2.4726e-12, 2.0922e-11, 9.9643e-11, 3.7526e-13,\n 7.4177e-14, 2.8088e-12, 1.1385e-12, 9.4596e-16, 6.0862e-13, 2.5349e-11,\n 5.2173e-13, 4.0576e-12, 2.1181e-12, 2.3901e-13, 2.1398e-13, 4.0670e-12,\n 4.1538e-12, 1.8786e-13, 2.9655e-12, 7.2851e-13, 2.3345e-12, 3.1099e-12,\n 2.1422e-12, 1.2538e-12, 6.3578e-13, 1.4919e-12, 3.5205e-13, 8.8402e-14,\n 8.1291e-12, 2.0805e-12, 2.4924e-13, 4.5616e-12, 3.4848e-13, 4.4779e-14,\n 8.3971e-12, 9.0260e-12, 2.2275e-11, 4.7570e-12, 1.4650e-13, 4.8423e-12,\n 3.0875e-13, 6.8247e-12, 9.2215e-14, 1.0930e-12], device='cuda:0')" }, "43": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([3.8372e-14, 2.8582e-14, 3.9635e-11, 1.1168e-13, 8.2233e-12, 1.0812e-11,\n 6.4034e-12, 1.0453e-12, 1.2806e-10, 3.7886e-12, 3.0397e-11, 2.3420e-11,\n 8.5709e-14, 2.8103e-12, 8.6232e-14, 2.8357e-13, 1.3065e-11, 7.9260e-12,\n 1.2533e-12, 6.0188e-13, 2.6196e-12, 3.6392e-13, 1.2754e-11, 6.8879e-12,\n 1.8344e-11, 2.1396e-11, 7.8352e-12, 9.9471e-12, 5.6990e-12, 5.9798e-11,\n 5.6411e-11, 2.6150e-12, 2.5393e-10, 1.9486e-12, 3.2871e-12, 4.3734e-13,\n 8.5655e-12, 5.9951e-13, 6.5889e-12, 2.3052e-12, 1.7875e-11, 5.5844e-13,\n 1.0677e-13, 1.2666e-10, 5.0235e-14, 4.7998e-12, 1.2132e-11, 1.1846e-11,\n 2.4029e-13, 7.4572e-12, 1.1787e-11, 1.2492e-12, 4.1549e-11, 2.7304e-12,\n 1.7745e-12, 2.8722e-11, 1.3733e-12, 2.5085e-12, 5.1431e-11, 5.6446e-11,\n 1.0211e-11, 1.2845e-11, 6.1667e-13, 7.7211e-12, 9.4705e-12, 4.5440e-11,\n 4.0474e-13, 1.2844e-11, 3.0411e-12, 6.8206e-12, 2.5546e-12, 2.6655e-12,\n 5.0823e-12, 7.1356e-11, 2.2580e-11, 5.8325e-11, 1.4045e-11, 2.1991e-12,\n 2.4917e-12, 7.6391e-14, 6.2233e-12, 9.9813e-13, 8.9076e-13, 1.6906e-11,\n 4.1738e-12, 8.4012e-13, 9.4449e-11, 1.0843e-11, 3.4772e-12, 6.9717e-14,\n 3.1853e-13, 5.6534e-12, 1.0140e-11, 3.1930e-12, 9.2452e-11, 3.6375e-13,\n 8.7267e-11, 6.0409e-11, 9.3522e-12, 1.7436e-12, 2.8077e-12, 2.0288e-12,\n 1.8045e-13, 4.5200e-11, 6.0468e-13, 7.5727e-12, 6.4819e-12, 1.4859e-12,\n 6.4989e-12, 2.7996e-11, 8.6049e-12, 6.7582e-13, 3.2690e-11, 2.8174e-12,\n 3.1235e-12, 1.0728e-11, 1.3052e-12, 7.1921e-12, 1.7676e-12, 2.7734e-11,\n 7.2600e-11, 1.0576e-11, 8.7110e-12, 6.3178e-13, 2.5184e-12, 9.8882e-11,\n 7.7047e-13, 6.6190e-12, 2.4453e-12, 7.3561e-12, 2.4729e-13, 2.6464e-12,\n 9.7647e-11, 4.6531e-12, 3.3588e-13, 8.4442e-12, 6.5463e-12, 7.7911e-13,\n 2.6410e-11, 4.7160e-11, 1.2673e-12, 2.9850e-12, 3.3607e-11, 2.7859e-13,\n 1.7236e-12, 8.7213e-12, 2.2254e-12, 2.9150e-11, 7.7986e-13, 1.9063e-11,\n 2.0103e-12, 4.5006e-14, 2.5394e-11, 6.2953e-12, 2.3425e-12, 2.5065e-11,\n 8.4711e-12, 3.2081e-11, 1.0592e-13, 5.0280e-12, 1.2092e-12, 2.6745e-11,\n 9.3083e-14, 4.1189e-11, 5.5673e-12, 2.4838e-11, 2.4674e-12, 8.8197e-12,\n 2.6688e-13, 1.9946e-14, 2.1752e-12, 9.6616e-12, 8.0301e-12, 1.0548e-11,\n 3.7527e-12, 3.2831e-12, 4.4374e-11, 9.8977e-12, 4.1077e-13, 2.7107e-14,\n 1.2611e-12, 9.8190e-12, 1.6585e-13, 1.3749e-12, 8.2252e-13, 1.2230e-10,\n 1.0205e-11, 3.0457e-13, 3.0699e-12, 3.4634e-12, 3.4389e-11, 5.3292e-13,\n 1.4902e-11, 4.0170e-11, 2.4017e-11, 3.2676e-13, 5.1703e-13, 1.4943e-12,\n 1.1542e-11, 3.4949e-11, 1.5784e-11, 6.7459e-13, 7.0958e-11, 1.6969e-13,\n 4.6826e-12, 5.0268e-14, 3.3309e-12, 2.1938e-11, 1.2596e-12, 9.0829e-13,\n 4.7524e-11, 1.2111e-12, 1.8752e-11, 9.6213e-11, 3.2716e-10, 2.4016e-12,\n 4.8934e-13, 1.3050e-11, 7.6979e-12, 1.7565e-14, 4.9380e-12, 1.3570e-10,\n 2.3473e-12, 1.3496e-11, 1.2515e-11, 1.1450e-12, 1.0147e-12, 1.8941e-11,\n 2.3434e-11, 1.3669e-12, 2.2276e-11, 3.9148e-12, 9.3192e-12, 1.5611e-11,\n 1.3433e-11, 7.5452e-12, 3.6201e-12, 8.3728e-12, 2.0787e-12, 4.8442e-13,\n 3.5726e-11, 8.7299e-12, 1.1567e-12, 3.1717e-11, 3.1880e-12, 3.7406e-13,\n 2.9842e-11, 3.3346e-11, 9.0927e-11, 2.9975e-11, 9.5348e-13, 1.9026e-11,\n 1.7026e-12, 3.0821e-11, 6.2195e-13, 3.0994e-12], device='cuda:0')" + "exp_avg_sq": "tensor([1.0965e-14, 8.1677e-15, 1.1326e-11, 3.1913e-14, 2.3499e-12, 3.0897e-12,\n 1.8298e-12, 2.9869e-13, 3.6595e-11, 1.0826e-12, 8.6862e-12, 6.6925e-12,\n 2.4492e-14, 8.0306e-13, 2.4642e-14, 8.1033e-14, 3.7334e-12, 2.2649e-12,\n 3.5815e-13, 1.7199e-13, 7.4858e-13, 1.0399e-13, 3.6446e-12, 1.9683e-12,\n 5.2420e-12, 6.1140e-12, 2.2390e-12, 2.8425e-12, 1.6285e-12, 1.7088e-11,\n 1.6120e-11, 7.4726e-13, 7.2563e-11, 5.5683e-13, 9.3930e-13, 1.2497e-13,\n 2.4477e-12, 1.7132e-13, 1.8828e-12, 6.5874e-13, 5.1079e-12, 1.5958e-13,\n 3.0512e-14, 3.6194e-11, 1.4355e-14, 1.3716e-12, 3.4667e-12, 3.3850e-12,\n 6.8664e-14, 2.1310e-12, 3.3683e-12, 3.5696e-13, 1.1873e-11, 7.8024e-13,\n 5.0708e-13, 8.2075e-12, 3.9243e-13, 7.1681e-13, 1.4697e-11, 1.6130e-11,\n 2.9179e-12, 3.6707e-12, 1.7622e-13, 2.2064e-12, 2.7063e-12, 1.2985e-11,\n 1.1566e-13, 3.6704e-12, 8.6901e-13, 1.9491e-12, 7.2999e-13, 7.6169e-13,\n 1.4523e-12, 2.0391e-11, 6.4525e-12, 1.6667e-11, 4.0135e-12, 6.2841e-13,\n 7.1202e-13, 2.1829e-14, 1.7784e-12, 2.8522e-13, 2.5454e-13, 4.8311e-12,\n 1.1927e-12, 2.4007e-13, 2.6989e-11, 3.0984e-12, 9.9364e-13, 1.9922e-14,\n 9.1022e-14, 1.6155e-12, 2.8975e-12, 9.1241e-13, 2.6419e-11, 1.0395e-13,\n 2.4937e-11, 1.7262e-11, 2.6725e-12, 4.9825e-13, 8.0232e-13, 5.7974e-13,\n 5.1565e-14, 1.2916e-11, 1.7279e-13, 2.1640e-12, 1.8522e-12, 4.2461e-13,\n 1.8571e-12, 8.0001e-12, 2.4589e-12, 1.9312e-13, 9.3416e-12, 8.0510e-13,\n 8.9256e-13, 3.0657e-12, 3.7297e-13, 2.0552e-12, 5.0510e-13, 7.9251e-12,\n 2.0746e-11, 3.0223e-12, 2.4892e-12, 1.8054e-13, 7.1965e-13, 2.8256e-11,\n 2.2017e-13, 1.8914e-12, 6.9876e-13, 2.1021e-12, 7.0664e-14, 7.5622e-13,\n 2.7903e-11, 1.3297e-12, 9.5982e-14, 2.4130e-12, 1.8707e-12, 2.2264e-13,\n 7.5469e-12, 1.3476e-11, 3.6214e-13, 8.5298e-13, 9.6035e-12, 7.9610e-14,\n 4.9252e-13, 2.4922e-12, 6.3592e-13, 8.3297e-12, 2.2285e-13, 5.4473e-12,\n 5.7446e-13, 1.2861e-14, 7.2566e-12, 1.7989e-12, 6.6940e-13, 7.1625e-12,\n 2.4207e-12, 9.1675e-12, 3.0266e-14, 1.4368e-12, 3.4553e-13, 7.6427e-12,\n 2.6599e-14, 1.1770e-11, 1.5909e-12, 7.0976e-12, 7.0508e-13, 2.5203e-12,\n 7.6264e-14, 5.6996e-15, 6.2157e-13, 2.7609e-12, 2.2947e-12, 3.0143e-12,\n 1.0724e-12, 9.3816e-13, 1.2680e-11, 2.8283e-12, 1.1738e-13, 7.7460e-15,\n 3.6037e-13, 2.8058e-12, 4.7394e-14, 3.9290e-13, 2.3504e-13, 3.4948e-11,\n 2.9163e-12, 8.7033e-14, 8.7725e-13, 9.8970e-13, 9.8269e-12, 1.5229e-13,\n 4.2584e-12, 1.1479e-11, 6.8631e-12, 9.3376e-14, 1.4775e-13, 4.2700e-13,\n 3.2983e-12, 9.9869e-12, 4.5104e-12, 1.9277e-13, 2.0277e-11, 4.8491e-14,\n 1.3381e-12, 1.4364e-14, 9.5183e-13, 6.2691e-12, 3.5995e-13, 2.5955e-13,\n 1.3580e-11, 3.4608e-13, 5.3586e-12, 2.7494e-11, 9.3488e-11, 6.8628e-13,\n 1.3983e-13, 3.7291e-12, 2.1997e-12, 5.0194e-15, 1.4111e-12, 3.8778e-11,\n 6.7075e-13, 3.8566e-12, 3.5763e-12, 3.2718e-13, 2.8994e-13, 5.4126e-12,\n 6.6965e-12, 3.9061e-13, 6.3657e-12, 1.1187e-12, 2.6630e-12, 4.4608e-12,\n 3.8386e-12, 2.1561e-12, 1.0345e-12, 2.3926e-12, 5.9401e-13, 1.3843e-13,\n 1.0209e-11, 2.4946e-12, 3.3053e-13, 9.0633e-12, 9.1098e-13, 1.0689e-13,\n 8.5275e-12, 9.5290e-12, 2.5983e-11, 8.5657e-12, 2.7246e-13, 5.4368e-12,\n 4.8654e-13, 8.8074e-12, 1.7773e-13, 8.8568e-13], device='cuda:0')" }, "44": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[ 5.6052e-45, -5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n 0.0000e+00, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [-5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[9.8859e-16, 1.6766e-15, 6.6839e-15, ..., 1.1997e-15, 0.0000e+00,\n 2.6504e-14],\n [4.3003e-15, 5.3131e-16, 4.8592e-14, ..., 3.1650e-14, 0.0000e+00,\n 1.4534e-13],\n [8.3781e-12, 9.9935e-13, 6.3691e-11, ..., 3.9717e-11, 0.0000e+00,\n 1.0498e-10],\n ...,\n [2.2334e-12, 2.7086e-13, 1.5270e-11, ..., 7.2101e-12, 0.0000e+00,\n 7.8554e-12],\n [9.7118e-15, 5.8925e-15, 2.1534e-14, ..., 1.6115e-13, 0.0000e+00,\n 9.7193e-14],\n [2.1374e-15, 9.0714e-16, 8.0825e-14, ..., 6.0784e-14, 0.0000e+00,\n 4.1355e-15]], device='cuda:0')" + "exp_avg_sq": "tensor([[2.8250e-16, 4.7909e-16, 1.9100e-15, ..., 3.4282e-16, 0.0000e+00,\n 7.5738e-15],\n [1.2288e-15, 1.5183e-16, 1.3885e-14, ..., 9.0444e-15, 0.0000e+00,\n 4.1533e-14],\n [2.3941e-12, 2.8557e-13, 1.8200e-11, ..., 1.1350e-11, 0.0000e+00,\n 2.9998e-11],\n ...,\n [6.3820e-13, 7.7401e-14, 4.3636e-12, ..., 2.0603e-12, 0.0000e+00,\n 2.2447e-12],\n [2.7752e-15, 1.6838e-15, 6.1535e-15, ..., 4.6049e-14, 0.0000e+00,\n 2.7774e-14],\n [6.1079e-16, 2.5922e-16, 2.3096e-14, ..., 1.7369e-14, 0.0000e+00,\n 1.1818e-15]], device='cuda:0')" }, "45": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([4.1587e-12, 1.0258e-11, 5.2222e-08, 1.6983e-12, 1.7321e-08, 1.0815e-08,\n 2.5725e-09, 1.5157e-09, 5.2759e-08, 1.0933e-09, 1.3958e-09, 4.9774e-09,\n 6.0315e-11, 1.1297e-09, 2.4159e-10, 4.3801e-10, 3.9493e-09, 9.7146e-10,\n 1.3592e-10, 4.3434e-11, 3.7198e-09, 1.7985e-10, 5.0085e-09, 5.3230e-10,\n 4.5657e-09, 1.6291e-08, 1.9914e-09, 2.0465e-09, 7.3020e-10, 1.6358e-08,\n 5.6904e-10, 2.2056e-09, 8.5879e-08, 1.0431e-09, 5.6125e-10, 3.2186e-11,\n 5.6316e-09, 3.2565e-11, 3.4699e-09, 1.1054e-09, 2.6281e-09, 9.7948e-11,\n 2.9297e-11, 8.1461e-09, 2.6495e-12, 3.0328e-09, 1.3411e-09, 1.8016e-09,\n 3.4272e-10, 1.3033e-09, 1.0961e-08, 6.9497e-10, 4.2804e-09, 7.9403e-10,\n 4.4144e-10, 2.1614e-08, 1.2758e-10, 1.8357e-09, 2.2877e-09, 3.9338e-08,\n 4.0604e-09, 3.8439e-09, 2.0423e-10, 6.1403e-09, 3.5784e-09, 1.3649e-08,\n 3.9537e-11, 7.4309e-09, 5.3982e-10, 1.8043e-09, 1.8412e-09, 5.4105e-09,\n 2.2138e-10, 1.4547e-08, 1.5592e-08, 1.2709e-08, 8.9123e-10, 9.3765e-10,\n 2.7034e-10, 8.4813e-11, 3.7463e-09, 2.6543e-10, 4.1189e-10, 2.9595e-10,\n 3.7295e-09, 9.4862e-12, 1.1541e-08, 1.6290e-09, 8.8268e-10, 3.4934e-12,\n 2.3644e-11, 2.0222e-09, 6.0008e-09, 4.0728e-09, 1.1907e-08, 6.8144e-10,\n 1.8938e-08, 1.2543e-08, 5.9406e-10, 6.5294e-10, 2.1073e-09, 7.4401e-10,\n 1.3450e-09, 1.9550e-08, 2.8603e-09, 2.6739e-09, 4.2038e-09, 2.9097e-10,\n 2.9240e-09, 4.4050e-10, 1.1259e-10, 7.1108e-10, 7.1173e-10, 1.7799e-09,\n 2.9149e-10, 3.9559e-09, 4.6741e-10, 2.1354e-08, 2.0059e-10, 1.2378e-08,\n 2.2909e-08, 1.8146e-09, 1.6237e-08, 9.5174e-11, 8.0831e-10, 2.9796e-08,\n 1.6219e-09, 8.5196e-10, 8.6713e-10, 2.6594e-09, 7.3665e-11, 1.2729e-09,\n 6.7353e-08, 5.9664e-09, 9.2641e-11, 4.0398e-09, 2.3743e-09, 2.5249e-09,\n 2.1754e-08, 9.2357e-09, 1.8878e-09, 4.1877e-10, 5.8957e-09, 3.4569e-11,\n 2.3900e-11, 4.2855e-09, 3.0735e-09, 3.6681e-09, 5.8919e-11, 9.3767e-10,\n 1.2395e-09, 6.3439e-13, 4.7891e-09, 1.1601e-09, 2.2135e-10, 2.4290e-08,\n 1.4539e-09, 6.0592e-09, 1.7241e-11, 7.2027e-10, 1.6535e-10, 1.4746e-08,\n 2.7548e-11, 7.3235e-09, 2.5575e-08, 4.6651e-09, 7.4057e-10, 3.9279e-09,\n 6.2221e-11, 1.3551e-12, 3.4023e-09, 7.8422e-09, 1.9002e-08, 2.1820e-09,\n 2.6323e-09, 4.9445e-11, 6.8330e-09, 1.6434e-09, 1.3776e-10, 1.6670e-10,\n 4.0919e-10, 4.1001e-09, 1.8277e-11, 1.2644e-10, 5.6685e-10, 7.4711e-08,\n 1.1766e-09, 1.6209e-11, 1.8771e-10, 5.9601e-10, 1.2020e-08, 5.0851e-11,\n 8.8749e-09, 2.8497e-09, 1.5560e-09, 1.8228e-09, 1.5711e-10, 4.5174e-09,\n 3.1762e-09, 5.1915e-09, 4.7196e-09, 2.3353e-11, 2.3779e-08, 7.4926e-12,\n 3.0716e-09, 1.3894e-11, 1.6466e-09, 7.2011e-09, 1.8283e-10, 1.4364e-09,\n 4.9520e-09, 1.7457e-09, 1.6435e-08, 2.3800e-08, 4.2778e-08, 6.0696e-09,\n 7.8245e-10, 1.0180e-09, 3.0060e-08, 1.6216e-10, 2.8963e-09, 4.1981e-08,\n 2.3454e-10, 3.2466e-09, 3.8639e-08, 5.9683e-10, 8.6793e-10, 4.0462e-09,\n 6.4395e-09, 2.2091e-09, 2.4878e-09, 1.8647e-09, 3.7747e-09, 1.3925e-09,\n 4.2200e-09, 5.5317e-09, 6.2899e-10, 4.6773e-10, 2.8839e-09, 8.6494e-11,\n 2.6354e-08, 1.8467e-09, 1.1390e-10, 1.3904e-08, 4.2821e-09, 1.9834e-11,\n 5.1507e-09, 2.0828e-09, 4.6503e-08, 3.2928e-08, 7.1124e-10, 2.0282e-09,\n 8.2025e-10, 1.2842e-08, 2.6166e-10, 7.1461e-11], device='cuda:0')" + "exp_avg_sq": "tensor([1.1884e-12, 2.9314e-12, 1.4923e-08, 4.8531e-13, 4.9497e-09, 3.0906e-09,\n 7.3512e-10, 4.3313e-10, 1.5076e-08, 3.1241e-10, 3.9887e-10, 1.4223e-09,\n 1.7236e-11, 3.2281e-10, 6.9037e-11, 1.2517e-10, 1.1286e-09, 2.7760e-10,\n 3.8841e-11, 1.2412e-11, 1.0630e-09, 5.1395e-11, 1.4312e-09, 1.5211e-10,\n 1.3047e-09, 4.6552e-09, 5.6906e-10, 5.8481e-10, 2.0866e-10, 4.6744e-09,\n 1.6261e-10, 6.3026e-10, 2.4541e-08, 2.9807e-10, 1.6038e-10, 9.1973e-12,\n 1.6093e-09, 9.3057e-12, 9.9154e-10, 3.1587e-10, 7.5100e-10, 2.7989e-11,\n 8.3720e-12, 2.3278e-09, 7.5713e-13, 8.6664e-10, 3.8324e-10, 5.1481e-10,\n 9.7936e-11, 3.7243e-10, 3.1321e-09, 1.9859e-10, 1.2232e-09, 2.2690e-10,\n 1.2615e-10, 6.1765e-09, 3.6456e-11, 5.2456e-10, 6.5373e-10, 1.1241e-08,\n 1.1603e-09, 1.0984e-09, 5.8362e-11, 1.7546e-09, 1.0226e-09, 3.9004e-09,\n 1.1298e-11, 2.1234e-09, 1.5426e-10, 5.1558e-10, 5.2615e-10, 1.5461e-09,\n 6.3260e-11, 4.1568e-09, 4.4556e-09, 3.6318e-09, 2.5468e-10, 2.6794e-10,\n 7.7251e-11, 2.4236e-11, 1.0705e-09, 7.5850e-11, 1.1770e-10, 8.4569e-11,\n 1.0657e-09, 2.7108e-12, 3.2979e-09, 4.6549e-10, 2.5223e-10, 9.9825e-13,\n 6.7565e-12, 5.7785e-10, 1.7148e-09, 1.1638e-09, 3.4024e-09, 1.9473e-10,\n 