| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.5, |
| "eval_steps": 10000, |
| "global_step": 100000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "grad_norm": 5.09375, |
| "learning_rate": 1.0040201005025126e-07, |
| "loss": 0.9541, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 5.0625, |
| "learning_rate": 1.0030150753768845e-07, |
| "loss": 1.0027, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 4.5625, |
| "learning_rate": 1.0020100502512563e-07, |
| "loss": 0.9509, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 5.09375, |
| "learning_rate": 1.0010050251256282e-07, |
| "loss": 0.9602, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 5.03125, |
| "learning_rate": 1e-07, |
| "loss": 0.9338, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.01, |
| "grad_norm": 6.34375, |
| "learning_rate": 9.989949748743718e-08, |
| "loss": 0.9006, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 6.09375, |
| "learning_rate": 9.979899497487437e-08, |
| "loss": 0.9381, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 4.40625, |
| "learning_rate": 9.969849246231156e-08, |
| "loss": 0.931, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.02, |
| "grad_norm": 6.4375, |
| "learning_rate": 9.959798994974874e-08, |
| "loss": 0.9787, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 8.25, |
| "learning_rate": 9.949748743718593e-08, |
| "loss": 0.9649, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 5.78125, |
| "learning_rate": 9.939698492462312e-08, |
| "loss": 0.9302, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 5.375, |
| "learning_rate": 9.92964824120603e-08, |
| "loss": 0.891, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.03, |
| "grad_norm": 3.078125, |
| "learning_rate": 9.919597989949749e-08, |
| "loss": 0.9733, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 12.25, |
| "learning_rate": 9.909547738693468e-08, |
| "loss": 0.8859, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 5.4375, |
| "learning_rate": 9.899497487437185e-08, |
| "loss": 0.9646, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 8.5, |
| "learning_rate": 9.889447236180904e-08, |
| "loss": 0.9013, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 12.5, |
| "learning_rate": 9.879396984924623e-08, |
| "loss": 0.9703, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 6.03125, |
| "learning_rate": 9.869346733668342e-08, |
| "loss": 1.035, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 7.5625, |
| "learning_rate": 9.85929648241206e-08, |
| "loss": 0.9458, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 3.78125, |
| "learning_rate": 9.849246231155779e-08, |
| "loss": 0.9454, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.05, |
| "grad_norm": 4.40625, |
| "learning_rate": 9.839195979899498e-08, |
| "loss": 0.9185, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 5.375, |
| "learning_rate": 9.829145728643217e-08, |
| "loss": 0.9754, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 9.625, |
| "learning_rate": 9.819095477386935e-08, |
| "loss": 0.9614, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 6.46875, |
| "learning_rate": 9.809045226130653e-08, |
| "loss": 0.9622, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 8.5, |
| "learning_rate": 9.798994974874371e-08, |
| "loss": 0.9266, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 6.125, |
| "learning_rate": 9.78894472361809e-08, |
| "loss": 0.9662, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 7.375, |
| "learning_rate": 9.778894472361809e-08, |
| "loss": 0.9502, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 5.375, |
| "learning_rate": 9.768844221105528e-08, |
| "loss": 0.9098, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 8.4375, |
| "learning_rate": 9.758793969849246e-08, |
| "loss": 0.9406, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.07, |
| "grad_norm": 5.40625, |
| "learning_rate": 9.748743718592965e-08, |
| "loss": 0.925, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 5.03125, |
| "learning_rate": 9.738693467336684e-08, |
| "loss": 0.9583, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 6.03125, |
| "learning_rate": 9.728643216080402e-08, |
| "loss": 0.9582, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 6.1875, |
| "learning_rate": 9.718592964824121e-08, |
| "loss": 0.9629, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 4.03125, |
| "learning_rate": 9.708542713567839e-08, |
| "loss": 0.9405, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 6.375, |
| "learning_rate": 9.698492462311557e-08, |
| "loss": 0.9867, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 6.90625, |
| "learning_rate": 9.688442211055276e-08, |
| "loss": 0.9221, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.09, |
| "grad_norm": 5.09375, |
| "learning_rate": 9.678391959798995e-08, |
| "loss": 0.9397, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 5.34375, |
| "learning_rate": 9.668341708542713e-08, |
| "loss": 0.9511, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 4.96875, |
| "learning_rate": 9.658291457286432e-08, |
| "loss": 0.9714, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 5.15625, |