5.4117e-09, 3.5843e-09, 1.6976e-10, 1.8658e-10, 6.0217e-10, 2.1261e-10,\n 3.8435e-10, 5.5867e-09, 8.1734e-10, 7.6408e-10, 1.2013e-09, 8.3148e-11,\n 8.3557e-10, 1.2588e-10, 3.2174e-11, 2.0320e-10, 2.0338e-10, 5.0862e-10,\n 8.3296e-11, 1.1304e-09, 1.3357e-10, 6.1020e-09, 5.7319e-11, 3.5370e-09,\n 6.5463e-09, 5.1853e-10, 4.6398e-09, 2.7197e-11, 2.3098e-10, 8.5146e-09,\n 4.6347e-10, 2.4345e-10, 2.4779e-10, 7.5995e-10, 2.1050e-11, 3.6374e-10,\n 1.9247e-08, 1.7050e-09, 2.6473e-11, 1.1544e-09, 6.7848e-10, 7.2152e-10,\n 6.2165e-09, 2.6392e-09, 5.3945e-10, 1.1967e-10, 1.6847e-09, 9.8783e-12,\n 6.8295e-12, 1.2246e-09, 8.7828e-10, 1.0482e-09, 1.6836e-11, 2.6795e-10,\n 3.5420e-10, 1.8128e-13, 1.3685e-09, 3.3152e-10, 6.3253e-11, 6.9411e-09,\n 4.1545e-10, 1.7315e-09, 4.9267e-12, 2.0582e-10, 4.7251e-11, 4.2138e-09,\n 7.8721e-12, 2.0928e-09, 7.3083e-09, 1.3331e-09, 2.1162e-10, 1.1224e-09,\n 1.7780e-11, 3.8724e-13, 9.7225e-10, 2.2410e-09, 5.4300e-09, 6.2352e-10,\n 7.5222e-10, 1.4129e-11, 1.9526e-09, 4.6961e-10, 3.9367e-11, 4.7636e-11,\n 1.1693e-10, 1.1716e-09, 5.2227e-12, 3.6131e-11, 1.6198e-10, 2.1349e-08,\n 3.3623e-10, 4.6319e-12, 5.3639e-11, 1.7031e-10, 3.4348e-09, 1.4531e-11,\n 2.5361e-09, 8.1431e-10, 4.4464e-10, 5.2088e-10, 4.4895e-11, 1.2909e-09,\n 9.0764e-10, 1.4835e-09, 1.3487e-09, 6.6733e-12, 6.7951e-09, 2.1411e-12,\n 8.7775e-10, 3.9703e-12, 4.7052e-10, 2.0578e-09, 5.2246e-11, 4.1046e-10,\n 1.4151e-09, 4.9886e-10, 4.6963e-09, 6.8011e-09, 1.2224e-08, 1.7344e-09,\n 2.2359e-10, 2.9090e-10, 8.5897e-09, 4.6339e-11, 8.2763e-10, 1.1997e-08,\n 6.7021e-11, 9.2774e-10, 1.1041e-08, 1.7055e-10, 2.4802e-10, 1.1562e-09,\n 1.8401e-09, 6.3127e-10, 7.1090e-10, 5.3285e-10, 1.0786e-09, 3.9791e-10,\n 1.2059e-09, 1.5807e-09, 1.7974e-10, 1.3366e-10, 8.2410e-10, 2.4716e-11,\n 7.5309e-09, 5.2770e-10, 3.2548e-11, 3.9731e-09, 1.2237e-09, 5.6678e-12,\n 1.4719e-09, 5.9519e-10, 1.3289e-08, 9.4095e-09, 2.0324e-10, 5.7958e-10,\n 2.3439e-10, 3.6698e-09, 7.4772e-11, 2.0421e-11], device='cuda:0')" }, "46": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([1.6910e-14, 1.2480e-13, 1.1601e-10, 2.3355e-15, 3.0010e-11, 2.0215e-11,\n 4.4793e-12, 4.8143e-12, 1.1492e-10, 2.6469e-12, 2.6985e-12, 8.5292e-12,\n 1.1463e-13, 1.8682e-12, 5.5823e-13, 1.6039e-12, 7.8503e-12, 2.0151e-12,\n 3.3576e-13, 2.2860e-13, 5.6979e-12, 9.2634e-13, 1.4247e-11, 8.6733e-13,\n 1.5434e-11, 1.9867e-11, 4.0770e-12, 3.8751e-12, 3.8925e-12, 3.1569e-11,\n 1.8073e-12, 4.2801e-12, 1.7605e-10, 1.4493e-12, 8.9209e-13, 1.5721e-13,\n 9.1879e-12, 5.9724e-13, 5.2179e-12, 2.8333e-12, 3.8141e-12, 1.4796e-13,\n 6.8793e-14, 1.5738e-11, 5.1115e-15, 6.4912e-12, 1.7920e-12, 5.3195e-12,\n 1.1283e-12, 1.7709e-12, 1.9503e-11, 8.7206e-13, 5.0479e-12, 1.2134e-12,\n 2.2808e-12, 6.0230e-11, 4.1495e-13, 2.7983e-12, 3.7216e-12, 7.9669e-11,\n 5.7428e-12, 6.2840e-12, 6.7150e-13, 6.4550e-12, 7.1654e-12, 2.1805e-11,\n 8.8047e-14, 1.2142e-11, 8.0042e-13, 2.6405e-12, 3.3404e-12, 1.3347e-11,\n 2.5740e-13, 2.5090e-11, 3.9393e-11, 1.8914e-11, 1.2205e-12, 1.2580e-12,\n 3.5579e-13, 5.7782e-14, 5.2497e-12, 3.6730e-13, 7.9667e-13, 4.2524e-13,\n 9.5440e-12, 6.8857e-14, 1.6096e-11, 6.9227e-12, 1.9753e-12, 9.1232e-15,\n 3.3390e-15, 3.7197e-12, 1.0894e-11, 5.9461e-12, 1.4687e-11, 6.3752e-13,\n 3.5383e-11, 1.9679e-11, 1.1714e-12, 2.5271e-12, 4.2596e-12, 1.4625e-12,\n 3.1234e-12, 4.6063e-11, 5.2533e-12, 3.5770e-12, 6.1751e-12, 5.9055e-13,\n 2.6355e-12, 1.0549e-12, 3.5702e-13, 2.8473e-12, 2.5490e-12, 2.3952e-12,\n 3.6645e-13, 1.0950e-11, 1.3492e-12, 7.3201e-11, 8.3040e-13, 2.8953e-11,\n 4.1346e-11, 3.9590e-12, 3.8575e-11, 1.0790e-12, 2.3888e-12, 6.4983e-11,\n 1.4648e-12, 9.7196e-13, 1.9230e-12, 5.2255e-12, 2.3006e-13, 2.0217e-12,\n 2.4254e-10, 6.7560e-12, 2.4160e-13, 9.0379e-12, 8.9043e-12, 5.1645e-12,\n 5.2840e-11, 9.6082e-12, 3.8779e-12, 1.0046e-12, 9.4522e-12, 1.1268e-13,\n 8.0429e-14, 1.4449e-11, 4.9085e-12, 5.7726e-12, 1.6560e-13, 1.3987e-12,\n 4.1301e-12, 7.8962e-15, 6.1451e-12, 1.8029e-12, 3.4600e-13, 3.4646e-11,\n 1.9719e-12, 8.0123e-12, 1.7280e-14, 1.3620e-12, 3.6441e-13, 3.0919e-11,\n 2.8876e-13, 1.6381e-11, 6.1469e-11, 6.1147e-12, 1.5772e-12, 8.6822e-12,\n 2.6229e-13, 4.7375e-14, 6.7130e-12, 3.4496e-11, 4.2818e-11, 3.0251e-12,\n 7.2428e-12, 1.3377e-13, 9.2929e-12, 2.5225e-12, 9.0127e-13, 9.4583e-13,\n 1.0173e-12, 1.0329e-11, 6.1456e-14, 3.1722e-13, 2.5713e-12, 2.0816e-10,\n 1.4206e-12, 5.5624e-14, 2.9226e-13, 1.9748e-12, 3.6542e-11, 2.1833e-13,\n 1.3833e-11, 4.0476e-12, 3.4298e-12, 4.5307e-12, 2.3525e-13, 6.3536e-12,\n 4.3504e-12, 7.2294e-12, 9.1392e-12, 1.1665e-13, 2.7484e-11, 4.8954e-14,\n 4.9409e-12, 4.2065e-14, 2.4373e-12, 1.2343e-11, 1.3425e-12, 3.1453e-12,\n 7.6306e-12, 2.2869e-12, 3.6529e-11, 3.6614e-11, 6.5797e-11, 1.6088e-11,\n 1.8145e-12, 1.4024e-12, 5.9630e-11, 5.8338e-13, 4.2019e-12, 1.0144e-10,\n 5.7183e-13, 8.1506e-12, 1.0684e-10, 1.0319e-12, 4.1657e-12, 4.7616e-12,\n 1.0626e-11, 4.4567e-12, 4.5365e-12, 2.5797e-12, 9.7351e-12, 2.4483e-12,\n 5.9293e-12, 1.1525e-11, 9.9460e-13, 7.3762e-13, 5.6430e-12, 2.3315e-13,\n 8.4356e-11, 3.0493e-12, 4.7529e-13, 2.0433e-11, 4.7109e-12, 4.8635e-14,\n 9.8702e-12, 2.4215e-12, 1.1779e-10, 1.3514e-10, 1.0660e-12, 2.3356e-12,\n 1.3917e-12, 3.1707e-11, 6.4994e-13, 1.4066e-13], device='cuda:0')" + "exp_avg_sq": "tensor([4.8322e-15, 3.5663e-14, 3.3152e-11, 6.6740e-16, 8.5755e-12, 5.7766e-12,\n 1.2800e-12, 1.3757e-12, 3.2840e-11, 7.5638e-13, 7.7111e-13, 2.4373e-12,\n 3.2758e-14, 5.3384e-13, 1.5952e-13, 4.5832e-13, 2.2433e-12, 5.7582e-13,\n 9.5947e-14, 6.5324e-14, 1.6282e-12, 2.6471e-13, 4.0711e-12, 2.4785e-13,\n 4.4104e-12, 5.6772e-12, 1.1650e-12, 1.1073e-12, 1.1123e-12, 9.0210e-12,\n 5.1644e-13, 1.2231e-12, 5.0309e-11, 4.1414e-13, 2.5492e-13, 4.4923e-14,\n 2.6255e-12, 1.7067e-13, 1.4911e-12, 8.0963e-13, 1.0899e-12, 4.2282e-14,\n 1.9658e-14, 4.4973e-12, 1.4607e-15, 1.8549e-12, 5.1209e-13, 1.5201e-12,\n 3.2241e-13, 5.0605e-13, 5.5731e-12, 2.4920e-13, 1.4425e-12, 3.4674e-13,\n 6.5176e-13, 1.7211e-11, 1.1857e-13, 7.9963e-13, 1.0635e-12, 2.2766e-11,\n 1.6411e-12, 1.7957e-12, 1.9188e-13, 1.8446e-12, 2.0476e-12, 6.2309e-12,\n 2.5160e-14, 3.4696e-12, 2.2873e-13, 7.5453e-13, 9.5455e-13, 3.8140e-12,\n 7.3554e-14, 7.1697e-12, 1.1257e-11, 5.4048e-12, 3.4877e-13, 3.5950e-13,\n 1.0167e-13, 1.6512e-14, 1.5002e-12, 1.0496e-13, 2.2766e-13, 1.2152e-13,\n 2.7273e-12, 1.9676e-14, 4.5996e-12, 1.9782e-12, 5.6446e-13, 2.6070e-15,\n 9.5414e-16, 1.0629e-12, 3.1129e-12, 1.6991e-12, 4.1968e-12, 1.8218e-13,\n 1.0111e-11, 5.6235e-12, 3.3473e-13, 7.2214e-13, 1.2172e-12, 4.1792e-13,\n 8.9253e-13, 1.3163e-11, 1.5012e-12, 1.0222e-12, 1.7646e-12, 1.6876e-13,\n 7.5313e-13, 3.0146e-13, 1.0202e-13, 8.1364e-13, 7.2840e-13, 6.8445e-13,\n 1.0471e-13, 3.1290e-12, 3.8553e-13, 2.0918e-11, 2.3729e-13, 8.2734e-12,\n 1.1815e-11, 1.1313e-12, 1.1023e-11, 3.0832e-13, 6.8262e-13, 1.8569e-11,\n 4.1859e-13, 2.7774e-13, 5.4951e-13, 1.4932e-12, 6.5743e-14, 5.7771e-13,\n 6.9309e-11, 1.9306e-12, 6.9040e-14, 2.5827e-12, 2.5445e-12, 1.4758e-12,\n 1.5099e-11, 2.7456e-12, 1.1081e-12, 2.8707e-13, 2.7010e-12, 3.2198e-14,\n 2.2983e-14, 4.1290e-12, 1.4026e-12, 1.6496e-12, 4.7321e-14, 3.9970e-13,\n 1.1802e-12, 2.2564e-15, 1.7560e-12, 5.1519e-13, 9.8871e-14, 9.9003e-12,\n 5.6349e-13, 2.2896e-12, 4.9378e-15, 3.8919e-13, 1.0413e-13, 8.8353e-12,\n 8.2515e-14, 4.6810e-12, 1.7565e-11, 1.7473e-12, 4.5070e-13, 2.4810e-12,\n 7.4951e-14, 1.3538e-14, 1.9183e-12, 9.8576e-12, 1.2236e-11, 8.6446e-13,\n 2.0697e-12, 3.8226e-14, 2.6555e-12, 7.2084e-13, 2.5755e-13, 2.7028e-13,\n 2.9071e-13, 2.9515e-12, 1.7561e-14, 9.0649e-14, 7.3478e-13, 5.9482e-11,\n 4.0596e-13, 1.5895e-14, 8.3515e-14, 5.6432e-13, 1.0442e-11, 6.2391e-14,\n 3.9528e-12, 1.1566e-12, 9.8010e-13, 1.2947e-12, 6.7226e-14, 1.8156e-12,\n 1.2432e-12, 2.0658e-12, 2.6116e-12, 3.3333e-14, 7.8536e-12, 1.3989e-14,\n 1.4119e-12, 1.2020e-14, 6.9648e-13, 3.5272e-12, 3.8364e-13, 8.9878e-13,\n 2.1805e-12, 6.5350e-13, 1.0438e-11, 1.0463e-11, 1.8802e-11, 4.5972e-12,\n 5.1852e-13, 4.0075e-13, 1.7040e-11, 1.6671e-13, 1.2007e-12, 2.8989e-11,\n 1.6340e-13, 2.3291e-12, 3.0531e-11, 2.9487e-13, 1.1904e-12, 1.3607e-12,\n 3.0363e-12, 1.2735e-12, 1.2964e-12, 7.3715e-13, 2.7819e-12, 6.9962e-13,\n 1.6944e-12, 3.2933e-12, 2.8422e-13, 2.1078e-13, 1.6125e-12, 6.6624e-14,\n 2.4105e-11, 8.7137e-13, 1.3582e-13, 5.8388e-12, 1.3462e-12, 1.3898e-14,\n 2.8205e-12, 6.9197e-13, 3.3660e-11, 3.8617e-11, 3.0462e-13, 6.6741e-13,\n 3.9768e-13, 9.0606e-12, 1.8573e-13, 4.0195e-14], device='cuda:0')" }, "47": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([4.2961e-14, 8.7322e-14, 1.4351e-10, 2.1152e-14, 4.2460e-11, 2.6469e-11,\n 7.3760e-12, 5.0806e-12, 1.4133e-10, 3.1921e-12, 4.4323e-12, 1.2431e-11,\n 2.5971e-13, 3.7291e-12, 1.0639e-12, 1.4075e-12, 1.1403e-11, 3.1617e-12,\n 5.0781e-13, 3.0541e-13, 1.0853e-11, 9.1541e-13, 1.2907e-11, 1.8306e-12,\n 1.4299e-11, 4.0954e-11, 6.2449e-12, 5.5499e-12, 3.2075e-12, 4.6348e-11,\n 1.6282e-12, 5.7856e-12, 2.3388e-10, 2.5118e-12, 1.7461e-12, 2.2901e-13,\n 1.3107e-11, 6.1610e-13, 1.0645e-11, 3.5860e-12, 6.6771e-12, 4.1129e-13,\n 1.2475e-13, 2.4782e-11, 1.3143e-14, 6.8017e-12, 3.8740e-12, 6.3224e-12,\n 1.2707e-12, 4.0446e-12, 3.3068e-11, 1.4648e-12, 1.0696e-11, 2.1730e-12,\n 1.8403e-12, 5.4709e-11, 6.1511e-13, 4.2284e-12, 5.9706e-12, 1.1046e-10,\n 9.3553e-12, 8.9044e-12, 8.1452e-13, 1.7360e-11, 1.0369e-11, 3.2166e-11,\n 1.9433e-13, 1.8360e-11, 1.6675e-12, 5.5954e-12, 5.9034e-12, 1.6726e-11,\n 4.6873e-13, 4.1145e-11, 4.4548e-11, 3.3011e-11, 2.8812e-12, 3.0741e-12,\n 9.2058e-13, 9.4088e-14, 1.1418e-11, 7.5864e-13, 1.2825e-12, 9.7047e-13,\n 8.4739e-12, 2.0804e-13, 3.3725e-11, 5.9136e-12, 3.1804e-12, 1.2696e-14,\n 3.2986e-14, 5.1512e-12, 1.4728e-11, 1.0767e-11, 3.0582e-11, 1.5479e-12,\n 4.9221e-11, 3.1968e-11, 1.5861e-12, 2.6995e-12, 6.5358e-12, 2.3632e-12,\n 4.4830e-12, 5.0080e-11, 8.4217e-12, 6.6929e-12, 1.2084e-11, 1.0948e-12,\n 6.5900e-12, 1.2920e-12, 3.5401e-13, 2.7275e-12, 1.9555e-12, 4.0471e-12,\n 1.0383e-12, 1.1129e-11, 1.8654e-12, 5.2068e-11, 1.0028e-12, 3.5284e-11,\n 6.4567e-11, 5.9778e-12, 4.1777e-11, 1.6377e-12, 2.9711e-12, 7.9615e-11,\n 3.7869e-12, 2.5918e-12, 2.8660e-12, 7.9046e-12, 3.8584e-13, 4.0753e-12,\n 1.7169e-10, 1.4956e-11, 3.8554e-13, 9.7635e-12, 7.0886e-12, 5.7711e-12,\n 5.3813e-11, 2.5319e-11, 5.5632e-12, 1.2976e-12, 1.4572e-11, 2.4282e-13,\n 1.1378e-13, 1.2989e-11, 6.4210e-12, 1.0382e-11, 3.6319e-13, 3.0793e-12,\n 4.3029e-12, 3.0514e-16, 1.2469e-11, 2.8085e-12, 6.9136e-13, 6.0078e-11,\n 3.8611e-12, 1.5585e-11, 2.3368e-14, 2.5282e-12, 7.0607e-13, 4.1235e-11,\n 3.6250e-13, 2.0890e-11, 7.2194e-11, 1.1930e-11, 1.7617e-12, 9.8458e-12,\n 3.9597e-13, 5.2749e-14, 7.9145e-12, 2.3298e-11, 4.7184e-11, 6.6788e-12,\n 6.5383e-12, 1.9624e-13, 1.9491e-11, 5.0178e-12, 9.2949e-13, 8.2414e-13,\n 1.6430e-12, 1.1008e-11, 8.2518e-14, 5.0388e-13, 2.2672e-12, 2.0388e-10,\n 2.9045e-12, 1.1521e-13, 6.7653e-13, 2.3480e-12, 3.0208e-11, 4.4789e-13,\n 2.2134e-11, 9.0086e-12, 4.0276e-12, 5.9263e-12, 5.2535e-13, 1.1372e-11,\n 8.3141e-12, 1.5294e-11, 1.1722e-11, 3.7822e-13, 6.3699e-11, 1.5565e-13,\n 7.1998e-12, 3.1474e-14, 5.0979e-12, 1.9223e-11, 1.1497e-12, 4.4189e-12,\n 1.3831e-11, 3.9063e-12, 4.5823e-11, 6.2403e-11, 1.2055e-10, 1.4502e-11,\n 2.8466e-12, 2.4143e-12, 7.6664e-11, 7.2881e-13, 6.7959e-12, 1.1203e-10,\n 8.8550e-13, 1.0051e-11, 1.0704e-10, 1.0447e-12, 3.1626e-12, 1.1666e-11,\n 1.8783e-11, 5.9869e-12, 7.4090e-12, 5.3331e-12, 8.5175e-12, 4.2652e-12,\n 1.1864e-11, 1.5898e-11, 1.8333e-12, 1.4560e-12, 6.7515e-12, 3.4409e-13,\n 6.5633e-11, 5.2101e-12, 8.4061e-13, 3.9307e-11, 1.0148e-11, 1.2214e-13,\n 1.4124e-11, 5.1210e-12, 1.2752e-10, 9.2959e-11, 2.3210e-12, 4.9945e-12,\n 1.8612e-12, 3.5700e-11, 9.6382e-13, 3.1836e-13], device='cuda:0')" + "exp_avg_sq": "tensor([1.2276e-14, 2.4953e-14, 4.1008e-11, 6.0443e-15, 1.2133e-11, 7.5637e-12,\n 2.1078e-12, 1.4518e-12, 4.0387e-11, 9.1216e-13, 1.2666e-12, 3.5523e-12,\n 7.4213e-14, 1.0656e-12, 3.0402e-13, 4.0221e-13, 3.2584e-12, 9.0349e-13,\n 1.4511e-13, 8.7274e-14, 3.1014e-12, 2.6159e-13, 3.6884e-12, 5.2312e-13,\n 4.0859e-12, 1.1703e-11, 1.7845e-12, 1.5859e-12, 9.1656e-13, 1.3244e-11,\n 4.6527e-13, 1.6533e-12, 6.6834e-11, 7.1776e-13, 4.9897e-13, 6.5442e-14,\n 3.7455e-12, 1.7605e-13, 3.0420e-12, 1.0247e-12, 1.9081e-12, 1.1753e-13,\n 3.5648e-14, 7.0818e-12, 3.7556e-15, 1.9437e-12, 1.1070e-12, 1.8067e-12,\n 3.6311e-13, 1.1558e-12, 9.4494e-12, 4.1858e-13, 3.0564e-12, 6.2096e-13,\n 5.2589e-13, 1.5633e-11, 1.7577e-13, 1.2083e-12, 1.7061e-12, 3.1566e-11,\n 2.6733e-12, 2.5445e-12, 2.3276e-13, 4.9608e-12, 2.9631e-12, 9.1916e-12,\n 5.5531e-14, 5.2466e-12, 4.7649e-13, 1.5989e-12, 1.6869e-12, 4.7795e-12,\n 1.3394e-13, 1.1758e-11, 1.2730e-11, 9.4331e-12, 8.2331e-13, 8.7845e-13,\n 2.6306e-13, 2.6886e-14, 3.2627e-12, 2.1679e-13, 3.6647e-13, 2.7732e-13,\n 2.4215e-12, 5.9448e-14, 9.6371e-12, 1.6899e-12, 9.0883e-13, 3.6280e-15,\n 9.4261e-15, 1.4720e-12, 4.2086e-12, 3.0766e-12, 8.7389e-12, 4.4233e-13,\n 1.4065e-11, 9.1352e-12, 4.5325e-13, 7.7141e-13, 1.8676e-12, 6.7530e-13,\n 1.2811e-12, 1.4311e-11, 2.4066e-12, 1.9125e-12, 3.4530e-12, 3.1286e-13,\n 1.8831e-12, 3.6920e-13, 1.0116e-13, 7.7941e-13, 5.5879e-13, 1.1565e-12,\n 2.9672e-13, 3.1801e-12, 5.3306e-13, 1.4879e-11, 2.8657e-13, 1.0083e-11,\n 1.8450e-11, 1.7082e-12, 1.1938e-11, 4.6800e-13, 8.4901e-13, 2.2750e-11,\n 1.0821e-12, 7.4063e-13, 8.1899e-13, 2.2588e-12, 1.1026e-13, 1.1646e-12,\n 4.9063e-11, 4.2739e-12, 1.1017e-13, 2.7900e-12, 2.0256e-12, 1.6491e-12,\n 1.5378e-11, 7.2352e-12, 1.5897e-12, 3.7080e-13, 4.1640e-12, 6.9388e-14,\n 3.2515e-14, 3.7117e-12, 1.8349e-12, 2.9666e-12, 1.0378e-13, 8.7995e-13,\n 1.2296e-12, 8.7196e-17, 3.5633e-12, 8.0256e-13, 1.9756e-13, 1.7168e-11,\n 1.1033e-12, 4.4537e-12, 6.6776e-15, 7.2246e-13, 2.0176e-13, 1.1783e-11,\n 1.0359e-13, 5.9694e-12, 2.0630e-11, 3.4092e-12, 5.0343e-13, 2.8135e-12,\n 1.1315e-13, 1.5073e-14, 2.2616e-12, 6.6576e-12, 1.3483e-11, 1.9085e-12,\n 1.8684e-12, 5.6077e-14, 5.5697e-12, 1.4339e-12, 2.6561e-13, 2.3550e-13,\n 4.6951e-13, 3.1455e-12, 2.3580e-14, 1.4399e-13, 6.4787e-13, 5.8259e-11,\n 8.2998e-13, 3.2923e-14, 1.9332e-13, 6.7096e-13, 8.6322e-12, 1.2799e-13,\n 6.3250e-12, 2.5743e-12, 1.1509e-12, 1.6935e-12, 1.5012e-13, 3.2495e-12,\n 2.3758e-12, 4.3705e-12, 3.3496e-12, 1.0808e-13, 1.8202e-11, 4.4479e-14,\n 2.0574e-12, 8.9940e-15, 1.4568e-12, 5.4931e-12, 3.2855e-13, 1.2627e-12,\n 3.9522e-12, 1.1162e-12, 1.3094e-11, 1.7832e-11, 3.4448e-11, 4.1441e-12,\n 8.1343e-13, 6.8991e-13, 2.1907e-11, 2.0826e-13, 1.9420e-12, 3.2013e-11,\n 2.5304e-13, 2.8721e-12, 3.0587e-11, 2.9853e-13, 9.0374e-13, 3.3336e-12,\n 5.3674e-12, 1.7108e-12, 2.1172e-12, 1.5240e-12, 2.4339e-12, 1.2188e-12,\n 3.3901e-12, 4.5429e-12, 5.2389e-13, 4.1606e-13, 1.9293e-12, 9.8326e-14,\n 1.8755e-11, 1.4888e-12, 2.4021e-13, 1.1232e-11, 2.8998e-12, 3.4903e-14,\n 4.0360e-12, 1.4634e-12, 3.6438e-11, 2.6564e-11, 6.6324e-13, 1.4272e-12,\n 5.3185e-13, 1.0202e-11, 2.7542e-13, 9.0974e-14], device='cuda:0')" }, "48": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[ 5.6052e-45, -5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [-5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45],\n [ 5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 0.0000e+00, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[5.9575e-15, 1.0825e-15, 1.6920e-14, ..., 1.9717e-14, 0.0000e+00,\n 2.6737e-15],\n [1.2008e-15, 4.9309e-15, 6.2334e-14, ..., 8.4273e-14, 0.0000e+00,\n 1.4732e-13],\n [9.9611e-12, 1.9475e-13, 4.1646e-11, ..., 1.0580e-11, 0.0000e+00,\n 5.9738e-11],\n ...,\n [1.3149e-12, 5.4443e-14, 1.5148e-11, ..., 4.4843e-12, 0.0000e+00,\n 4.8208e-12],\n [1.6362e-13, 1.9082e-16, 4.9669e-15, ..., 2.0852e-14, 0.0000e+00,\n 2.1374e-14],\n [6.1442e-16, 3.6497e-16, 1.3591e-13, ..., 1.1544e-14, 0.0000e+00,\n 9.1086e-14]], device='cuda:0')" + "exp_avg_sq": "tensor([[1.7024e-15, 3.0932e-16, 4.8349e-15, ..., 5.6343e-15, 0.0000e+00,\n 7.6404e-16],\n [3.4315e-16, 1.4091e-15, 1.7813e-14, ..., 2.4082e-14, 0.0000e+00,\n 4.2098e-14],\n [2.8464e-12, 5.5650e-14, 1.1901e-11, ..., 3.0233e-12, 0.0000e+00,\n 1.7071e-11],\n ...,\n [3.7574e-13, 1.5558e-14, 4.3287e-12, ..., 1.2814e-12, 0.0000e+00,\n 1.3776e-12],\n [4.6756e-14, 5.4528e-17, 1.4193e-15, ..., 5.9585e-15, 0.0000e+00,\n 6.1078e-15],\n [1.7558e-16, 1.0429e-16, 3.8839e-14, ..., 3.2987e-15, 0.0000e+00,\n 2.6029e-14]], device='cuda:0')" }, "49": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([5.0947e-12, 4.4473e-11, 2.3858e-08, 2.4776e-11, 3.8914e-09, 1.2540e-08,\n 5.8821e-09, 1.3252e-09, 6.2616e-10, 1.2800e-09, 1.6520e-08, 4.8650e-09,\n 9.0830e-10, 3.2470e-09, 1.4097e-11, 9.2087e-10, 2.4485e-09, 1.1582e-09,\n 1.8504e-10, 4.1756e-10, 5.9863e-09, 1.1563e-10, 4.8867e-10, 1.8430e-09,\n 3.0868e-09, 1.6117e-09, 2.5546e-09, 5.1959e-09, 9.2708e-10, 9.3659e-09,\n 1.2694e-08, 1.4827e-08, 8.0513e-08, 2.5380e-10, 3.5108e-09, 6.9901e-11,\n 7.1885e-09, 1.1870e-10, 2.6520e-08, 7.6939e-10, 4.2342e-09, 9.9113e-11,\n 4.3879e-10, 1.7272e-08, 7.8637e-12, 8.6879e-10, 3.7030e-09, 2.5574e-09,\n 5.0259e-10, 4.8721e-09, 8.9902e-09, 3.0031e-09, 9.8094e-09, 7.9978e-10,\n 1.0168e-09, 2.2526e-08, 7.7947e-11, 3.4379e-09, 1.7969e-08, 2.2301e-08,\n 2.8514e-09, 2.5430e-09, 1.0117e-10, 1.7587e-08, 8.3959e-10, 2.3366e-08,\n 2.8388e-11, 5.9323e-09, 3.3395e-09, 3.3541e-09, 6.8937e-09, 1.1108e-08,\n 2.9647e-09, 8.1908e-09, 3.6194e-09, 1.3858e-08, 7.9422e-09, 3.1311e-09,\n 4.8211e-10, 1.8586e-11, 1.6506e-08, 1.1945e-09, 9.5226e-10, 6.4057e-09,\n 1.5079e-09, 2.2129e-10, 3.5769e-08, 1.1017e-09, 1.2612e-09, 5.3666e-11,\n 1.3298e-11, 1.3558e-09, 5.0779e-09, 1.1164e-09, 3.1166e-08, 9.6190e-10,\n 1.8446e-08, 2.5491e-08, 3.6745e-09, 2.2209e-09, 2.2460e-09, 9.5263e-10,\n 4.2489e-10, 7.3610e-09, 4.2538e-09, 3.0078e-09, 4.3458e-09, 3.3514e-10,\n 2.8249e-09, 3.4042e-09, 1.6088e-09, 4.0317e-10, 1.9726e-08, 4.9554e-10,\n 6.1749e-10, 1.9661e-08, 6.3633e-10, 4.0883e-09, 7.4328e-10, 1.2593e-09,\n 4.2560e-08, 3.5303e-09, 1.5861e-08, 2.8327e-10, 1.6853e-10, 4.9753e-08,\n 6.4933e-09, 3.3917e-10, 1.7083e-10, 7.2506e-09, 1.0543e-12, 1.4841e-09,\n 2.6074e-08, 1.6477e-08, 6.6996e-11, 1.6666e-09, 4.0297e-09, 1.3637e-09,\n 1.4747e-08, 1.0786e-08, 1.7243e-09, 1.5718e-09, 7.3272e-09, 1.1334e-11,\n 5.6128e-10, 8.4648e-10, 5.4316e-10, 8.9691e-09, 5.2223e-10, 8.5303e-10,\n 1.1820e-09, 1.5417e-09, 3.4850e-09, 2.6297e-09, 1.0105e-09, 1.2003e-08,\n 2.6695e-10, 3.9351e-08, 1.4173e-11, 7.2817e-10, 2.9532e-10, 1.0118e-09,\n 1.5196e-11, 1.8407e-09, 1.1375e-08, 7.4185e-09, 4.5602e-10, 2.1160e-09,\n 2.1356e-12, 3.6921e-10, 1.7152e-10, 3.1605e-09, 2.5365e-09, 3.1362e-09,\n 1.6517e-09, 4.1372e-10, 1.3535e-08, 7.8884e-09, 7.7845e-11, 1.9900e-11,\n 2.1493e-09, 5.8254e-09, 1.6315e-11, 1.2494e-09, 8.7076e-10, 1.7093e-08,\n 3.4473e-10, 8.8320e-11, 9.2788e-10, 4.0309e-10, 1.7197e-09, 2.0035e-11,\n 4.0691e-09, 1.3312e-09, 9.4103e-09, 1.1462e-09, 5.2530e-10, 3.8064e-09,\n 1.8495e-08, 1.7844e-09, 2.6190e-09, 5.5531e-11, 3.1594e-08, 7.0392e-12,\n 3.9121e-09, 1.5901e-11, 4.3035e-09, 4.2775e-09, 1.8034e-10, 2.6475e-09,\n 2.4955e-08, 1.4796e-09, 1.7597e-09, 2.1040e-08, 8.3409e-08, 1.0594e-09,\n 8.6943e-10, 8.8088e-10, 4.3999e-08, 4.8261e-11, 8.1679e-10, 8.3923e-08,\n 1.1339e-10, 6.3925e-09, 3.0066e-08, 5.2207e-10, 8.9954e-10, 1.3215e-08,\n 3.2176e-09, 9.5854e-10, 1.7540e-08, 3.7799e-09, 2.1637e-10, 1.3968e-09,\n 6.0081e-09, 5.9623e-09, 8.9291e-10, 1.1121e-09, 6.4941e-10, 4.0829e-10,\n 3.6919e-09, 2.7609e-09, 2.2553e-10, 3.8084e-08, 5.0894e-09, 4.6730e-11,\n 4.1755e-09, 1.7187e-08, 2.1480e-08, 1.8723e-08, 5.9150e-10, 2.0446e-09,\n 8.1711e-10, 7.7826e-09, 8.4481e-11, 2.1635e-11], device='cuda:0')" + "exp_avg_sq": "tensor([1.4558e-12, 1.2708e-11, 6.8176e-09, 7.0799e-12, 1.1120e-09, 3.5835e-09,\n 1.6809e-09, 3.7869e-10, 1.7893e-10, 3.6577e-10, 4.7207e-09, 1.3902e-09,\n 2.5955e-10, 9.2785e-10, 4.0284e-12, 2.6315e-10, 6.9969e-10, 3.3096e-10,\n 5.2877e-11, 1.1932e-10, 1.7106e-09, 3.3043e-11, 1.3964e-10, 5.2666e-10,\n 8.8209e-10, 4.6056e-10, 7.2999e-10, 1.4848e-09, 2.6492e-10, 2.6764e-09,\n 3.6274e-09, 4.2369e-09, 2.3007e-08, 7.2525e-11, 1.0032e-09, 1.9975e-11,\n 2.0542e-09, 3.3920e-11, 7.5783e-09, 2.1986e-10, 1.2100e-09, 2.8322e-11,\n 1.2539e-10, 4.9357e-09, 2.2471e-12, 2.4826e-10, 1.0582e-09, 7.3079e-10,\n 1.4362e-10, 1.3922e-09, 2.5690e-09, 8.5815e-10, 2.8031e-09, 2.2854e-10,\n 2.9056e-10, 6.4370e-09, 2.2274e-11, 9.8242e-10, 5.1347e-09, 6.3727e-09,\n 8.1481e-10, 7.2667e-10, 2.8910e-11, 5.0256e-09, 2.3992e-10, 6.6770e-09,\n 8.1121e-12, 1.6952e-09, 9.5430e-10, 9.5847e-10, 1.9699e-09, 3.1742e-09,\n 8.4720e-10, 2.3406e-09, 1.0343e-09, 3.9599e-09, 2.2696e-09, 8.9473e-10,\n 1.3777e-10, 5.3111e-12, 4.7166e-09, 3.4133e-10, 2.7211e-10, 1.8305e-09,\n 4.3090e-10, 6.3235e-11, 1.0221e-08, 3.1482e-10, 3.6039e-10, 1.5336e-11,\n 3.7999e-12, 3.8743e-10, 1.4510e-09, 3.1903e-10, 8.9061e-09, 2.7487e-10,\n 5.2711e-09, 7.2843e-09, 1.0500e-09, 6.3463e-10, 6.4181e-10, 2.7222e-10,\n 1.2141e-10, 2.1035e-09, 1.2155e-09, 8.5951e-10, 1.2418e-09, 9.5768e-11,\n 8.0724e-10, 9.7279e-10, 4.5971e-10, 1.1521e-10, 5.6368e-09, 1.4160e-10,\n 1.7645e-10, 5.6184e-09, 1.8184e-10, 1.1683e-09, 2.1240e-10, 3.5986e-10,\n 1.2162e-08, 1.0088e-09, 4.5323e-09, 8.0948e-11, 4.8158e-11, 1.4217e-08,\n 1.8555e-09, 9.6920e-11, 4.8816e-11, 2.0719e-09, 3.0127e-13, 4.2409e-10,\n 7.4508e-09, 4.7086e-09, 1.9145e-11, 4.7626e-10, 1.1515e-09, 3.8969e-10,\n 4.2139e-09, 3.0823e-09, 4.9274e-10, 4.4915e-10, 2.0938e-09, 3.2388e-12,\n 1.6039e-10, 2.4189e-10, 1.5521e-10, 2.5630e-09, 1.4923e-10, 2.4376e-10,\n 3.3776e-10, 4.4056e-10, 9.9586e-10, 7.5147e-10, 2.8876e-10, 3.4300e-09,\n 7.6282e-11, 1.1245e-08, 4.0500e-12, 2.0808e-10, 8.4389e-11, 2.8912e-10,\n 4.3424e-12, 5.2600e-10, 3.2505e-09, 2.1199e-09, 1.3031e-10, 6.0468e-10,\n 6.1027e-13, 1.0551e-10, 4.9012e-11, 9.0314e-10, 7.2484e-10, 8.9621e-10,\n 4.7199e-10, 1.1822e-10, 3.8676e-09, 2.2542e-09, 2.2245e-11, 5.6865e-12,\n 6.1418e-10, 1.6647e-09, 4.6620e-12, 3.5704e-10, 2.4883e-10, 4.8845e-09,\n 9.8510e-11, 2.5238e-11, 2.6515e-10, 1.1519e-10, 4.9142e-10, 5.7253e-12,\n 1.1628e-09, 3.8041e-10, 2.6891e-09, 3.2753e-10, 1.5011e-10, 1.0877e-09,\n 5.2852e-09, 5.0990e-10, 7.4841e-10, 1.5869e-11, 9.0282e-09, 2.0115e-12,\n 1.1179e-09, 4.5439e-12, 1.2297e-09, 1.2223e-09, 5.1532e-11, 7.5656e-10,\n 7.1312e-09, 4.2280e-10, 5.0284e-10, 6.0123e-09, 2.3835e-08, 3.0273e-10,\n 2.4845e-10, 2.5172e-10, 1.2573e-08, 1.3791e-11, 2.3341e-10, 2.3982e-08,\n 3.2403e-11, 1.8267e-09, 8.5916e-09, 1.4919e-10, 2.5705e-10, 3.7763e-09,\n 9.1945e-10, 2.7391e-10, 5.0123e-09, 1.0801e-09, 6.1830e-11, 3.9915e-10,\n 1.7169e-09, 1.7038e-09, 2.5516e-10, 3.1779e-10, 1.8557e-10, 1.1667e-10,\n 1.0550e-09, 7.8894e-10, 6.4448e-11, 1.0883e-08, 1.4543e-09, 1.3353e-11,\n 1.1932e-09, 4.9114e-09, 6.1381e-09, 5.3502e-09, 1.6903e-10, 5.8425e-10,\n 2.3350e-10, 2.2239e-09, 2.4141e-11, 6.1823e-12], device='cuda:0')" }, "50": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([3.9446e-15, 1.7013e-13, 3.5363e-11, 3.7492e-13, 3.3681e-12, 2.8597e-11,\n 1.6346e-11, 2.9354e-12, 6.1217e-12, 2.2087e-12, 3.5292e-11, 8.6545e-12,\n 3.9509e-12, 5.9049e-12, 5.7731e-14, 3.2940e-12, 4.0845e-12, 2.3014e-12,\n 7.1147e-13, 7.6498e-13, 1.1166e-11, 1.8883e-13, 9.4184e-13, 2.6065e-12,\n 5.8489e-12, 3.3253e-12, 3.2642e-12, 2.2662e-11, 6.0461e-12, 1.1400e-11,\n 1.6907e-11, 3.2851e-11, 1.1364e-10, 2.9929e-13, 1.3659e-11, 9.7027e-14,\n 1.4386e-11, 8.8877e-13, 8.3707e-11, 1.0545e-12, 4.3594e-12, 2.8653e-13,\n 7.8511e-13, 2.5977e-11, 7.0402e-16, 1.0098e-12, 8.8536e-12, 5.5622e-12,\n 2.5512e-12, 8.1338e-12, 1.7832e-11, 5.7583e-12, 1.7500e-11, 1.0162e-12,\n 2.1618e-12, 5.2277e-11, 2.0231e-13, 9.8465e-12, 2.5800e-11, 2.8825e-11,\n 3.0533e-12, 3.6857e-12, 1.6759e-13, 3.2632e-11, 1.3117e-12, 6.8141e-11,\n 7.9564e-14, 7.5297e-12, 8.4702e-12, 3.9836e-12, 1.6289e-11, 2.5989e-11,\n 7.2750e-12, 1.1007e-11, 4.7434e-12, 2.1002e-11, 1.6529e-11, 4.5184e-12,\n 6.7395e-13, 5.6244e-14, 2.7756e-11, 3.4398e-12, 2.5473e-12, 2.3198e-11,\n 1.9107e-12, 5.8781e-13, 7.6221e-11, 1.2223e-12, 2.4039e-12, 8.3100e-14,\n 8.3892e-16, 2.5367e-12, 8.1515e-12, 2.1474e-12, 6.4814e-11, 1.3938e-12,\n 2.4542e-11, 5.5510e-11, 5.1320e-12, 1.2880e-11, 4.2488e-12, 1.6749e-12,\n 7.2756e-13, 1.2870e-11, 7.7191e-12, 4.4262e-12, 7.2851e-12, 1.0160e-12,\n 3.7101e-12, 4.0345e-12, 2.5730e-12, 5.5181e-13, 3.5068e-11, 4.4474e-13,\n 6.8119e-13, 2.8753e-11, 1.5980e-12, 7.0007e-12, 2.6739e-12, 2.0930e-12,\n 1.0492e-10, 5.0731e-12, 3.6473e-11, 2.3785e-12, 5.0099e-13, 1.4207e-10,\n 1.8123e-11, 6.5164e-13, 4.7694e-13, 1.8496e-11, 4.1208e-15, 3.1593e-12,\n 3.6330e-11, 2.9122e-11, 1.7786e-13, 2.5210e-12, 1.3638e-11, 1.5864e-12,\n 2.6609e-11, 1.4127e-11, 3.4335e-12, 2.6931e-12, 9.5511e-12, 8.0765e-14,\n 7.9081e-13, 1.2941e-12, 4.2906e-13, 2.8804e-11, 1.7371e-12, 1.3982e-12,\n 3.5023e-12, 3.5601e-12, 5.6386e-12, 4.6609e-12, 2.2291e-12, 2.4865e-11,\n 4.5483e-13, 9.2786e-11, 1.9414e-14, 1.0120e-12, 6.6343e-13, 2.3220e-12,\n 6.3974e-14, 3.1198e-12, 2.3792e-11, 1.3209e-11, 4.7651e-13, 3.2238e-12,\n 2.8018e-14, 9.0746e-13, 2.7184e-13, 7.0493e-12, 3.6847e-12, 4.9255e-12,\n 3.0911e-12, 6.0458e-13, 2.0842e-11, 1.2759e-11, 8.6891e-13, 7.2465e-14,\n 7.7692e-12, 1.6348e-11, 2.9469e-14, 2.4112e-12, 3.3939e-12, 2.0863e-11,\n 7.9514e-13, 1.9575e-13, 2.9000e-12, 7.0354e-13, 2.6198e-12, 1.0132e-13,\n 4.8448e-12, 1.9972e-12, 1.7503e-11, 2.1070e-12, 7.6691e-13, 5.5358e-12,\n 6.0560e-11, 2.7921e-12, 5.4803e-12, 3.3723e-13, 3.7292e-11, 1.0099e-13,\n 9.6197e-12, 1.3410e-14, 1.4026e-11, 4.8940e-12, 1.2014e-12, 7.7915e-12,\n 7.4482e-11, 2.2428e-12, 4.2812e-12, 2.1836e-11, 1.3815e-10, 1.8123e-12,\n 1.8576e-12, 8.0104e-13, 1.3213e-10, 1.5174e-13, 6.6309e-13, 1.8447e-10,\n 2.9563e-13, 2.1325e-11, 6.4421e-11, 8.4506e-13, 1.7039e-12, 3.4314e-11,\n 3.1699e-12, 1.3728e-12, 3.4108e-11, 7.6055e-12, 3.7477e-13, 1.8525e-12,\n 9.8873e-12, 1.4587e-11, 1.7873e-12, 1.8545e-12, 1.0156e-12, 9.6140e-13,\n 4.7545e-12, 5.6392e-12, 1.0308e-12, 8.6286e-11, 7.9691e-12, 4.2131e-13,\n 6.2941e-12, 3.7936e-11, 2.9666e-11, 3.8387e-11, 9.4171e-13, 3.3925e-12,\n 7.0121e-13, 1.0920e-11, 1.8017e-13, 1.1953e-13], device='cuda:0')" + "exp_avg_sq": "tensor([1.1272e-15, 4.8616e-14, 1.0105e-11, 1.0714e-13, 9.6246e-13, 8.1719e-12,\n 4.6711e-12, 8.3882e-13, 1.7493e-12, 6.3115e-13, 1.0085e-11, 2.4731e-12,\n 1.1290e-12, 1.6874e-12, 1.6497e-14, 9.4127e-13, 1.1672e-12, 6.5764e-13,\n 2.0331e-13, 2.1860e-13, 3.1908e-12, 5.3960e-14, 2.6914e-13, 7.4482e-13,\n 1.6714e-12, 9.5024e-13, 9.3276e-13, 6.4758e-12, 1.7277e-12, 3.2576e-12,\n 4.8314e-12, 9.3874e-12, 3.2473e-11, 8.5524e-14, 3.9030e-12, 2.7726e-14,\n 4.1109e-12, 2.5397e-13, 2.3920e-11, 3.0134e-13, 1.2457e-12, 8.1877e-14,\n 2.2435e-13, 7.4230e-12, 2.0118e-16, 2.8855e-13, 2.5300e-12, 1.5894e-12,\n 7.2904e-13, 2.3243e-12, 5.0957e-12, 1.6455e-12, 5.0008e-12, 2.9039e-13,\n 6.1776e-13, 1.4939e-11, 5.7811e-14, 2.8137e-12, 7.3725e-12, 8.2368e-12,\n 8.7251e-13, 1.0532e-12, 4.7889e-14, 9.3248e-12, 3.7482e-13, 1.9472e-11,\n 2.2736e-14, 2.1517e-12, 2.4204e-12, 1.1384e-12, 4.6547e-12, 7.4265e-12,\n 2.0789e-12, 3.1454e-12, 1.3555e-12, 6.0014e-12, 4.7234e-12, 1.2912e-12,\n 1.9259e-13, 1.6072e-14, 7.9314e-12, 9.8296e-13, 7.2790e-13, 6.6290e-12,\n 5.4600e-13, 1.6797e-13, 2.1781e-11, 3.4928e-13, 6.8693e-13, 2.3747e-14,\n 2.3973e-16, 7.2489e-13, 2.3294e-12, 6.1363e-13, 1.8521e-11, 3.9830e-13,\n 7.0132e-12, 1.5862e-11, 1.4665e-12, 3.6807e-12, 1.2141e-12, 4.7863e-13,\n 2.0790e-13, 3.6778e-12, 2.2058e-12, 1.2648e-12, 2.0818e-12, 2.9033e-13,\n 1.0602e-12, 1.1529e-12, 7.3527e-13, 1.5768e-13, 1.0021e-11, 1.2709e-13,\n 1.9466e-13, 8.2164e-12, 4.5665e-13, 2.0005e-12, 7.6408e-13, 5.9808e-13,\n 2.9980e-11, 1.4497e-12, 1.0422e-11, 6.7966e-13, 1.4316e-13, 4.0599e-11,\n 5.1788e-12, 1.8621e-13, 1.3629e-13, 5.2853e-12, 1.1776e-15, 9.0280e-13,\n 1.0382e-11, 8.3219e-12, 5.0824e-14, 7.2038e-13, 3.8972e-12, 4.5332e-13,\n 7.6038e-12, 4.0370e-12, 9.8116e-13, 7.6956e-13, 2.7293e-12, 2.3079e-14,\n 2.2598e-13, 3.6979e-13, 1.2261e-13, 8.2311e-12, 4.9638e-13, 3.9956e-13,\n 1.0008e-12, 1.0173e-12, 1.6113e-12, 1.3319e-12, 6.3698e-13, 7.1055e-12,\n 1.2997e-13, 2.6514e-11, 5.5478e-15, 2.8919e-13, 1.8958e-13, 6.6353e-13,\n 1.8281e-14, 8.9152e-13, 6.7987e-12, 3.7747e-12, 1.3617e-13, 9.2122e-13,\n 8.0063e-15, 2.5931e-13, 7.7681e-14, 2.0144e-12, 1.0529e-12, 1.4075e-12,\n 8.8331e-13, 1.7276e-13, 5.9558e-12, 3.6459e-12, 2.4830e-13, 2.0707e-14,\n 2.2201e-12, 4.6714e-12, 8.4211e-15, 6.8903e-13, 9.6985e-13, 5.9617e-12,\n 2.2722e-13, 5.5938e-14, 8.2871e-13, 2.0104e-13, 7.4862e-13, 2.8953e-14,\n 1.3845e-12, 5.7072e-13, 5.0016e-12, 6.0210e-13, 2.1915e-13, 1.5819e-12,\n 1.7305e-11, 7.9787e-13, 1.5660e-12, 9.6367e-14, 1.0656e-11, 2.8859e-14,\n 2.7489e-12, 3.8319e-15, 4.0081e-12, 1.3985e-12, 3.4330e-13, 2.2265e-12,\n 2.1284e-11, 6.4090e-13, 1.2234e-12, 6.2399e-12, 3.9476e-11, 5.1788e-13,\n 5.3084e-13, 2.2890e-13, 3.7757e-11, 4.3360e-14, 1.8948e-13, 5.2715e-11,\n 8.4478e-14, 6.0939e-12, 1.8409e-11, 2.4148e-13, 4.8692e-13, 9.8056e-12,\n 9.0582e-13, 3.9229e-13, 9.7465e-12, 2.1733e-12, 1.0709e-13, 5.2936e-13,\n 2.8254e-12, 4.1683e-12, 5.1074e-13, 5.2993e-13, 2.9021e-13, 2.7473e-13,\n 1.3586e-12, 1.6114e-12, 2.9457e-13, 2.4657e-11, 2.2772e-12, 1.2039e-13,\n 1.7986e-12, 1.0841e-11, 8.4774e-12, 1.0970e-11, 2.6910e-13, 9.6942e-13,\n 2.0038e-13, 3.1204e-12, 5.1486e-14, 3.4156e-14], device='cuda:0')" }, "51": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([ 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([4.0085e-14, 1.9896e-13, 6.3540e-11, 2.1188e-13, 1.0350e-11, 3.1923e-11,\n 1.6062e-11, 4.1986e-12, 2.0162e-12, 3.9005e-12, 4.5847e-11, 1.1898e-11,\n 2.4898e-12, 9.4389e-12, 7.3258e-14, 2.8766e-12, 7.0997e-12, 3.3496e-12,\n 7.1951e-13, 1.3068e-12, 1.6906e-11, 3.5412e-13, 1.9017e-12, 5.5879e-12,\n 8.9230e-12, 4.6037e-12, 6.9851e-12, 1.3850e-11, 4.7247e-12, 2.4627e-11,\n 3.3518e-11, 3.8740e-11, 2.1753e-10, 6.3721e-13, 1.0240e-11, 2.2075e-13,\n 1.8404e-11, 1.0046e-12, 7.2519e-11, 2.3052e-12, 1.1353e-11, 3.5135e-13,\n 1.2879e-12, 4.6827e-11, 2.1746e-14, 2.1334e-12, 1.0096e-11, 7.5828e-12,\n 1.7853e-12, 1.3663e-11, 2.4683e-11, 6.9277e-12, 2.5599e-11, 2.1567e-12,\n 2.7751e-12, 5.6628e-11, 3.5789e-13, 8.7624e-12, 4.6061e-11, 5.8191e-11,\n 7.5362e-12, 6.2666e-12, 3.2223e-13, 4.6420e-11, 2.5206e-12, 6.0031e-11,\n 9.6485e-14, 1.5275e-11, 9.3053e-12, 9.0184e-12, 1.9217e-11, 3.0065e-11,\n 7.1173e-12, 2.2132e-11, 1.0586e-11, 3.5385e-11, 2.1528e-11, 8.7994e-12,\n 1.4174e-12, 5.5902e-14, 4.3463e-11, 3.2915e-12, 2.7927e-12, 1.7783e-11,\n 3.6524e-12, 7.5615e-13, 9.5006e-11, 3.0006e-12, 4.1401e-12, 6.3401e-14,\n 5.5399e-14, 3.7052e-12, 1.3467e-11, 3.2279e-12, 8.0151e-11, 2.3562e-12,\n 4.7693e-11, 6.5444e-11, 8.6732e-12, 7.2659e-12, 6.5077e-12, 2.9720e-12,\n 1.3360e-12, 2.0173e-11, 1.2261e-11, 7.9687e-12, 1.2155e-11, 1.3028e-12,\n 7.0867e-12, 9.0318e-12, 4.7541e-12, 1.1375e-12, 5.1654e-11, 1.1918e-12,\n 1.8711e-12, 5.4707e-11, 2.1565e-12, 1.0726e-11, 2.8754e-12, 3.7087e-12,\n 1.1534e-10, 1.0239e-11, 4.0964e-11, 2.8234e-12, 6.5787e-13, 1.2725e-10,\n 1.5988e-11, 1.0602e-12, 6.1938e-13, 1.9822e-11, 1.3761e-14, 4.5356e-12,\n 6.5283e-11, 4.1253e-11, 2.5504e-13, 4.5234e-12, 1.1883e-11, 3.3010e-12,\n 3.7629e-11, 2.7906e-11, 4.9464e-12, 4.6959e-12, 1.8779e-11, 3.3781e-13,\n 1.7459e-12, 2.3829e-12, 1.3694e-12, 2.3795e-11, 1.5385e-12, 2.4355e-12,\n 3.4555e-12, 3.5687e-12, 9.4784e-12, 6.5100e-12, 3.2782e-12, 3.3078e-11,\n 7.9779e-13, 1.0371e-10, 3.9868e-14, 2.2999e-12, 9.8232e-13, 3.1593e-12,\n 1.4712e-13, 5.7708e-12, 3.0069e-11, 1.9207e-11, 1.2066e-12, 6.6168e-12,\n 1.0036e-13, 1.2355e-12, 4.1535e-13, 9.2143e-12, 7.0252e-12, 9.2337e-12,\n 4.8243e-12, 1.2716e-12, 3.5453e-11, 2.1932e-11, 1.0855e-12, 9.8796e-14,\n 6.0680e-12, 1.6950e-11, 5.7725e-14, 3.5970e-12, 2.9387e-12, 4.7509e-11,\n 9.1291e-13, 3.6445e-13, 2.9615e-12, 1.1935e-12, 5.6111e-12, 3.1009e-13,\n 9.9820e-12, 3.9080e-12, 2.3213e-11, 3.6136e-12, 1.7276e-12, 9.9397e-12,\n 4.6262e-11, 5.5634e-12, 7.4620e-12, 6.2364e-13, 7.8575e-11, 2.4546e-13,\n 9.3465e-12, 1.8893e-14, 1.1978e-11, 1.0655e-11, 1.3060e-12, 8.0197e-12,\n 6.7673e-11, 3.1013e-12, 5.4417e-12, 5.9721e-11, 2.2024e-10, 3.3319e-12,\n 2.3911e-12, 2.0263e-12, 1.1678e-10, 2.3709e-13, 1.9108e-12, 2.2420e-10,\n 4.0373e-13, 1.7079e-11, 8.0098e-11, 1.1628e-12, 2.5577e-12, 3.6156e-11,\n 9.3960e-12, 2.7998e-12, 4.7994e-11, 1.0147e-11, 4.7224e-13, 3.9185e-12,\n 1.7163e-11, 1.6187e-11, 2.7015e-12, 3.5302e-12, 1.8195e-12, 1.1401e-12,\n 1.0332e-11, 8.0481e-12, 1.3956e-12, 1.0014e-10, 1.4000e-11, 5.9840e-13,\n 1.1506e-11, 4.6302e-11, 5.7676e-11, 5.2285e-11, 1.6572e-12, 5.5282e-12,\n 1.9896e-12, 2.1313e-11, 2.5921e-13, 8.9746e-14], device='cuda:0')" + "exp_avg_sq": "tensor([1.1455e-14, 5.6854e-14, 1.8157e-11, 6.0547e-14, 2.9576e-12, 9.1221e-12,\n 4.5897e-12, 1.1998e-12, 5.7614e-13, 1.1146e-12, 1.3101e-11, 3.4001e-12,\n 7.1147e-13, 2.6972e-12, 2.0934e-14, 8.2203e-13, 2.0288e-12, 9.5717e-13,\n 2.0561e-13, 3.7344e-13, 4.8311e-12, 1.0119e-13, 5.4341e-13, 1.5968e-12,\n 2.5498e-12, 1.3155e-12, 1.9961e-12, 3.9577e-12, 1.3501e-12, 7.0374e-12,\n 9.5780e-12, 1.1070e-11, 6.2162e-11, 1.8209e-13, 2.9262e-12, 6.3082e-14,\n 5.2592e-12, 2.8707e-13, 2.0723e-11, 6.5874e-13, 3.2442e-12, 1.0040e-13,\n 3.6803e-13, 1.3381e-11, 6.2140e-15, 6.0965e-13, 2.8851e-12, 2.1668e-12,\n 5.1016e-13, 3.9043e-12, 7.0533e-12, 1.9796e-12, 7.3151e-12, 6.1630e-13,\n 7.9300e-13, 1.6182e-11, 1.0227e-13, 2.5039e-12, 1.3162e-11, 1.6629e-11,\n 2.1535e-12, 1.7907e-12, 9.2079e-14, 1.3265e-11, 7.2028e-13, 1.7154e-11,\n 2.7571e-14, 4.3648e-12, 2.6591e-12, 2.5771e-12, 5.4913e-12, 8.5914e-12,\n 2.0338e-12, 6.3245e-12, 3.0251e-12, 1.0112e-11, 6.1518e-12, 2.5145e-12,\n 4.0502e-13, 1.5975e-14, 1.2420e-11, 9.4057e-13, 7.9803e-13, 5.0817e-12,\n 1.0437e-12, 2.1608e-13, 2.7149e-11, 8.5745e-13, 1.1831e-12, 1.8117e-14,\n 1.5831e-14, 1.0588e-12, 3.8484e-12, 9.2240e-13, 2.2904e-11, 6.7329e-13,\n 1.3629e-11, 1.8701e-11, 2.4784e-12, 2.0763e-12, 1.8596e-12, 8.4928e-13,\n 3.8177e-13, 5.7645e-12, 3.5036e-12, 2.2771e-12, 3.4735e-12, 3.7228e-13,\n 2.0251e-12, 2.5809e-12, 1.3585e-12, 3.2504e-13, 1.4760e-11, 3.4056e-13,\n 5.3469e-13, 1.5633e-11, 6.1624e-13, 3.0651e-12, 8.2166e-13, 1.0598e-12,\n 3.2960e-11, 2.9260e-12, 1.1706e-11, 8.0680e-13, 1.8799e-13, 3.6363e-11,\n 4.5687e-12, 3.0295e-13, 1.7699e-13, 5.6642e-12, 3.9323e-15, 1.2961e-12,\n 1.8655e-11, 1.1788e-11, 7.2880e-14, 1.2926e-12, 3.3957e-12, 9.4328e-13,\n 1.0753e-11, 7.9743e-12, 1.4135e-12, 1.3419e-12, 5.3663e-12, 9.6533e-14,\n 4.9889e-13, 6.8095e-13, 3.9130e-13, 6.7995e-12, 4.3963e-13, 6.9596e-13,\n 9.8743e-13, 1.0198e-12, 2.7085e-12, 1.8603e-12, 9.3677e-13, 9.4523e-12,\n 2.2797e-13, 2.9635e-11, 1.1393e-14, 6.5721e-13, 2.8071e-13, 9.0278e-13,\n 4.2041e-14, 1.6490e-12, 8.5924e-12, 5.4885e-12, 3.4480e-13, 1.8908e-12,\n 2.8678e-14, 3.5304e-13, 1.1869e-13, 2.6331e-12, 2.0075e-12, 2.6386e-12,\n 1.3786e-12, 3.6338e-13, 1.0131e-11, 6.2671e-12, 3.1020e-13, 2.8232e-14,\n 1.7340e-12, 4.8436e-12, 1.6495e-14, 1.0279e-12, 8.3975e-13, 1.3576e-11,\n 2.6087e-13, 1.0414e-13, 8.4627e-13, 3.4105e-13, 1.6034e-12, 8.8611e-14,\n 2.8524e-12, 1.1167e-12, 6.6332e-12, 1.0326e-12, 4.9367e-13, 2.8403e-12,\n 1.3220e-11, 1.5898e-12, 2.1323e-12, 1.7821e-13, 2.2453e-11, 7.0142e-14,\n 2.6708e-12, 5.3987e-15, 3.4229e-12, 3.0447e-12, 3.7320e-13, 2.2917e-12,\n 1.9338e-11, 8.8623e-13, 1.5550e-12, 1.7066e-11, 6.2936e-11, 9.5213e-13,\n 6.8327e-13, 5.7904e-13, 3.3370e-11, 6.7751e-14, 5.4603e-13, 6.4067e-11,\n 1.1537e-13, 4.8804e-12, 2.2889e-11, 3.3228e-13, 7.3088e-13, 1.0332e-11,\n 2.6850e-12, 8.0007e-13, 1.3715e-11, 2.8995e-12, 1.3495e-13, 1.1198e-12,\n 4.9045e-12, 4.6254e-12, 7.7198e-13, 1.0088e-12, 5.1995e-13, 3.2581e-13,\n 2.9525e-12, 2.2998e-12, 3.9882e-13, 2.8617e-11, 4.0006e-12, 1.7100e-13,\n 3.2881e-12, 1.3231e-11, 1.6481e-11, 1.4941e-11, 4.7355e-13, 1.5797e-12,\n 5.6854e-13, 6.0903e-12, 7.4071e-14, 2.5646e-14], device='cuda:0')" }, "52": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[-5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, -5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [-5.6052e-45, -5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n -5.6052e-45, -5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.0017e-14, 7.8582e-15, 1.4592e-14, ..., 9.4973e-16, 4.5026e-16,\n 7.1545e-14],\n [2.2888e-15, 6.7599e-15, 2.2422e-14, ..., 8.8568e-14, 5.2899e-15,\n 1.1728e-14],\n [1.3947e-14, 5.0148e-16, 7.4101e-15, ..., 4.2796e-14, 4.4201e-16,\n 6.2229e-16],\n ...,\n [1.9419e-11, 3.1606e-10, 4.2962e-10, ..., 7.3530e-10, 3.7532e-10,\n 1.8565e-10],\n [5.7855e-12, 9.5852e-11, 1.3089e-10, ..., 2.1153e-10, 1.1913e-10,\n 6.1779e-11],\n [2.3785e-12, 2.7955e-11, 4.1566e-11, ..., 6.8377e-11, 2.9529e-11,\n 1.4632e-11]], device='cuda:0')" + "exp_avg_sq": "tensor([[2.8626e-15, 2.2455e-15, 4.1699e-15, ..., 2.7139e-16, 1.2867e-16,\n 2.0444e-14],\n [6.5405e-16, 1.9317e-15, 6.4072e-15, ..., 2.5309e-14, 1.5116e-15,\n 3.3513e-15],\n [3.9856e-15, 1.4330e-16, 2.1175e-15, ..., 1.2229e-14, 1.2631e-16,\n 1.7782e-16],\n ...,\n [5.5491e-12, 9.0316e-11, 1.2277e-10, ..., 2.1012e-10, 1.0725e-10,\n 5.3050e-11],\n [1.6532e-12, 2.7390e-11, 3.7404e-11, ..., 6.0448e-11, 3.4043e-11,\n 1.7654e-11],\n [6.7967e-13, 7.9884e-12, 1.1878e-11, ..., 1.9539e-11, 8.4381e-12,\n 4.1811e-12]], device='cuda:0')" }, "53": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([7.8881e-14, 6.3152e-14, 1.8977e-13, 2.3054e-14, 9.4219e-14, 1.5038e-13,\n 4.2869e-14, 1.8302e-13, 4.4785e-13, 1.5137e-12, 5.4721e-14, 2.2723e-14,\n 6.3096e-13, 1.7238e-13, 1.1678e-12, 6.8947e-16, 1.1747e-13, 2.6376e-14,\n 3.0992e-13, 8.5476e-13, 7.3031e-14, 1.9072e-16, 3.2698e-15, 1.8174e-14,\n 1.1050e-13, 7.6735e-14, 1.6513e-13, 2.3113e-13, 2.1382e-14, 1.7573e-13,\n 4.8293e-13, 5.5988e-14, 3.3364e-13, 5.2803e-13, 1.1853e-12, 7.2859e-13,\n 1.8720e-13, 1.9919e-13, 3.6595e-13, 8.2147e-14, 1.4604e-14, 2.5379e-13,\n 2.6537e-13, 1.2814e-15, 1.6022e-12, 7.6221e-13, 7.2534e-15, 2.2835e-15,\n 8.6447e-14, 2.7267e-13, 2.9742e-13, 1.8446e-13, 2.1107e-13, 9.5885e-14,\n 8.7379e-13, 3.2639e-13, 4.8503e-14, 9.0398e-13, 6.9266e-15, 1.7224e-13,\n 2.5514e-13, 6.4299e-14, 5.3708e-15, 1.6619e-12, 4.1839e-14, 9.9817e-13,\n 6.1899e-15, 1.0451e-12, 1.9340e-13, 1.9650e-12, 2.3198e-12, 5.6223e-14,\n 5.5957e-13, 1.5452e-12, 1.1014e-12, 1.0241e-13, 2.7244e-12, 1.0432e-12,\n 3.8233e-12, 1.0297e-14, 1.0914e-13, 3.2736e-13, 1.2960e-13, 4.3734e-13,\n 1.3825e-12, 5.0885e-14, 1.0296e-13, 3.3356e-13, 3.9776e-12, 9.5502e-14,\n 7.5428e-13, 5.3684e-15, 3.1300e-12, 3.0092e-12, 5.1618e-13, 4.3518e-13,\n 3.4397e-15, 5.5677e-13, 5.7189e-13, 4.3511e-13, 1.2684e-12, 3.0959e-15,\n 1.6421e-14, 6.4785e-15, 3.3432e-15, 5.2926e-14, 1.3468e-14, 7.7023e-14,\n 3.1301e-13, 9.9282e-13, 1.5244e-13, 7.9610e-14, 4.5644e-14, 2.5114e-13,\n 6.9103e-14, 9.6152e-13, 3.5900e-13, 7.0068e-13, 1.0938e-12, 3.6962e-13,\n 1.4837e-14, 3.5075e-15, 6.9838e-13, 9.4324e-14, 6.4651e-14, 3.7231e-14,\n 1.2123e-12, 1.9558e-13, 3.0082e-14, 1.1104e-12, 4.7382e-14, 1.3405e-13,\n 1.8978e-13, 7.3520e-13, 1.3907e-12, 1.2823e-14, 5.3291e-14, 1.0583e-13,\n 7.5930e-14, 1.9925e-15, 3.1427e-13, 8.7406e-14, 2.5568e-14, 1.1490e-15,\n 2.5335e-13, 4.0270e-13, 6.0054e-13, 2.6327e-13, 4.6681e-14, 1.9170e-13,\n 4.4947e-15, 6.3730e-13, 6.6222e-15, 1.9210e-14, 9.5676e-13, 4.7198e-13,\n 1.5384e-15, 9.4925e-16, 6.3416e-14, 4.2448e-13, 1.8915e-12, 1.1733e-13,\n 1.1364e-13, 6.0120e-14, 1.6646e-13, 1.9633e-13, 1.6038e-13, 3.1349e-13,\n 3.6340e-15, 3.9254e-14, 4.4693e-13, 9.0008e-14, 1.1346e-13, 6.6382e-13,\n 4.1716e-16, 4.7976e-13, 6.2869e-13, 5.7207e-13, 4.8554e-15, 3.1855e-14,\n 3.6443e-14, 2.3160e-13, 5.9311e-16, 1.5005e-14, 1.0812e-13, 1.2628e-12,\n 1.2067e-13, 1.9709e-14, 5.3055e-13, 6.0151e-14, 5.1844e-14, 4.3504e-13,\n 3.0636e-13, 3.6488e-13, 1.9835e-13, 5.6412e-13, 6.8088e-16, 4.1053e-14,\n 3.0615e-13, 3.4820e-13, 5.7890e-15, 8.9699e-13, 4.2852e-14, 5.1847e-13,\n 2.8156e-14, 2.2078e-14, 8.6286e-15, 1.6982e-13, 1.0927e-12, 3.5052e-14,\n 1.7854e-13, 2.8843e-13, 6.9098e-13, 7.3444e-14, 3.8175e-15, 8.2270e-14,\n 4.7991e-14, 4.6594e-14, 2.3304e-14, 5.2066e-14, 1.3280e-14, 3.6116e-12,\n 9.5892e-13, 5.9606e-13, 1.8000e-15, 1.9781e-14, 1.9741e-13, 6.3775e-13,\n 1.8912e-12, 6.0356e-13, 2.1282e-14, 1.1591e-13, 6.7058e-13, 5.8677e-15,\n 3.8415e-13, 2.9286e-13, 7.9222e-13, 1.7309e-13, 1.2496e-13, 2.5783e-14,\n 5.1645e-14, 2.2981e-14, 6.1061e-13, 1.5739e-13, 1.5153e-13, 1.0705e-12,\n 1.8871e-13, 1.8069e-15, 4.1283e-13, 4.1875e-13, 1.7457e-12, 3.5813e-13,\n 5.1397e-13, 2.8108e-13, 4.4070e-16, 2.2262e-13, 1.2059e-27, 2.9533e-29,\n 1.8786e-28, 3.4172e-29, 1.9300e-28, 5.7458e-29, 3.1024e-28, 7.4796e-30,\n 1.6943e-28, 2.6350e-28, 2.4658e-28, 6.3341e-30, 1.6009e-31, 1.3458e-29,\n 7.5691e-29, 5.6567e-30, 1.3143e-28, 1.0067e-29, 1.1986e-28, 2.8843e-29,\n 1.8533e-30, 9.2954e-29, 5.2572e-30, 1.0832e-28, 1.5336e-30, 9.2693e-29,\n 2.0479e-28, 8.8563e-30, 5.3455e-28, 1.8232e-28, 2.9657e-29, 4.2721e-29,\n 5.4718e-29, 3.2744e-28, 3.0566e-29, 2.5194e-29, 1.0241e-29, 5.5503e-29,\n 1.4950e-28, 2.9559e-29, 5.2424e-30, 6.0785e-29, 1.2225e-29, 2.7036e-29,\n 2.5552e-29, 2.3045e-28, 9.4324e-29, 3.1600e-29, 3.3879e-29, 2.1037e-29,\n 4.2117e-29, 1.3955e-28, 1.4076e-28, 1.7641e-28, 7.1848e-29, 1.6523e-29,\n 1.4690e-28, 2.5285e-28, 3.4655e-28, 1.5126e-28, 1.2217e-28, 8.4675e-29,\n 5.8254e-30, 5.2432e-29, 3.9511e-31, 2.0464e-28, 5.4184e-28, 1.6891e-28,\n 4.5634e-30, 2.7119e-29, 1.6489e-28, 5.5138e-29, 8.9425e-29, 5.3785e-30,\n 4.6064e-29, 2.4913e-29, 8.7643e-29, 2.4689e-28, 8.5054e-28, 3.1158e-28,\n 3.3752e-29, 2.5570e-28, 3.0974e-29, 4.1328e-28, 5.0564e-29, 4.6668e-29,\n 4.1546e-28, 1.4445e-28, 7.4850e-29, 9.5776e-30, 1.6729e-29, 9.1067e-29,\n 9.5516e-29, 2.0099e-29, 7.7656e-29, 5.5221e-30, 7.3314e-30, 2.4000e-28,\n 1.2608e-29, 7.2246e-29, 1.0817e-28, 7.5457e-30, 6.5915e-29, 6.2286e-29,\n 2.5000e-29, 3.4303e-29, 2.1408e-29, 3.5735e-28, 1.4515e-28, 2.5846e-29,\n 1.9768e-29, 1.0513e-28, 1.1384e-28, 2.1380e-28, 9.0044e-29, 4.2135e-28,\n 1.1571e-28, 1.5676e-28, 3.6452e-28, 4.4757e-28, 2.7082e-28, 1.4265e-28,\n 1.6829e-28, 8.2413e-29, 1.8870e-29, 1.6904e-29, 9.3875e-28, 1.1028e-30,\n 1.7832e-28, 1.0616e-29, 4.1707e-29, 5.0833e-29, 1.3087e-28, 4.2289e-29,\n 1.1091e-29, 1.2456e-32, 2.8480e-30, 1.6753e-29, 3.7604e-29, 9.7723e-29,\n 9.2936e-30, 2.2078e-28, 8.4587e-28, 9.5487e-29, 2.7320e-28, 3.6403e-29,\n 1.4937e-28, 4.1200e-28, 1.8984e-28, 1.2722e-29, 1.3700e-29, 5.5483e-28,\n 9.5530e-29, 1.0740e-28, 4.9810e-30, 6.8336e-29, 2.3657e-28, 7.7465e-29,\n 4.0592e-31, 1.0684e-28, 3.1446e-29, 3.0608e-29, 1.7375e-28, 1.6061e-29,\n 1.1227e-29, 9.7449e-29, 6.4443e-29, 4.1266e-30, 3.7941e-29, 4.4174e-28,\n 2.6089e-29, 9.0657e-28, 4.6126e-28, 6.4994e-30, 1.5028e-28, 4.7655e-28,\n 4.7156e-28, 1.7825e-29, 2.2368e-29, 3.9071e-29, 2.7428e-29, 1.1799e-28,\n 2.3370e-30, 1.1898e-28, 6.0865e-29, 1.5310e-29, 5.5554e-29, 1.0276e-30,\n 1.9238e-28, 1.1782e-30, 2.3662e-28, 1.3208e-28, 1.4571e-28, 3.3100e-28,\n 3.0172e-28, 1.0937e-29, 6.1652e-28, 5.5999e-29, 1.4270e-28, 3.0816e-29,\n 9.1167e-31, 5.1665e-29, 7.1095e-29, 1.1859e-28, 1.3693e-29, 4.2502e-29,\n 2.7936e-30, 6.4468e-29, 9.6965e-29, 4.2475e-30, 1.9012e-28, 3.6103e-30,\n 1.0901e-28, 4.0638e-29, 2.0797e-28, 9.5189e-29, 5.4217e-30, 4.3073e-29,\n 1.0285e-28, 2.8053e-29, 3.1106e-29, 5.6677e-29, 1.1090e-28, 6.5107e-29,\n 8.5778e-29, 2.4811e-30, 1.1116e-28, 1.1952e-28, 3.1926e-28, 3.2411e-28,\n 8.0598e-30, 5.9329e-29, 8.4628e-30, 5.6572e-29, 1.2108e-30, 1.5605e-29,\n 1.8729e-29, 9.9933e-29, 3.4385e-30, 5.0681e-30, 2.2277e-29, 2.4897e-28,\n 1.7774e-29, 5.1055e-29, 1.3583e-29, 1.7013e-28, 3.5099e-28, 3.6544e-32,\n 3.3407e-28, 2.3512e-28, 3.2173e-28, 1.5142e-30, 5.2955e-29, 3.6289e-28,\n 1.5986e-29, 7.1608e-30, 4.6065e-10, 6.9577e-10, 7.3203e-12, 5.8465e-09,\n 1.5907e-09, 3.9905e-10, 1.0059e-09, 2.8342e-09, 5.6410e-11, 7.0456e-10,\n 4.6601e-09, 1.1058e-11, 3.2767e-10, 1.1874e-09, 5.8934e-09, 1.3348e-10,\n 1.7447e-09, 5.8619e-09, 4.0275e-10, 6.7440e-10, 1.8774e-09, 3.3208e-12,\n 2.4690e-09, 7.5427e-09, 4.7542e-10, 8.2207e-09, 1.7392e-10, 1.3149e-09,\n 1.0810e-09, 4.9524e-09, 1.0573e-09, 1.1469e-11, 1.8648e-10, 1.2138e-11,\n 2.2710e-09, 3.1490e-09, 3.1549e-09, 4.8447e-10, 9.2270e-11, 3.0847e-09,\n 5.8501e-09, 2.9384e-09, 3.4777e-09, 2.7137e-09, 1.1788e-09, 7.7262e-12,\n 2.1123e-09, 6.0154e-10, 2.8835e-09, 2.0549e-09, 1.1501e-08, 2.2368e-09,\n 3.8118e-10, 1.7719e-11, 9.8428e-10, 9.8174e-10, 6.3035e-09, 5.3795e-10,\n 1.5077e-09, 1.5234e-09, 1.5621e-09, 5.9962e-10, 5.7680e-10, 4.2371e-10,\n 1.6382e-10, 1.3198e-11, 3.4268e-09, 1.2779e-09, 1.7378e-09, 2.2539e-09,\n 1.6295e-10, 5.2861e-11, 3.2192e-11, 7.6448e-09, 2.4748e-10, 1.0440e-08,\n 2.0000e-09, 5.7968e-09, 2.6598e-09, 2.6915e-10, 1.6855e-09, 1.9677e-08,\n 6.8846e-10, 4.3500e-09, 1.5803e-10, 5.9127e-09, 2.1364e-09, 1.3540e-10,\n 3.7929e-11, 2.3665e-09, 3.4421e-09, 1.6055e-09, 1.1620e-11, 3.2958e-09,\n 8.4400e-09, 4.1595e-10, 1.7773e-10, 3.4495e-09, 3.2476e-10, 3.2649e-10,\n 4.3707e-10, 2.0542e-09, 5.4998e-11, 2.6963e-10, 7.2852e-10, 7.2741e-09,\n 4.4020e-09, 6.2670e-10, 6.6686e-11, 7.8158e-10, 3.5236e-10, 1.1754e-09,\n 2.1964e-09, 1.2169e-10, 1.6865e-09, 2.3472e-09, 6.2072e-10, 2.9321e-10,\n 4.1434e-09, 8.1221e-11, 2.8912e-10, 1.5894e-08, 2.9077e-09, 1.0040e-08,\n 3.6226e-09, 1.2015e-09, 8.9331e-11, 1.4694e-10, 1.8154e-09, 2.0351e-09,\n 1.1355e-09, 1.6054e-09, 4.0184e-10, 6.7986e-12, 1.7374e-09, 6.4987e-10,\n 9.3318e-10, 4.3033e-09, 4.8560e-10, 9.4932e-11, 7.6813e-13, 3.5586e-09,\n 1.5403e-09, 3.5899e-11, 1.0581e-09, 8.5527e-10, 2.2316e-10, 7.6102e-09,\n 3.2499e-10, 5.8725e-12, 1.0048e-09, 1.1367e-08, 5.9032e-10, 4.8207e-10,\n 4.5078e-09, 1.3521e-10, 5.1234e-11, 8.5184e-10, 7.4108e-13, 1.4445e-09,\n 3.6361e-09, 5.1320e-09, 1.4837e-10, 9.7027e-10, 1.3523e-09, 2.4325e-09,\n 1.1545e-09, 5.6041e-10, 9.3139e-11, 6.0844e-09, 3.7516e-10, 7.8498e-09,\n 4.9659e-09, 3.7467e-10, 1.2649e-09, 3.2852e-10, 2.0277e-11, 4.1618e-10,\n 3.3939e-09, 1.3091e-11, 2.3606e-09, 6.7454e-09, 1.7815e-11, 1.6044e-09,\n 4.4573e-09, 4.1934e-09, 1.1337e-09, 6.5518e-10, 5.1178e-11, 4.3137e-09,\n 7.8639e-11, 1.4605e-09, 6.6196e-09, 6.8517e-09, 1.1926e-09, 2.0088e-09,\n 3.5838e-09, 8.6628e-09, 8.3651e-11, 3.8425e-09, 4.7748e-10, 2.4063e-10,\n 8.9216e-09, 6.1204e-10, 2.8275e-10, 1.0915e-09, 3.5383e-10, 6.4193e-10,\n 9.9752e-10, 3.2701e-10, 1.8837e-09, 1.4131e-09, 4.5348e-09, 3.5161e-10,\n 5.2451e-11, 1.4271e-08, 4.4924e-10, 8.4499e-09, 1.7143e-11, 9.2384e-10,\n 7.2292e-10, 3.2112e-09, 3.6661e-09, 2.1645e-09, 4.6861e-09, 1.6604e-10,\n 2.9713e-09, 3.9374e-10, 1.8311e-08, 1.1854e-10, 4.3243e-10, 1.8001e-09,\n 1.8395e-09, 1.0749e-09, 3.7574e-09, 1.3596e-09, 3.8129e-10, 3.3836e-10,\n 2.2952e-09, 1.3751e-11, 3.3432e-09, 1.2618e-10, 3.1017e-13, 6.2582e-10,\n 3.1831e-09, 4.2028e-11, 8.4115e-10, 8.1981e-11, 1.1073e-10, 7.4330e-10,\n 2.6609e-09, 1.9219e-09, 1.4463e-09, 6.2773e-09, 1.8970e-09, 5.5348e-10],\n device='cuda:0')" + "exp_avg_sq": "tensor([2.2541e-14, 1.8046e-14, 5.4227e-14, 6.5878e-15, 2.6924e-14, 4.2974e-14,\n 1.2250e-14, 5.2301e-14, 1.2798e-13, 4.3255e-13, 1.5637e-14, 6.4933e-15,\n 1.8030e-13, 4.9259e-14, 3.3371e-13, 1.9702e-16, 3.3567e-14, 7.5371e-15,\n 8.8562e-14, 2.4425e-13, 2.0869e-14, 5.4500e-17, 9.3438e-16, 5.1933e-15,\n 3.1578e-14, 2.1928e-14, 4.7186e-14, 6.6046e-14, 6.1100e-15, 5.0218e-14,\n 1.3800e-13, 1.5999e-14, 9.5342e-14, 1.5089e-13, 3.3871e-13, 2.0820e-13,\n 5.3494e-14, 5.6921e-14, 1.0457e-13, 2.3474e-14, 4.1733e-15, 7.2522e-14,\n 7.5832e-14, 3.6618e-16, 4.5785e-13, 2.1781e-13, 2.0727e-15, 6.5254e-16,\n 2.4703e-14, 7.7916e-14, 8.4991e-14, 5.2712e-14, 6.0315e-14, 2.7400e-14,\n 2.4969e-13, 9.3269e-14, 1.3860e-14, 2.5832e-13, 1.9793e-15, 4.9220e-14,\n 7.2907e-14, 1.8374e-14, 1.5348e-15, 4.7491e-13, 1.1956e-14, 2.8524e-13,\n 1.7688e-15, 2.9863e-13, 5.5267e-14, 5.6152e-13, 6.6292e-13, 1.6066e-14,\n 1.5990e-13, 4.4154e-13, 3.1474e-13, 2.9266e-14, 7.7851e-13, 2.9810e-13,\n 1.0925e-12, 2.9426e-15, 3.1189e-14, 9.3546e-14, 3.7034e-14, 1.2497e-13,\n 3.9507e-13, 1.4541e-14, 2.9422e-14, 9.5318e-14, 1.1366e-12, 2.7290e-14,\n 2.1554e-13, 1.5340e-15, 8.9443e-13, 8.5989e-13, 1.4750e-13, 1.2436e-13,\n 9.8292e-16, 1.5910e-13, 1.6342e-13, 1.2434e-13, 3.6247e-13, 8.8466e-16,\n 4.6925e-15, 1.8513e-15, 9.5535e-16, 1.5124e-14, 3.8485e-15, 2.2010e-14,\n 8.9445e-14, 2.8371e-13, 4.3562e-14, 2.2749e-14, 1.3043e-14, 7.1764e-14,\n 1.9747e-14, 2.7476e-13, 1.0259e-13, 2.0023e-13, 3.1255e-13, 1.0562e-13,\n 4.2397e-15, 1.0023e-15, 1.9957e-13, 2.6954e-14, 1.8475e-14, 1.0639e-14,\n 3.4644e-13, 5.5889e-14, 8.5963e-15, 3.1731e-13, 1.3540e-14, 3.8305e-14,\n 5.4232e-14, 2.1009e-13, 3.9741e-13, 3.6642e-15, 1.5228e-14, 3.0243e-14,\n 2.1698e-14, 5.6938e-16, 8.9804e-14, 2.4977e-14, 7.3063e-15, 3.2835e-16,\n 7.2395e-14, 1.1507e-13, 1.7161e-13, 7.5231e-14, 1.3339e-14, 5.4778e-14,\n 1.2844e-15, 1.8211e-13, 1.8923e-15, 5.4893e-15, 2.7340e-13, 1.3487e-13,\n 4.3960e-16, 2.7126e-16, 1.8122e-14, 1.2130e-13, 5.4050e-13, 3.3527e-14,\n 3.2475e-14, 1.7180e-14, 4.7567e-14, 5.6103e-14, 4.5830e-14, 8.9582e-14,\n 1.0385e-15, 1.1217e-14, 1.2771e-13, 2.5721e-14, 3.2422e-14, 1.8969e-13,\n 1.1921e-16, 1.3709e-13, 1.7965e-13, 1.6347e-13, 1.3875e-15, 9.1028e-15,\n 1.0414e-14, 6.6182e-14, 1.6949e-16, 4.2877e-15, 3.0896e-14, 3.6086e-13,\n 3.4481e-14, 5.6320e-15, 1.5161e-13, 1.7188e-14, 1.4815e-14, 1.2431e-13,\n 8.7545e-14, 1.0427e-13, 5.6681e-14, 1.6120e-13, 1.9457e-16, 1.1731e-14,\n 8.7484e-14, 9.9500e-14, 1.6542e-15, 2.5632e-13, 1.2245e-14, 1.4816e-13,\n 8.0458e-15, 6.3091e-15, 2.4657e-15, 4.8529e-14, 3.1224e-13, 1.0017e-14,\n 5.1021e-14, 8.2421e-14, 1.9745e-13, 2.0987e-14, 1.0909e-15, 2.3509e-14,\n 1.3714e-14, 1.3315e-14, 6.6592e-15, 1.4878e-14, 3.7948e-15, 1.0321e-12,\n 2.7402e-13, 1.7033e-13, 5.1437e-16, 5.6526e-15, 5.6413e-14, 1.8224e-13,\n 5.4043e-13, 1.7247e-13, 6.0816e-15, 3.3121e-14, 1.9162e-13, 1.6767e-15,\n 1.0978e-13, 8.3686e-14, 2.2638e-13, 4.9461e-14, 3.5708e-14, 7.3677e-15,\n 1.4758e-14, 6.5670e-15, 1.7449e-13, 4.4977e-14, 4.3300e-14, 3.0591e-13,\n 5.3924e-14, 5.1634e-16, 1.1797e-13, 1.1966e-13, 4.9884e-13, 1.0234e-13,\n 1.4687e-13, 8.0319e-14, 1.2593e-16, 6.3615e-14, 3.4460e-28, 8.4392e-30,\n 5.3681e-29, 9.7649e-30, 5.5152e-29, 1.6419e-29, 8.8654e-29, 2.1374e-30,\n 4.8417e-29, 7.5299e-29, 7.0462e-29, 1.8100e-30, 4.5747e-32, 3.8459e-30,\n 2.1629e-29, 1.6164e-30, 3.7556e-29, 2.8768e-30, 3.4252e-29, 8.2420e-30,\n 5.2959e-31, 2.6562e-29, 1.5023e-30, 3.0955e-29, 4.3825e-31, 2.6488e-29,\n 5.8520e-29, 2.5308e-30, 1.5275e-28, 5.2099e-29, 8.4749e-30, 1.2208e-29,\n 1.5636e-29, 9.3568e-29, 8.7345e-30, 7.1994e-30, 2.9265e-30, 1.5860e-29,\n 4.2721e-29, 8.4467e-30, 1.4980e-30, 1.7370e-29, 3.4934e-30, 7.7256e-30,\n 7.3016e-30, 6.5852e-29, 2.6954e-29, 9.0301e-30, 9.6811e-30, 6.0116e-30,\n 1.2035e-29, 3.9878e-29, 4.0222e-29, 5.0412e-29, 2.0531e-29, 4.7215e-30,\n 4.1977e-29, 7.2254e-29, 9.9030e-29, 4.3225e-29, 3.4912e-29, 2.4196e-29,\n 1.6647e-30, 1.4983e-29, 1.1290e-31, 5.8477e-29, 1.5483e-28, 4.8267e-29,\n 1.3040e-30, 7.7493e-30, 4.7119e-29, 1.5756e-29, 2.5554e-29, 1.5369e-30,\n 1.3163e-29, 7.1190e-30, 2.5045e-29, 7.0552e-29, 2.4305e-28, 8.9035e-29,\n 9.6448e-30, 7.3067e-29, 8.8511e-30, 1.1810e-28, 1.4449e-29, 1.3336e-29,\n 1.1872e-28, 4.1278e-29, 2.1389e-29, 2.7369e-30, 4.7803e-30, 2.6023e-29,\n 2.7294e-29, 5.7435e-30, 2.2191e-29, 1.5780e-30, 2.0950e-30, 6.8582e-29,\n 3.6027e-30, 2.0645e-29, 3.0912e-29, 2.1562e-30, 1.8836e-29, 1.7799e-29,\n 7.1440e-30, 9.8023e-30, 6.1174e-30, 1.0212e-28, 4.1479e-29, 7.3857e-30,\n 5.6487e-30, 3.0042e-29, 3.2531e-29, 6.1094e-29, 2.5731e-29, 1.2040e-28,\n 3.3065e-29, 4.4794e-29, 1.0417e-28, 1.2790e-28, 7.7389e-29, 4.0763e-29,\n 4.8091e-29, 2.3550e-29, 5.3923e-30, 4.8306e-30, 2.6826e-28, 3.1515e-31,\n 5.0956e-29, 3.0336e-30, 1.1918e-29, 1.4526e-29, 3.7398e-29, 1.2084e-29,\n 3.1694e-30, 3.5593e-33, 8.1385e-31, 4.7872e-30, 1.0746e-29, 2.7925e-29,\n 2.6557e-30, 6.3090e-29, 2.4171e-28, 2.7286e-29, 7.8069e-29, 1.0403e-29,\n 4.2684e-29, 1.1773e-28, 5.4248e-29, 3.6355e-30, 3.9148e-30, 1.5855e-28,\n 2.7299e-29, 3.0690e-29, 1.4233e-30, 1.9528e-29, 6.7603e-29, 2.2136e-29,\n 1.1599e-31, 3.0531e-29, 8.9860e-30, 8.7466e-30, 4.9650e-29, 4.5895e-30,\n 3.2083e-30, 2.7847e-29, 1.8415e-29, 1.1792e-30, 1.0842e-29, 1.2623e-28,\n 7.4552e-30, 2.5906e-28, 1.3181e-28, 1.8573e-30, 4.2943e-29, 1.3618e-28,\n 1.3475e-28, 5.0938e-30, 6.3919e-30, 1.1165e-29, 7.8378e-30, 3.3716e-29,\n 6.6780e-31, 3.3999e-29, 1.7393e-29, 4.3751e-30, 1.5875e-29, 2.9364e-31,\n 5.4974e-29, 3.3667e-31, 6.7616e-29, 3.7743e-29, 4.1638e-29, 9.4586e-29,\n 8.6220e-29, 3.1253e-30, 1.7617e-28, 1.6002e-29, 4.0779e-29, 8.8059e-30,\n 2.6052e-31, 1.4764e-29, 2.0316e-29, 3.3888e-29, 3.9130e-30, 1.2145e-29,\n 7.9830e-31, 1.8422e-29, 2.7709e-29, 1.2137e-30, 5.4329e-29, 1.0317e-30,\n 3.1151e-29, 1.1613e-29, 5.9430e-29, 2.7201e-29, 1.5493e-30, 1.2308e-29,\n 2.9389e-29, 8.0165e-30, 8.8889e-30, 1.6196e-29, 3.1691e-29, 1.8605e-29,\n 2.4512e-29, 7.0899e-31, 3.1764e-29, 3.4155e-29, 9.1230e-29, 9.2616e-29,\n 2.3031e-30, 1.6954e-29, 2.4183e-30, 1.6166e-29, 3.4601e-31, 4.4593e-30,\n 5.3518e-30, 2.8557e-29, 9.8258e-31, 1.4483e-30, 6.3659e-30, 7.1144e-29,\n 5.0792e-30, 1.4589e-29, 3.8814e-30, 4.8616e-29, 1.0030e-28, 1.0443e-32,\n 9.5463e-29, 6.7188e-29, 9.1938e-29, 4.3269e-31, 1.5132e-29, 1.0370e-28,\n 4.5682e-30, 2.0463e-30, 1.3163e-10, 1.9882e-10, 2.0918e-12, 1.6707e-09,\n 4.5456e-10, 1.1403e-10, 2.8746e-10, 8.0989e-10, 1.6120e-11, 2.0133e-10,\n 1.3317e-09, 3.1600e-12, 9.3635e-11, 3.3931e-10, 1.6841e-09, 3.8144e-11,\n 4.9856e-10, 1.6751e-09, 1.1509e-10, 1.9272e-10, 5.3648e-10, 9.4895e-13,\n 7.0554e-10, 2.1554e-09, 1.3585e-10, 2.3491e-09, 4.9699e-11, 3.7573e-10,\n 3.0889e-10, 1.4152e-09, 3.0213e-10, 3.2774e-12, 5.3289e-11, 3.4686e-12,\n 6.4895e-10, 8.9984e-10, 9.0153e-10, 1.3844e-10, 2.6367e-11, 8.8146e-10,\n 1.6717e-09, 8.3968e-10, 9.9377e-10, 7.7546e-10, 3.3684e-10, 2.2078e-12,\n 6.0361e-10, 1.7189e-10, 8.2397e-10, 5.8720e-10, 3.2865e-09, 6.3918e-10,\n 1.0893e-10, 5.0632e-12, 2.8127e-10, 2.8054e-10, 1.8013e-09, 1.5372e-10,\n 4.3085e-10, 4.3532e-10, 4.4640e-10, 1.7134e-10, 1.6483e-10, 1.2108e-10,\n 4.6813e-11, 3.7714e-12, 9.7924e-10, 3.6517e-10, 4.9658e-10, 6.4407e-10,\n 4.6564e-11, 1.5105e-11, 9.1990e-12, 2.1845e-09, 7.0719e-11, 2.9833e-09,\n 5.7151e-10, 1.6565e-09, 7.6006e-10, 7.6911e-11, 4.8164e-10, 5.6230e-09,\n 1.9673e-10, 1.2430e-09, 4.5159e-11, 1.6896e-09, 6.1050e-10, 3.8693e-11,\n 1.0838e-11, 6.7623e-10, 9.8361e-10, 4.5879e-10, 3.3205e-12, 9.4180e-10,\n 2.4118e-09, 1.1886e-10, 5.0789e-11, 9.8573e-10, 9.2803e-11, 9.3297e-11,\n 1.2490e-10, 5.8699e-10, 1.5716e-11, 7.7048e-11, 2.0818e-10, 2.0786e-09,\n 1.2579e-09, 1.7909e-10, 1.9056e-11, 2.2334e-10, 1.0069e-10, 3.3589e-10,\n 6.2765e-10, 3.4774e-11, 4.8193e-10, 6.7073e-10, 1.7738e-10, 8.3788e-11,\n 1.1840e-09, 2.3209e-11, 8.2619e-11, 4.5417e-09, 8.3089e-10, 2.8690e-09,\n 1.0352e-09, 3.4332e-10, 2.5527e-11, 4.1988e-11, 5.1877e-10, 5.8155e-10,\n 3.2448e-10, 4.5876e-10, 1.1483e-10, 1.9428e-12, 4.9648e-10, 1.8570e-10,\n 2.6666e-10, 1.2297e-09, 1.3876e-10, 2.7128e-11, 2.1950e-13, 1.0169e-09,\n 4.4016e-10, 1.0258e-11, 3.0235e-10, 2.4440e-10, 6.3769e-11, 2.1747e-09,\n 9.2868e-11, 1.6781e-12, 2.8714e-10, 3.2482e-09, 1.6869e-10, 1.3776e-10,\n 1.2881e-09, 3.8637e-11, 1.4640e-11, 2.4342e-10, 2.1177e-13, 4.1277e-10,\n 1.0390e-09, 1.4665e-09, 4.2398e-11, 2.7726e-10, 3.8642e-10, 6.9511e-10,\n 3.2991e-10, 1.6014e-10, 2.6615e-11, 1.7387e-09, 1.0720e-10, 2.2431e-09,\n 1.4190e-09, 1.0707e-10, 3.6145e-10, 9.3877e-11, 5.7943e-12, 1.1893e-10,\n 9.6984e-10, 3.7409e-12, 6.7456e-10, 1.9276e-09, 5.0907e-12, 4.5848e-10,\n 1.2737e-09, 1.1983e-09, 3.2396e-10, 1.8722e-10, 1.4624e-11, 1.2327e-09,\n 2.2472e-11, 4.1734e-10, 1.8916e-09, 1.9579e-09, 3.4081e-10, 5.7404e-10,\n 1.0241e-09, 2.4755e-09, 2.3904e-11, 1.0980e-09, 1.3644e-10, 6.8762e-11,\n 2.5494e-09, 1.7490e-10, 8.0797e-11, 3.1190e-10, 1.0111e-10, 1.8344e-10,\n 2.8505e-10, 9.3447e-11, 5.3829e-10, 4.0379e-10, 1.2958e-09, 1.0047e-10,\n 1.4988e-11, 4.0781e-09, 1.2837e-10, 2.4146e-09, 4.8989e-12, 2.6399e-10,\n 2.0658e-10, 9.1763e-10, 1.0476e-09, 6.1853e-10, 1.3391e-09, 4.7446e-11,\n 8.4908e-10, 1.1251e-10, 5.2326e-09, 3.3874e-11, 1.2357e-10, 5.1439e-10,\n 5.2565e-10, 3.0716e-10, 1.0737e-09, 3.8852e-10, 1.0896e-10, 9.6690e-11,\n 6.5587e-10, 3.9293e-12, 9.5535e-10, 3.6057e-11, 8.8632e-14, 1.7883e-10,\n 9.0959e-10, 1.2010e-11, 2.4036e-10, 2.3427e-11, 3.1642e-11, 2.1240e-10,\n 7.6038e-10, 5.4919e-10, 4.1328e-10, 1.7938e-09, 5.4208e-10, 1.5816e-10],\n device='cuda:0')" }, "54": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [-5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, -5.6052e-45, ..., 5.6052e-45,\n -5.6052e-45, -5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.2702e-10, 4.1625e-12, 2.8990e-12, ..., 9.6173e-12, 5.1048e-11,\n 6.6579e-11],\n [6.6957e-11, 3.0669e-12, 2.1936e-12, ..., 8.1221e-12, 2.9007e-11,\n 3.7984e-11],\n [1.2248e-09, 2.7770e-11, 2.4284e-11, ..., 4.0189e-11, 4.8968e-10,\n 6.1500e-10],\n ...,\n [3.8525e-10, 9.6462e-12, 7.0010e-12, ..., 1.7725e-11, 1.5404e-10,\n 2.0452e-10],\n [9.5558e-12, 7.2371e-13, 2.2394e-13, ..., 3.6473e-12, 3.7576e-12,\n 3.5078e-12],\n [2.1599e-11, 1.7497e-12, 1.3623e-12, ..., 5.8641e-12, 1.0624e-11,\n 1.1174e-11]], device='cuda:0')" + "exp_avg_sq": "tensor([[3.6296e-11, 1.1895e-12, 8.2840e-13, ..., 2.7482e-12, 1.4587e-11,\n 1.9026e-11],\n [1.9133e-11, 8.7639e-13, 6.2685e-13, ..., 2.3209e-12, 8.2890e-12,\n 1.0854e-11],\n [3.5000e-10, 7.9356e-12, 6.9392e-12, ..., 1.1484e-11, 1.3993e-10,\n 1.7574e-10],\n ...,\n [1.1009e-10, 2.7565e-12, 2.0006e-12, ..., 5.0650e-12, 4.4018e-11,\n 5.8445e-11],\n [2.7306e-12, 2.0681e-13, 6.3994e-14, ..., 1.0422e-12, 1.0738e-12,\n 1.0024e-12],\n [6.1721e-12, 4.9998e-13, 3.8928e-13, ..., 1.6757e-12, 3.0359e-12,\n 3.1931e-12]], device='cuda:0')" }, "55": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n -5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, -5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45,\n -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45, -5.6052e-45,\n -5.6052e-45, 5.6052e-45, -5.6052e-45, -5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, -5.6052e-45, 5.6052e-45,\n -5.6052e-45], device='cuda:0')", - "exp_avg_sq": "tensor([2.2734e-09, 1.2179e-09, 2.1638e-08, 1.1654e-08, 8.7356e-09, 1.1925e-11,\n 2.9041e-09, 3.0143e-09, 5.2079e-09, 5.1978e-10, 9.6106e-09, 1.2649e-10,\n 7.6180e-09, 4.4636e-12, 7.4967e-09, 1.8304e-11, 4.0195e-10, 3.5517e-09,\n 3.3235e-09, 5.6938e-09, 8.1539e-09, 1.2462e-09, 9.5351e-09, 4.3001e-09,\n 2.4014e-08, 6.4869e-09, 5.1684e-09, 1.2054e-09, 4.8196e-09, 1.1611e-08,\n 6.0805e-09, 2.1202e-08, 2.3599e-09, 2.7316e-10, 9.7926e-10, 1.0508e-09,\n 2.6955e-09, 8.9280e-09, 5.5316e-09, 1.1384e-09, 3.0832e-09, 8.6444e-09,\n 9.0908e-11, 3.3556e-09, 5.9787e-09, 2.1044e-09, 2.2044e-09, 6.9534e-09,\n 2.5981e-08, 1.2957e-09, 9.8671e-09, 5.0366e-09, 2.4580e-09, 2.0853e-09,\n 5.4186e-11, 1.9056e-08, 3.5981e-10, 8.0027e-09, 1.9482e-09, 1.8788e-08,\n 1.0672e-08, 7.2747e-09, 2.3001e-10, 3.8754e-09, 1.6653e-09, 1.8335e-09,\n 2.0162e-09, 1.9103e-08, 1.9124e-10, 9.0638e-09, 6.8541e-11, 7.3417e-09,\n 1.2151e-10, 7.5243e-09, 4.5795e-11, 2.0102e-09, 1.6758e-08, 9.5867e-10,\n 7.0868e-09, 1.3461e-08, 6.2308e-09, 4.4051e-09, 5.0777e-09, 7.0036e-09,\n 6.3432e-09, 8.9455e-09, 3.0707e-09, 4.3107e-09, 5.2586e-09, 2.8862e-10,\n 1.5060e-09, 1.8918e-09, 3.3457e-08, 9.1239e-10, 4.4927e-09, 4.6961e-12,\n 2.4087e-10, 7.6529e-11, 1.6932e-11, 5.2065e-10, 3.3444e-08, 2.3802e-10,\n 3.1675e-09, 2.7801e-08, 2.1891e-09, 5.8025e-10, 9.1144e-10, 2.3418e-10,\n 6.4023e-09, 1.5225e-08, 8.6314e-11, 4.4757e-09, 1.7830e-09, 1.7441e-09,\n 5.6885e-09, 1.3426e-09, 3.3304e-10, 1.8830e-10, 4.3342e-09, 3.3787e-09,\n 6.6496e-11, 4.7511e-09, 3.3609e-10, 3.3881e-08, 1.3898e-08, 3.4882e-08,\n 1.2996e-08, 8.1725e-10, 7.2263e-10, 2.0222e-09, 1.5701e-08, 2.5724e-10,\n 1.6144e-09, 7.2453e-09, 2.5374e-11, 1.0173e-10, 4.5093e-09, 1.0612e-08,\n 5.1931e-09, 1.2518e-09, 2.3298e-09, 6.1176e-09, 3.1834e-09, 2.9513e-09,\n 3.2016e-09, 7.0853e-09, 2.6427e-09, 4.0349e-09, 2.8563e-09, 1.8513e-08,\n 1.0303e-11, 2.1214e-09, 2.5928e-10, 2.1435e-10, 4.4752e-08, 1.9482e-08,\n 9.7065e-10, 1.8198e-08, 4.1317e-09, 2.5391e-08, 5.1226e-10, 5.2466e-10,\n 3.3700e-09, 1.4313e-11, 2.4735e-08, 4.1492e-09, 5.4605e-09, 8.7364e-10,\n 3.9040e-09, 1.7910e-08, 2.4529e-10, 4.2409e-09, 5.7701e-09, 1.5852e-09,\n 1.7589e-08, 4.3128e-09, 3.4382e-09, 1.2104e-10, 3.0907e-09, 9.3106e-10,\n 8.2117e-09, 3.8059e-09, 5.7461e-10, 1.1728e-09, 1.1351e-10, 2.2711e-09,\n 9.8124e-10, 1.4496e-08, 3.5843e-10, 1.3741e-09, 1.1874e-09, 2.6215e-10,\n 8.5794e-10, 9.3135e-09, 3.3762e-09, 2.9627e-09, 9.0343e-09, 4.9274e-09,\n 5.3025e-09, 2.4659e-09, 2.6375e-09, 1.1096e-08, 6.7776e-09, 1.8295e-10,\n 1.1666e-08, 3.9677e-10, 8.9851e-09, 1.1886e-08, 1.2233e-08, 1.1763e-08,\n 3.3699e-09, 2.5505e-09, 1.0891e-08, 4.3692e-10, 9.9659e-09, 4.7206e-09,\n 6.3555e-12, 6.8700e-09, 1.6387e-11, 5.1280e-09, 2.9559e-09, 7.3815e-08,\n 8.6311e-09, 2.6375e-08, 2.4055e-10, 1.1926e-09, 5.5636e-09, 4.2671e-10,\n 1.5367e-09, 8.5672e-11, 2.6454e-09, 6.7589e-09, 7.0970e-09, 1.0648e-09,\n 2.5231e-08, 3.3046e-09, 1.4597e-09, 1.5932e-10, 1.4346e-09, 1.8586e-11,\n 7.9435e-10, 6.0046e-09, 4.3985e-10, 4.0197e-10, 2.8913e-09, 1.8120e-08,\n 3.0262e-09, 9.4218e-09, 4.6120e-09, 2.2490e-09, 1.9008e-08, 4.4192e-10,\n 1.4920e-08, 6.8747e-09, 1.2970e-10, 3.7848e-10], device='cuda:0')" + "exp_avg_sq": "tensor([6.4964e-10, 3.4802e-10, 6.1832e-09, 3.3303e-09, 2.4963e-09, 3.4077e-12,\n 8.2988e-10, 8.6137e-10, 1.4882e-09, 1.4853e-10, 2.7463e-09, 3.6145e-11,\n 2.1769e-09, 1.2755e-12, 2.1422e-09, 5.2306e-12, 1.1486e-10, 1.0149e-09,\n 9.4972e-10, 1.6270e-09, 2.3300e-09, 3.5611e-10, 2.7247e-09, 1.2288e-09,\n 6.8623e-09, 1.8537e-09, 1.4769e-09, 3.4444e-10, 1.3772e-09, 3.3179e-09,\n 1.7375e-09, 6.0587e-09, 6.7435e-10, 7.8058e-11, 2.7983e-10, 3.0028e-10,\n 7.7025e-10, 2.5513e-09, 1.5807e-09, 3.2531e-10, 8.8104e-10, 2.4702e-09,\n 2.5978e-11, 9.5890e-10, 1.7084e-09, 6.0136e-10, 6.2992e-10, 1.9870e-09,\n 7.4244e-09, 3.7025e-10, 2.8196e-09, 1.4392e-09, 7.0238e-10, 5.9590e-10,\n 1.5484e-11, 5.4454e-09, 1.0282e-10, 2.2868e-09, 5.5673e-10, 5.3689e-09,\n 3.0495e-09, 2.0788e-09, 6.5728e-11, 1.1074e-09, 4.7587e-10, 5.2394e-10,\n 5.7614e-10, 5.4587e-09, 5.4647e-11, 2.5900e-09, 1.9586e-11, 2.0980e-09,\n 3.4721e-11, 2.1501e-09, 1.3086e-11, 5.7442e-10, 4.7888e-09, 2.7395e-10,\n 2.0251e-09, 3.8466e-09, 1.7805e-09, 1.2588e-09, 1.4510e-09, 2.0013e-09,\n 1.8126e-09, 2.5562e-09, 8.7748e-10, 1.2318e-09, 1.5027e-09, 8.2476e-11,\n 4.3036e-10, 5.4061e-10, 9.5607e-09, 2.6072e-10, 1.2838e-09, 1.3419e-12,\n 6.8831e-11, 2.1869e-11, 4.8386e-12, 1.4878e-10, 9.5570e-09, 6.8017e-11,\n 9.0512e-10, 7.9444e-09, 6.2556e-10, 1.6581e-10, 2.6045e-10, 6.6920e-11,\n 1.8295e-09, 4.3508e-09, 2.4665e-11, 1.2790e-09, 5.0952e-10, 4.9838e-10,\n 1.6255e-09, 3.8365e-10, 9.5169e-11, 5.3808e-11, 1.2385e-09, 9.6550e-10,\n 1.9002e-11, 1.3577e-09, 9.6039e-11, 9.6817e-09, 3.9715e-09, 9.9677e-09,\n 3.7137e-09, 2.3354e-10, 2.0650e-10, 5.7787e-10, 4.4868e-09, 7.3508e-11,\n 4.6133e-10, 2.0704e-09, 7.2508e-12, 2.9070e-11, 1.2886e-09, 3.0324e-09,\n 1.4840e-09, 3.5772e-10, 6.6577e-10, 1.7481e-09, 9.0968e-10, 8.4337e-10,\n 9.1489e-10, 2.0247e-09, 7.5517e-10, 1.1530e-09, 8.1621e-10, 5.2903e-09,\n 2.9442e-12, 6.0621e-10, 7.4091e-11, 6.1252e-11, 1.2788e-08, 5.5672e-09,\n 2.7737e-10, 5.2002e-09, 1.1807e-09, 7.2557e-09, 1.4638e-10, 1.4993e-10,\n 9.6299e-10, 4.0902e-12, 7.0683e-09, 1.1857e-09, 1.5604e-09, 2.4965e-10,\n 1.1156e-09, 5.1180e-09, 7.0094e-11, 1.2119e-09, 1.6488e-09, 4.5298e-10,\n 5.0262e-09, 1.2324e-09, 9.8250e-10, 3.4587e-11, 8.8320e-10, 2.6606e-10,\n 2.3466e-09, 1.0876e-09, 1.6420e-10, 3.3512e-10, 3.2435e-11, 6.4898e-10,\n 2.8040e-10, 4.1423e-09, 1.0242e-10, 3.9266e-10, 3.3931e-10, 7.4911e-11,\n 2.4516e-10, 2.6614e-09, 9.6478e-10, 8.4663e-10, 2.5816e-09, 1.4080e-09,\n 1.5152e-09, 7.0465e-10, 7.5369e-10, 3.1706e-09, 1.9367e-09, 5.2280e-11,\n 3.3337e-09, 1.1338e-10, 2.5676e-09, 3.3966e-09, 3.4958e-09, 3.3612e-09,\n 9.6297e-10, 7.2883e-10, 3.1120e-09, 1.2485e-10, 2.8478e-09, 1.3490e-09,\n 1.8161e-12, 1.9632e-09, 4.6827e-12, 1.4654e-09, 8.4466e-10, 2.1093e-08,\n 2.4664e-09, 7.5367e-09, 6.8740e-11, 3.4080e-10, 1.5898e-09, 1.2194e-10,\n 4.3911e-10, 2.4481e-11, 7.5595e-10, 1.9314e-09, 2.0280e-09, 3.0428e-10,\n 7.2099e-09, 9.4430e-10, 4.1711e-10, 4.5527e-11, 4.0994e-10, 5.3111e-12,\n 2.2699e-10, 1.7159e-09, 1.2569e-10, 1.1486e-10, 8.2622e-10, 5.1779e-09,\n 8.6475e-10, 2.6924e-09, 1.3179e-09, 6.4267e-10, 5.4316e-09, 1.2628e-10,\n 4.2634e-09, 1.9645e-09, 3.7063e-11, 1.0815e-10], device='cuda:0')" }, "56": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[-5.6052e-45, -5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[5.5761e-08, 1.2097e-08, 3.3469e-08, ..., 1.4493e-08, 1.2408e-09,\n 1.9037e-10],\n [5.0245e-10, 1.3094e-10, 2.9727e-10, ..., 1.1948e-10, 1.1835e-11,\n 2.3523e-12],\n [6.8433e-10, 1.4440e-10, 4.0833e-10, ..., 1.8091e-10, 1.5308e-11,\n 2.1946e-12],\n ...,\n [5.5947e-10, 1.1797e-10, 3.3771e-10, ..., 1.4378e-10, 1.3257e-11,\n 2.7554e-12],\n [6.9308e-10, 1.4966e-10, 4.1660e-10, ..., 1.8031e-10, 1.2851e-11,\n 1.2891e-12],\n [6.1503e-10, 1.2976e-10, 3.8465e-10, ..., 1.6464e-10, 1.4854e-11,\n 3.7643e-12]], device='cuda:0')" + "exp_avg_sq": "tensor([[1.5934e-08, 3.4567e-09, 9.5640e-09, ..., 4.1415e-09, 3.5456e-10,\n 5.4400e-11],\n [1.4358e-10, 3.7418e-11, 8.4946e-11, ..., 3.4141e-11, 3.3819e-12,\n 6.7219e-13],\n [1.9555e-10, 4.1263e-11, 1.1668e-10, ..., 5.1697e-11, 4.3744e-12,\n 6.2711e-13],\n ...,\n [1.5987e-10, 3.3710e-11, 9.6504e-11, ..., 4.1086e-11, 3.7883e-12,\n 7.8738e-13],\n [1.9805e-10, 4.2768e-11, 1.1905e-10, ..., 5.1525e-11, 3.6721e-12,\n 3.6837e-13],\n [1.7575e-10, 3.7081e-11, 1.0992e-10, ..., 4.7047e-11, 4.2448e-12,\n 1.0757e-12]], device='cuda:0')" }, "57": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45],\n device='cuda:0')", - "exp_avg_sq": "tensor([1.4373e-06, 1.3031e-08, 1.7601e-08, 1.8265e-08, 2.0941e-08, 2.4055e-08,\n 1.9125e-08, 1.4163e-08, 1.7889e-08, 1.5897e-08], device='cuda:0')" + "exp_avg_sq": "tensor([4.1073e-07, 3.7238e-09, 5.0296e-09, 5.2193e-09, 5.9840e-09, 6.8739e-09,\n 5.4651e-09, 4.0472e-09, 5.1120e-09, 4.5427e-09], device='cuda:0')" }, "58": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[-5.6052e-45, -5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[5.5755e-08, 1.2094e-08, 3.3459e-08, ..., 1.4493e-08, 1.2380e-09,\n 1.8845e-10],\n [5.0237e-10, 1.3091e-10, 2.9713e-10, ..., 1.1948e-10, 1.1793e-11,\n 2.3242e-12],\n [6.8427e-10, 1.4437e-10, 4.0823e-10, ..., 1.8091e-10, 1.5277e-11,\n 2.1731e-12],\n ...,\n [5.5938e-10, 1.1793e-10, 3.3755e-10, ..., 1.4378e-10, 1.3210e-11,\n 2.7234e-12],\n [6.9305e-10, 1.4966e-10, 4.1656e-10, ..., 1.8031e-10, 1.2839e-11,\n 1.2815e-12],\n [6.1486e-10, 1.2970e-10, 3.8437e-10, ..., 1.6464e-10, 1.4771e-11,\n 3.7080e-12]], device='cuda:0')" + "exp_avg_sq": "tensor([[1.5932e-08, 3.4560e-09, 9.5613e-09, ..., 4.1415e-09, 3.5376e-10,\n 5.3851e-11],\n [1.4356e-10, 3.7409e-11, 8.4906e-11, ..., 3.4141e-11, 3.3701e-12,\n 6.6416e-13],\n [1.9553e-10, 4.1255e-11, 1.1665e-10, ..., 5.1697e-11, 4.3654e-12,\n 6.2098e-13],\n ...,\n [1.5985e-10, 3.3699e-11, 9.6459e-11, ..., 4.1085e-11, 3.7750e-12,\n 7.7823e-13],\n [1.9805e-10, 4.2765e-11, 1.1904e-10, ..., 5.1525e-11, 3.6689e-12,\n 3.6619e-13],\n [1.7570e-10, 3.7062e-11, 1.0984e-10, ..., 4.7047e-11, 4.2211e-12,\n 1.0596e-12]], device='cuda:0')" }, "59": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45],\n device='cuda:0')", - "exp_avg_sq": "tensor([1.4373e-06, 1.3031e-08, 1.7601e-08, 1.8265e-08, 2.0941e-08, 2.4055e-08,\n 1.9125e-08, 1.4163e-08, 1.7889e-08, 1.5897e-08], device='cuda:0')" + "exp_avg_sq": "tensor([4.1073e-07, 3.7238e-09, 5.0296e-09, 5.2192e-09, 5.9840e-09, 6.8739e-09,\n 5.4650e-09, 4.0472e-09, 5.1119e-09, 4.5427e-09], device='cuda:0')" }, "60": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([[-5.6052e-45, -5.6052e-45, -5.6052e-45, ..., -5.6052e-45,\n -5.6052e-45, -5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n ...,\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45],\n [ 5.6052e-45, 5.6052e-45, 5.6052e-45, ..., 5.6052e-45,\n 5.6052e-45, 5.6052e-45]], device='cuda:0')", - "exp_avg_sq": "tensor([[5.5761e-08, 1.2097e-08, 3.3469e-08, ..., 1.4493e-08, 1.2408e-09,\n 1.9037e-10],\n [5.0245e-10, 1.3094e-10, 2.9727e-10, ..., 1.1948e-10, 1.1835e-11,\n 2.3523e-12],\n [6.8433e-10, 1.4440e-10, 4.0833e-10, ..., 1.8091e-10, 1.5308e-11,\n 2.1946e-12],\n ...,\n [5.5947e-10, 1.1797e-10, 3.3771e-10, ..., 1.4378e-10, 1.3257e-11,\n 2.7554e-12],\n [6.9308e-10, 1.4966e-10, 4.1660e-10, ..., 1.8031e-10, 1.2851e-11,\n 1.2891e-12],\n [6.1503e-10, 1.2976e-10, 3.8465e-10, ..., 1.6464e-10, 1.4854e-11,\n 3.7643e-12]], device='cuda:0')" + "exp_avg_sq": "tensor([[1.5934e-08, 3.4567e-09, 9.5640e-09, ..., 4.1415e-09, 3.5456e-10,\n 5.4400e-11],\n [1.4358e-10, 3.7418e-11, 8.4946e-11, ..., 3.4141e-11, 3.3819e-12,\n 6.7219e-13],\n [1.9555e-10, 4.1263e-11, 1.1668e-10, ..., 5.1697e-11, 4.3744e-12,\n 6.2711e-13],\n ...,\n [1.5987e-10, 3.3710e-11, 9.6504e-11, ..., 4.1086e-11, 3.7883e-12,\n 7.8738e-13],\n [1.9805e-10, 4.2768e-11, 1.1905e-10, ..., 5.1525e-11, 3.6721e-12,\n 3.6837e-13],\n [1.7575e-10, 3.7081e-11, 1.0992e-10, ..., 4.7047e-11, 4.2448e-12,\n 1.0757e-12]], device='cuda:0')" }, "61": { - "step": "tensor(5008.)", + "step": "tensor(6260.)", "exp_avg": "tensor([-5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45,\n 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45, 5.6052e-45],\n device='cuda:0')", - "exp_avg_sq": "tensor([1.4373e-06, 1.3031e-08, 1.7601e-08, 1.8265e-08, 2.0941e-08, 2.4055e-08,\n 1.9125e-08, 1.4163e-08, 1.7889e-08, 1.5897e-08], device='cuda:0')" + "exp_avg_sq": "tensor([4.1073e-07, 3.7238e-09, 5.0296e-09, 5.2193e-09, 5.9840e-09, 6.8739e-09,\n 5.4651e-09, 4.0472e-09, 5.1120e-09, 4.5427e-09], device='cuda:0')" }, "8": { - "step": "tensor(3756.)", - "exp_avg": "tensor([[ 3.3050e-06, -1.4826e-06, -2.4800e-07, ..., -2.0102e-07,\n 0.0000e+00, -1.5793e-08],\n [-1.4022e-06, 2.0576e-08, -5.3183e-07, ..., -7.3770e-07,\n 0.0000e+00, 2.0642e-06],\n [-1.6411e-07, 9.4075e-09, -6.2795e-07, ..., -4.2783e-07,\n 0.0000e+00, 1.3053e-07],\n ...,\n [-5.7752e-07, 3.2708e-08, 1.2804e-06, ..., -2.7579e-06,\n 0.0000e+00, 1.0007e-06],\n [ 8.3237e-08, -8.3440e-09, 4.8899e-07, ..., -5.5080e-07,\n 0.0000e+00, 1.0731e-06],\n [-5.8672e-07, 2.2360e-08, -6.9772e-08, ..., 8.9757e-07,\n 0.0000e+00, -3.7203e-07]], device='cuda:0')", - "exp_avg_sq": "tensor([[1.3802e-11, 2.4263e-12, 3.1701e-11, ..., 3.9639e-11, 0.0000e+00,\n 4.7108e-12],\n [2.2565e-11, 6.6864e-13, 6.1874e-11, ..., 1.6449e-11, 0.0000e+00,\n 3.5699e-11],\n [5.6193e-12, 4.7289e-12, 5.1355e-11, ..., 1.0132e-11, 0.0000e+00,\n 9.5389e-12],\n ...,\n [1.2220e-11, 2.8527e-12, 2.1157e-11, ..., 4.9039e-11, 0.0000e+00,\n 2.5976e-11],\n [1.2748e-11, 2.3249e-12, 1.7470e-10, ..., 8.4237e-11, 0.0000e+00,\n 8.7184e-11],\n [4.4334e-11, 1.4753e-11, 5.2298e-11, ..., 4.7643e-11, 0.0000e+00,\n 1.5019e-11]], device='cuda:0')" + "step": "tensor(5008.)", + "exp_avg": "tensor([[-5.0826e-09, 5.6962e-07, -9.4582e-07, ..., -6.9808e-07,\n 0.0000e+00, 1.8751e-10],\n [-3.1985e-07, 1.3486e-08, 3.2612e-07, ..., -2.2639e-08,\n 0.0000e+00, -2.3643e-07],\n [-2.3131e-07, -7.9868e-08, -1.2513e-07, ..., -6.5193e-08,\n 0.0000e+00, 2.6853e-06],\n ...,\n [ 7.5336e-07, 2.2135e-07, 3.8334e-07, ..., -1.2551e-06,\n 0.0000e+00, 5.1076e-06],\n [-1.6613e-07, 6.3618e-09, 1.9841e-06, ..., 1.9786e-06,\n 0.0000e+00, 8.6103e-08],\n [ 5.9936e-07, -1.5554e-07, 1.7261e-06, ..., -2.4410e-06,\n 0.0000e+00, -4.6505e-08]], device='cuda:0')", + "exp_avg_sq": "tensor([[6.9293e-12, 1.4506e-12, 2.6802e-11, ..., 2.7772e-11, 0.0000e+00,\n 3.6257e-12],\n [1.9960e-11, 4.8525e-13, 3.7170e-11, ..., 7.7162e-12, 0.0000e+00,\n 2.4445e-11],\n [5.6413e-12, 2.6440e-12, 3.3237e-11, ..., 5.8931e-12, 0.0000e+00,\n 6.1625e-12],\n ...,\n [7.2599e-12, 1.8992e-12, 1.3625e-11, ..., 5.0757e-11, 0.0000e+00,\n 1.8631e-11],\n [6.7263e-12, 1.2203e-12, 1.0565e-10, ..., 6.9829e-11, 0.0000e+00,\n 6.8267e-11],\n [3.2867e-11, 8.3434e-12, 3.3946e-11, ..., 3.8466e-11, 0.0000e+00,\n 1.3044e-11]], device='cuda:0')" }, "9": { - "step": "tensor(3756.)", - "exp_avg": "tensor([-7.2511e-06, 1.2043e-05, 9.9536e-07, ..., -1.9313e-06,\n -1.7364e-06, 1.1259e-06], device='cuda:0')", - "exp_avg_sq": "tensor([2.5431e-09, 2.8160e-09, 2.0961e-09, ..., 2.1836e-09, 2.7892e-09,\n 2.6763e-09], device='cuda:0')" + "step": "tensor(5008.)", + "exp_avg": "tensor([ 2.2834e-06, 5.4557e-06, -2.3191e-06, ..., -6.5588e-06,\n 1.4447e-07, -7.7950e-06], device='cuda:0')", + "exp_avg_sq": "tensor([1.9485e-09, 2.1639e-09, 1.6383e-09, ..., 1.5327e-09, 1.8287e-09,\n 1.9270e-09], device='cuda:0')" }, "10": { - "step": "tensor(3756.)", - "exp_avg": "tensor([[-1.4582e-07, -3.2010e-08, 8.8345e-08, ..., -8.3714e-07,\n 2.0524e-07, -1.4161e-07],\n [ 2.5680e-07, -6.5429e-07, -1.0099e-06, ..., 5.4951e-08,\n 2.9744e-07, -5.2585e-07],\n [ 2.2452e-08, 3.6399e-07, -1.0430e-06, ..., 2.9101e-07,\n -3.9084e-08, 2.8615e-07],\n ...,\n [ 1.3794e-07, 4.5533e-07, -2.2350e-07, ..., -1.5740e-06,\n -4.9998e-07, -2.8954e-07],\n [-1.9641e-07, 4.6601e-07, 1.3403e-06, ..., 8.0811e-08,\n -3.8842e-07, 1.8245e-07],\n [-4.6505e-08, -2.4896e-07, 1.7009e-07, ..., -4.5994e-07,\n 2.9078e-07, 4.7989e-07]], device='cuda:0')", - "exp_avg_sq": "tensor([[2.2996e-12, 2.7972e-12, 3.1599e-12, ..., 6.2085e-12, 3.3779e-12,\n 3.2575e-12],\n [6.1283e-12, 7.4758e-12, 4.7177e-12, ..., 7.6014e-12, 5.9618e-12,\n 5.2781e-12],\n [5.7199e-12, 7.3769e-12, 6.4712e-12, ..., 7.2539e-12, 7.4178e-12,\n 5.9844e-12],\n ...,\n [7.8544e-12, 5.5159e-12, 6.4635e-12, ..., 1.2175e-11, 9.2106e-12,\n 5.3115e-12],\n [5.7846e-12, 5.2673e-12, 7.7382e-12, ..., 7.8881e-12, 1.0459e-11,\n 5.6582e-12],\n [8.8593e-12, 7.0071e-12, 5.0220e-12, ..., 7.6055e-12, 6.9697e-12,\n 8.3465e-12]], device='cuda:0')" + "step": "tensor(5008.)", + "exp_avg": "tensor([[-6.1133e-08, -2.2699e-07, 1.1548e-07, ..., 5.4343e-07,\n 1.1997e-07, -2.4621e-07],\n [-1.7459e-08, 2.6073e-07, 5.4724e-07, ..., 1.3667e-06,\n 2.5627e-07, -6.5040e-07],\n [-1.8486e-07, -2.2381e-07, -1.5276e-07, ..., -3.7863e-07,\n 1.2714e-07, -1.7166e-07],\n ...,\n [-5.4842e-07, 2.1764e-07, -6.7904e-07, ..., 1.8933e-06,\n 7.6690e-08, -4.1044e-07],\n [ 1.3948e-07, 4.6015e-07, 8.5891e-07, ..., 1.2815e-07,\n -1.2013e-07, -2.5677e-07],\n [-5.4631e-09, 4.2475e-08, -1.0664e-07, ..., -3.0244e-07,\n 2.1689e-08, -9.1923e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[1.4586e-12, 1.7521e-12, 2.3510e-12, ..., 3.7267e-12, 2.0205e-12,\n 1.6216e-12],\n [3.3512e-12, 4.5592e-12, 3.0438e-12, ..., 4.8338e-12, 3.6955e-12,\n 2.9662e-12],\n [3.4701e-12, 5.0114e-12, 4.2479e-12, ..., 4.3990e-12, 4.9313e-12,\n 3.2922e-12],\n ...,\n [4.6657e-12, 3.2303e-12, 3.8175e-12, ..., 7.1166e-12, 5.8046e-12,\n 3.1217e-12],\n [3.6557e-12, 3.3538e-12, 5.1507e-12, ..., 5.1126e-12, 6.6289e-12,\n 3.2912e-12],\n [5.3020e-12, 4.5069e-12, 3.2411e-12, ..., 4.7514e-12, 4.0652e-12,\n 4.7530e-12]], device='cuda:0')" }, "11": { - "step": "tensor(2504.)", - "exp_avg": "tensor([[-8.8237e-07, -2.4691e-06, 8.1761e-07, ..., 3.4484e-06,\n 0.0000e+00, -1.1416e-06],\n [-6.1348e-07, -1.0298e-06, -1.4337e-07, ..., 3.6510e-07,\n 0.0000e+00, -1.6920e-06],\n [ 3.2471e-07, -9.0841e-08, -6.9405e-07, ..., 5.1907e-06,\n 0.0000e+00, -1.2068e-07],\n ...,\n [ 1.7956e-07, 2.7996e-07, 1.6133e-06, ..., 1.5855e-07,\n 0.0000e+00, 2.8758e-07],\n [ 7.8083e-07, -9.2970e-07, -2.8678e-07, ..., 5.4130e-08,\n 0.0000e+00, -9.7476e-07],\n [-3.9340e-07, -1.8520e-10, 1.7226e-07, ..., -1.1919e-06,\n 0.0000e+00, -7.6001e-08]], device='cuda:0')", - "exp_avg_sq": "tensor([[4.6084e-11, 5.9483e-11, 1.5442e-11, ..., 1.3043e-11, 0.0000e+00,\n 4.6604e-11],\n [1.6724e-11, 3.6630e-11, 8.9965e-12, ..., 2.2404e-11, 0.0000e+00,\n 1.1194e-10],\n [9.5505e-12, 4.8172e-12, 5.9238e-11, ..., 8.9772e-11, 0.0000e+00,\n 2.8023e-11],\n ...,\n [1.8681e-12, 4.6385e-12, 1.7741e-11, ..., 8.7402e-12, 0.0000e+00,\n 1.5037e-11],\n [2.3605e-11, 4.3548e-11, 8.2598e-11, ..., 9.4976e-11, 0.0000e+00,\n 1.0488e-10],\n [1.8798e-11, 3.3586e-13, 1.0003e-11, ..., 4.4677e-11, 0.0000e+00,\n 2.4950e-11]], device='cuda:0')" + "step": "tensor(3756.)", + "exp_avg": "tensor([[-8.7423e-07, -5.2383e-08, 5.3545e-07, ..., -7.5333e-08,\n 0.0000e+00, -6.1715e-07],\n [-8.7395e-08, 9.3032e-08, -3.2399e-07, ..., 6.2396e-07,\n 0.0000e+00, -1.0548e-06],\n [-1.5694e-06, -2.7130e-07, 2.4359e-08, ..., -4.0639e-06,\n 0.0000e+00, 3.0444e-07],\n ...,\n [ 3.2692e-09, 1.2632e-07, 3.7692e-08, ..., 2.9991e-09,\n 0.0000e+00, 5.6342e-08],\n [ 3.9501e-07, -1.6667e-06, -1.5671e-06, ..., 8.9747e-07,\n 0.0000e+00, 1.2149e-06],\n [ 1.0849e-08, -8.3909e-11, -3.4838e-06, ..., -2.1884e-07,\n 0.0000e+00, -2.2243e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[3.0486e-11, 2.8112e-11, 8.4956e-12, ..., 7.0286e-12, 0.0000e+00,\n 2.7041e-11],\n [9.6333e-12, 4.3216e-11, 7.6159e-12, ..., 1.0726e-11, 0.0000e+00,\n 7.4497e-11],\n [8.0672e-12, 1.7191e-12, 3.8055e-11, ..., 7.9303e-11, 0.0000e+00,\n 1.7529e-11],\n ...,\n [7.7232e-13, 1.8286e-12, 1.3275e-11, ..., 5.8410e-12, 0.0000e+00,\n 9.5213e-12],\n [1.5686e-11, 2.5534e-11, 5.1617e-11, ..., 9.3988e-11, 0.0000e+00,\n 7.2957e-11],\n [1.5464e-11, 2.3095e-13, 7.0624e-12, ..., 2.8732e-11, 0.0000e+00,\n 1.2743e-11]], device='cuda:0')" }, "12": { - "step": "tensor(2504.)", - "exp_avg": "tensor([ 2.9945e-05, -5.4609e-06, 9.1374e-06, ..., -2.6447e-05,\n -2.7284e-06, 3.6155e-06], device='cuda:0')", - "exp_avg_sq": "tensor([3.4838e-09, 2.6143e-09, 2.7865e-09, ..., 2.2366e-09, 4.0454e-09,\n 2.6550e-09], device='cuda:0')" + "step": "tensor(3756.)", + "exp_avg": "tensor([ 8.7109e-07, 9.0327e-07, 9.5557e-06, ..., -2.4036e-05,\n -1.6593e-05, -1.9980e-05], device='cuda:0')", + "exp_avg_sq": "tensor([2.1482e-09, 1.8897e-09, 1.7735e-09, ..., 1.7096e-09, 2.5549e-09,\n 1.7763e-09], device='cuda:0')" }, "13": { - "step": "tensor(2504.)", - "exp_avg": "tensor([[-2.7803e-07, -3.7306e-08, 1.1099e-06, ..., 4.3446e-07,\n -2.4451e-07, 5.3858e-07],\n [ 1.9896e-06, -2.9311e-07, -1.4487e-06, ..., -7.3788e-08,\n 1.1764e-07, 3.2649e-07],\n [-2.2956e-07, -4.2627e-08, 5.5378e-07, ..., -1.2870e-07,\n 1.8180e-07, 5.0446e-07],\n ...,\n [-4.1148e-07, -4.8784e-07, -1.2887e-06, ..., -7.5846e-07,\n -1.7294e-07, 2.4146e-07],\n [ 3.0531e-07, -1.0245e-06, 3.2995e-07, ..., -8.1471e-08,\n -2.0297e-07, -7.3061e-07],\n [-1.2335e-06, 4.5136e-07, 5.9361e-08, ..., 1.0099e-07,\n 3.9649e-07, 2.4582e-07]], device='cuda:0')", - "exp_avg_sq": "tensor([[2.1799e-12, 1.9794e-12, 2.9180e-12, ..., 3.1936e-12, 4.0457e-12,\n 6.2768e-12],\n [6.2242e-12, 3.3000e-12, 8.1354e-12, ..., 7.6155e-12, 8.7068e-12,\n 1.2365e-11],\n [5.2473e-12, 5.0067e-12, 7.2966e-12, ..., 8.2098e-12, 8.7280e-12,\n 3.6841e-12],\n ...,\n [6.8874e-12, 5.9269e-12, 9.9386e-12, ..., 1.0752e-11, 9.0643e-12,\n 1.1276e-11],\n [4.8607e-12, 8.6427e-12, 4.2208e-12, ..., 3.9808e-12, 8.7024e-12,\n 1.2994e-11],\n [5.0820e-12, 3.4031e-12, 7.2314e-12, ..., 8.0791e-12, 9.0745e-12,\n 5.7353e-12]], device='cuda:0')" + "step": "tensor(3756.)", + "exp_avg": "tensor([[-4.0615e-08, -7.6590e-09, 2.6700e-07, ..., -7.8619e-07,\n 1.4897e-08, 4.4327e-07],\n [-5.4131e-08, 1.1470e-07, 6.9014e-07, ..., 1.0658e-07,\n -3.8449e-07, -8.0544e-07],\n [-2.2191e-07, 1.1909e-07, 2.2803e-07, ..., 8.4710e-07,\n -1.0537e-06, -4.8787e-07],\n ...,\n [ 7.6485e-09, -1.9177e-08, 3.6948e-07, ..., -8.3713e-07,\n 1.1823e-06, 6.3953e-07],\n [-3.9626e-07, 2.4206e-07, -8.0434e-08, ..., -9.7338e-07,\n 6.8367e-07, -5.0031e-07],\n [-4.9867e-07, -7.1097e-08, -6.1115e-07, ..., -4.4276e-07,\n 3.7702e-07, -4.0410e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[1.2249e-12, 1.0002e-12, 1.5767e-12, ..., 2.0271e-12, 2.0718e-12,\n 3.7156e-12],\n [2.9573e-12, 1.7176e-12, 4.7034e-12, ..., 4.7428e-12, 4.7378e-12,\n 6.3935e-12],\n [2.6634e-12, 2.6115e-12, 4.4838e-12, ..., 5.0130e-12, 5.1257e-12,\n 2.0812e-12],\n ...,\n [3.5846e-12, 3.4829e-12, 5.5874e-12, ..., 4.8846e-12, 4.9540e-12,\n 6.9845e-12],\n [2.5216e-12, 4.0041e-12, 2.1589e-12, ..., 2.6732e-12, 4.3798e-12,\n 8.1408e-12],\n [2.6900e-12, 1.9710e-12, 3.6146e-12, ..., 5.7217e-12, 4.8766e-12,\n 3.3781e-12]], device='cuda:0')" }, "14": { - "step": "tensor(1252.)", - "exp_avg": "tensor([[-5.5836e-08, 5.9145e-08, 3.1468e-06, ..., 5.0826e-06,\n 0.0000e+00, 1.1467e-06],\n [-1.2405e-06, 1.2115e-08, 3.8742e-07, ..., 8.6341e-07,\n 0.0000e+00, 2.7410e-06],\n [ 1.2883e-06, 2.8321e-07, 1.7832e-07, ..., -3.5196e-07,\n 0.0000e+00, 8.1146e-08],\n ...,\n [-7.3957e-06, 3.4333e-07, 1.1083e-06, ..., -1.6878e-07,\n 0.0000e+00, 8.5872e-08],\n [ 3.9460e-08, -1.4093e-07, -1.3078e-08, ..., 1.3118e-06,\n 0.0000e+00, -1.2148e-06],\n [-6.4453e-07, 2.9878e-08, -1.1689e-06, ..., 6.4297e-06,\n 0.0000e+00, 1.5700e-06]], device='cuda:0')", - "exp_avg_sq": "tensor([[9.2732e-12, 2.8539e-12, 2.8385e-10, ..., 7.6557e-10, 0.0000e+00,\n 1.0338e-10],\n [3.4943e-11, 2.6239e-12, 1.9723e-10, ..., 5.9832e-11, 0.0000e+00,\n 5.7539e-11],\n [5.1052e-11, 9.2555e-12, 1.2074e-11, ..., 1.0524e-11, 0.0000e+00,\n 1.0740e-11],\n ...,\n [3.2242e-11, 5.1022e-12, 3.0993e-11, ..., 1.1228e-11, 0.0000e+00,\n 1.1824e-11],\n [8.6551e-12, 7.5511e-12, 9.6267e-12, ..., 6.9740e-11, 0.0000e+00,\n 3.0929e-10],\n [6.4763e-11, 2.8280e-12, 3.2130e-10, ..., 1.6215e-09, 0.0000e+00,\n 3.4952e-11]], device='cuda:0')" + "step": "tensor(2504.)", + "exp_avg": "tensor([[ 1.2320e-08, 2.8497e-08, 2.8481e-06, ..., -3.8289e-06,\n 0.0000e+00, -5.0457e-08],\n [ 2.4219e-08, 1.3136e-08, -2.4104e-06, ..., 8.7400e-08,\n 0.0000e+00, 1.7123e-06],\n [-2.5052e-07, 6.6894e-07, -1.1433e-06, ..., 9.1624e-07,\n 0.0000e+00, -8.0019e-08],\n ...,\n [-9.3269e-08, 1.3985e-08, -8.7144e-07, ..., 3.9684e-08,\n 0.0000e+00, 2.3400e-08],\n [-6.7383e-07, -3.2639e-07, 3.2539e-07, ..., 7.3459e-07,\n 0.0000e+00, 9.1112e-07],\n [-6.0822e-07, 9.1417e-09, -6.1563e-07, ..., -8.4254e-06,\n 0.0000e+00, 5.8302e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[3.9672e-12, 1.5374e-12, 1.8064e-10, ..., 5.0522e-10, 0.0000e+00,\n 4.3604e-11],\n [1.7096e-11, 8.8400e-13, 1.4141e-10, ..., 4.1273e-11, 0.0000e+00,\n 4.3009e-11],\n [3.0785e-11, 4.8158e-12, 7.1354e-12, ..., 4.5958e-12, 0.0000e+00,\n 5.2565e-12],\n ...,\n [3.1924e-11, 1.9915e-12, 1.6619e-11, ..., 9.3013e-12, 0.0000e+00,\n 9.9534e-12],\n [5.7493e-12, 2.9328e-12, 1.3822e-11, ..., 7.3988e-11, 0.0000e+00,\n 2.1592e-10],\n [2.7939e-11, 1.2076e-12, 2.0357e-10, ..., 1.1915e-09, 0.0000e+00,\n 2.4390e-11]], device='cuda:0')" }, "15": { - "step": "tensor(1252.)", - "exp_avg": "tensor([-8.1167e-06, 1.7618e-05, -1.1373e-05, ..., 1.3225e-05,\n -2.5681e-06, 9.6634e-06], device='cuda:0')", - "exp_avg_sq": "tensor([8.3071e-09, 5.5925e-09, 6.2431e-09, ..., 3.4018e-09, 4.5848e-09,\n 7.2586e-09], device='cuda:0')" + "step": "tensor(2504.)", + "exp_avg": "tensor([ 1.1307e-05, 3.2308e-06, 1.2916e-05, ..., -3.6718e-05,\n 4.1557e-07, -3.6273e-07], device='cuda:0')", + "exp_avg_sq": "tensor([4.8060e-09, 3.3538e-09, 3.9253e-09, ..., 3.1399e-09, 3.6870e-09,\n 4.8002e-09], device='cuda:0')" }, "16": { + "step": "tensor(2504.)", + "exp_avg": "tensor([[ 5.3414e-07, 4.5119e-07, -1.9630e-07, ..., 2.3438e-08,\n 1.7424e-07, -3.9592e-08],\n [-4.5617e-08, -4.4612e-08, -7.9842e-07, ..., 1.8522e-08,\n 9.8481e-08, 2.0186e-07],\n [-1.3206e-07, 1.0917e-07, -3.4518e-07, ..., 1.2383e-07,\n 8.0408e-08, 2.0186e-08],\n ...,\n [-4.7195e-07, 2.0320e-07, 1.2332e-07, ..., -1.3182e-07,\n 6.7670e-08, -1.2215e-07],\n [-2.0476e-07, -2.2044e-07, -1.8906e-07, ..., -2.1482e-08,\n 1.2997e-06, 3.2573e-07],\n [-1.1426e-08, -5.2560e-07, 4.3108e-07, ..., -1.6561e-07,\n -2.3323e-08, -1.6453e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[2.6884e-12, 2.2491e-12, 2.4067e-12, ..., 1.2575e-12, 2.2583e-12,\n 1.5742e-12],\n [3.4229e-12, 3.2604e-12, 5.4071e-12, ..., 4.0876e-12, 5.5884e-12,\n 2.7649e-12],\n [3.3258e-12, 3.5086e-12, 3.7874e-12, ..., 3.4707e-12, 1.1775e-11,\n 2.0980e-12],\n ...,\n [5.2257e-12, 4.2729e-12, 2.6095e-12, ..., 3.3140e-12, 2.8782e-12,\n 2.5719e-12],\n [3.7905e-12, 3.4517e-12, 4.3633e-12, ..., 3.2973e-12, 1.5785e-11,\n 2.0082e-12],\n [4.2450e-12, 7.2899e-12, 3.9194e-12, ..., 3.1013e-12, 5.3088e-12,\n 3.4798e-12]], device='cuda:0')" + }, + "17": { + "step": "tensor(1252.)", + "exp_avg": "tensor([[-5.7060e-08, 7.3802e-08, -1.9580e-07, ..., 1.6809e-08,\n 0.0000e+00, 1.0926e-09],\n [-1.3783e-07, 2.0248e-07, -7.2158e-07, ..., -1.1465e-06,\n 0.0000e+00, 1.8002e-06],\n [-5.1600e-07, -1.8121e-07, 4.2179e-06, ..., 2.3470e-07,\n 0.0000e+00, 6.0861e-07],\n ...,\n [ 8.4494e-08, 2.4882e-07, 2.1328e-06, ..., 2.2136e-07,\n 0.0000e+00, -1.4351e-06],\n [ 1.4450e-06, 2.6426e-06, 2.7397e-06, ..., 4.0668e-06,\n 0.0000e+00, 7.4458e-07],\n [-1.3917e-06, 6.2669e-07, 6.1212e-07, ..., -5.7251e-07,\n 0.0000e+00, -7.5933e-08]], device='cuda:0')", + "exp_avg_sq": "tensor([[1.5703e-11, 1.7458e-12, 1.0888e-11, ..., 3.0022e-12, 0.0000e+00,\n 4.6481e-12],\n [1.3639e-11, 8.6552e-12, 6.0112e-11, ..., 4.3898e-10, 0.0000e+00,\n 4.6581e-10],\n [9.9996e-11, 3.3545e-12, 7.5301e-11, ..., 7.4812e-12, 0.0000e+00,\n 2.0622e-11],\n ...,\n [5.1079e-11, 2.6154e-11, 3.0075e-10, ..., 4.7013e-11, 0.0000e+00,\n 5.1699e-11],\n [9.7127e-11, 6.3865e-11, 8.1712e-11, ..., 1.0127e-10, 0.0000e+00,\n 5.0273e-11],\n [8.6476e-12, 5.5097e-12, 2.5212e-10, ..., 6.4465e-11, 0.0000e+00,\n 1.5434e-11]], device='cuda:0')" + }, + "18": { + "step": "tensor(1252.)", + "exp_avg": "tensor([-3.4119e-06, -6.7615e-06, 3.0917e-06, ..., 4.1757e-06,\n 2.3947e-05, 6.9000e-06], device='cuda:0')", + "exp_avg_sq": "tensor([3.5767e-09, 6.3875e-09, 4.6260e-09, ..., 5.9041e-09, 8.6187e-09,\n 7.0263e-09], device='cuda:0')" + }, + "19": { "step": "tensor(1252.)", - "exp_avg": "tensor([[-1.5481e-07, 2.8678e-07, 5.5108e-07, ..., -2.0205e-07,\n 1.5189e-08, 2.9531e-07],\n [-3.4451e-08, 6.0454e-08, 6.9110e-07, ..., -2.5115e-07,\n -2.3807e-07, -1.4570e-07],\n [ 5.8860e-09, 3.2474e-07, -5.4785e-07, ..., 3.3104e-08,\n 1.4735e-08, 3.6657e-07],\n ...,\n [ 9.7897e-07, -7.7870e-07, 1.5313e-07, ..., -5.1950e-07,\n 9.6189e-08, 4.9757e-07],\n [-1.3136e-06, -5.9382e-07, 3.5333e-07, ..., -1.2665e-08,\n 2.4294e-07, 3.7891e-07],\n [-1.3422e-07, 1.2921e-06, -6.8474e-07, ..., 1.9939e-07,\n 2.1495e-07, -3.0081e-07]], device='cuda:0')", - "exp_avg_sq": "tensor([[5.1165e-12, 4.8524e-12, 4.6993e-12, ..., 2.6035e-12, 3.6581e-12,\n 3.5904e-12],\n [6.9779e-12, 6.6144e-12, 1.1463e-11, ..., 6.5572e-12, 1.0694e-11,\n 5.1223e-12],\n [6.9161e-12, 7.6088e-12, 7.9918e-12, ..., 8.8075e-12, 3.1320e-11,\n 4.0475e-12],\n ...,\n [9.1454e-12, 9.5437e-12, 5.1974e-12, ..., 6.5395e-12, 5.4946e-12,\n 4.7305e-12],\n [7.3494e-12, 6.6373e-12, 8.6532e-12, ..., 5.7239e-12, 2.3864e-11,\n 3.8321e-12],\n [8.5844e-12, 1.5904e-11, 6.9594e-12, ..., 7.6194e-12, 8.6619e-12,\n 6.7029e-12]], device='cuda:0')" + "exp_avg": "tensor([[-1.4149e-07, 4.2814e-07, -2.5375e-07, ..., 6.3136e-08,\n -2.1584e-07, 1.5275e-07],\n [-3.0422e-07, -4.9220e-08, -3.8916e-07, ..., 5.1340e-08,\n -5.8419e-07, 5.7229e-08],\n [-3.7718e-07, 2.0629e-07, 4.5861e-07, ..., 6.3865e-07,\n 2.6689e-07, -3.1508e-07],\n ...,\n [ 1.5038e-07, 2.0739e-07, 4.5431e-08, ..., -3.0124e-07,\n -2.4864e-07, 3.3370e-07],\n [-1.5353e-07, -6.7129e-09, 2.7592e-07, ..., 3.0316e-07,\n -2.5395e-08, -4.5650e-07],\n [-1.2458e-07, -2.3178e-07, 3.5717e-07, ..., 8.6123e-08,\n 3.5469e-07, -1.8134e-07]], device='cuda:0')", + "exp_avg_sq": "tensor([[3.8504e-12, 5.0707e-12, 1.7906e-12, ..., 2.7861e-12, 4.1834e-12,\n 5.9459e-12],\n [2.4100e-12, 1.1127e-11, 4.0108e-12, ..., 5.3189e-12, 7.1865e-12,\n 6.5747e-12],\n [3.6657e-12, 1.1867e-11, 5.1476e-12, ..., 7.3983e-12, 9.1723e-12,\n 9.1207e-12],\n ...,\n [7.5244e-12, 5.3572e-12, 3.5871e-12, ..., 6.1545e-12, 7.0339e-12,\n 5.9241e-12],\n [4.5302e-12, 1.0013e-11, 4.6293e-12, ..., 5.1168e-12, 1.0642e-11,\n 1.1755e-11],\n [1.6663e-12, 4.1387e-12, 1.3617e-11, ..., 7.2629e-12, 8.2895e-12,\n 5.6313e-12]], device='cuda:0')" } }, "param_groups": [ { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "shared", "betas": [ 0.9, @@ -257,7 +272,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_256", "betas": [ 0.9, @@ -280,7 +295,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_512", "betas": [ 0.9, @@ -303,7 +318,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_768", "betas": [ 0.9, @@ -326,7 +341,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_1024", "betas": [ 0.9, @@ -349,7 +364,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_1280", "betas": [ 0.9, @@ -372,7 +387,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_1536", "betas": [ 0.9, @@ -395,7 +410,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_1792", "betas": [ 0.9, @@ -418,7 +433,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_2048", "betas": [ 0.9, @@ -441,7 +456,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_2304", "betas": [ 0.9, @@ -464,7 +479,7 @@ ] }, { - "lr": 0.005000500000000001, + "lr": 0.0034555695366224513, "name": "scale_2560", "betas": [ 0.9, @@ -487,7 +502,7 @@ ] }, { - "lr": 0.0025005, + "lr": 0.001728112022559819, "name": "fusion", "betas": [ 0.9, @@ -543,7 +558,7 @@ "T_i": 10, "T_mult": 2, "eta_min": 1e-06, - "T_cur": 5, + "T_cur": 6, "base_lrs": [ 0.01, 0.01, @@ -558,34 +573,35 @@ 0.01, 0.005 ], - "last_epoch": 5, + "last_epoch": 6, "_step_count": 0, "_is_initial": false, "_get_lr_called_within_step": false, "_last_lr": [ - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.005000500000000001, - 0.0025005 + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.0034555695366224513, + 0.001728112022559819 ] }, "metrics": { - "best_val_acc": 81.822, - "best_epoch": 4, + "best_val_acc": 82.1, + "best_epoch": 5, "scale_accuracies": { - "256": 81.822, - "512": 81.94, - "768": 81.974, - "1024": 81.794, - "1280": 81.448 + "256": 82.1, + "512": 82.336, + "768": 82.284, + "1024": 82.008, + "1280": 81.898, + "1536": 81.582 } }, "train_config": {