| "learning_rate": 9.648241206030151e-08, |
| "loss": 1.0139, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 7.875, |
| "learning_rate": 9.63819095477387e-08, |
| "loss": 0.9871, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 5.4375, |
| "learning_rate": 9.628140703517588e-08, |
| "loss": 0.9687, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 6.78125, |
| "learning_rate": 9.618090452261306e-08, |
| "loss": 1.0132, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 6.40625, |
| "learning_rate": 9.608040201005024e-08, |
| "loss": 0.9381, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.11, |
| "grad_norm": 10.375, |
| "learning_rate": 9.597989949748743e-08, |
| "loss": 0.9664, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 6.53125, |
| "learning_rate": 9.587939698492462e-08, |
| "loss": 1.0202, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 8.25, |
| "learning_rate": 9.57788944723618e-08, |
| "loss": 0.9774, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 7.78125, |
| "learning_rate": 9.567839195979899e-08, |
| "loss": 0.9352, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 6.8125, |
| "learning_rate": 9.557788944723618e-08, |
| "loss": 0.9894, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 7.46875, |
| "learning_rate": 9.547738693467337e-08, |
| "loss": 0.9708, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 6.46875, |
| "learning_rate": 9.537688442211055e-08, |
| "loss": 0.9545, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 6.34375, |
| "learning_rate": 9.527638190954774e-08, |
| "loss": 0.9299, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.13, |
| "grad_norm": 3.890625, |
| "learning_rate": 9.517587939698492e-08, |
| "loss": 1.0221, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 6.3125, |
| "learning_rate": 9.50753768844221e-08, |
| "loss": 1.0046, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 5.59375, |
| "learning_rate": 9.497487437185929e-08, |
| "loss": 0.9403, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 6.375, |
| "learning_rate": 9.487437185929648e-08, |
| "loss": 0.8997, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 6.90625, |
| "learning_rate": 9.477386934673366e-08, |
| "loss": 1.0223, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 11.0625, |
| "learning_rate": 9.467336683417085e-08, |
| "loss": 0.9861, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 7.5625, |
| "learning_rate": 9.457286432160804e-08, |
| "loss": 0.8784, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 8.75, |
| "learning_rate": 9.447236180904523e-08, |
| "loss": 0.9643, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 7.6875, |
| "learning_rate": 9.437185929648241e-08, |
| "loss": 0.9867, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.15, |
| "grad_norm": 4.625, |
| "learning_rate": 9.42713567839196e-08, |
| "loss": 0.9421, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 11.5, |
| "learning_rate": 9.417085427135679e-08, |
| "loss": 1.0064, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 7.0, |
| "learning_rate": 9.407035175879398e-08, |
| "loss": 0.9879, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 5.5625, |
| "learning_rate": 9.396984924623116e-08, |
| "loss": 0.941, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 6.8125, |
| "learning_rate": 9.386934673366835e-08, |
| "loss": 0.9675, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 7.59375, |
| "learning_rate": 9.376884422110554e-08, |
| "loss": 1.008, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 4.5625, |
| "learning_rate": 9.366834170854272e-08, |
| "loss": 0.9385, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 5.875, |
| "learning_rate": 9.356783919597991e-08, |
| "loss": 0.9606, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.17, |
| "grad_norm": 6.5625, |
| "learning_rate": 9.34673366834171e-08, |
| "loss": 0.9215, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 8.6875, |
| "learning_rate": 9.336683417085427e-08, |
| "loss": 0.9214, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 5.8125, |
| "learning_rate": 9.326633165829146e-08, |
| "loss": 0.9777, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 5.59375, |
| "learning_rate": 9.316582914572865e-08, |
| "loss": 1.0172, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 5.375, |
| "learning_rate": 9.306532663316583e-08, |
| "loss": 0.9319, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 6.9375, |
| "learning_rate": 9.296482412060302e-08, |
| "loss": 0.9762, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 7.4375, |
| "learning_rate": 9.286432160804021e-08, |
| "loss": 0.9869, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.19, |
| "grad_norm": 5.71875, |
| "learning_rate": 9.27638190954774e-08, |
| "loss": 0.973, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 6.90625, |
| "learning_rate": 9.266331658291458e-08, |
| "loss": 1.0055, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 5.21875, |
| "learning_rate": 9.256281407035177e-08, |
| "loss": 0.893, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 5.46875, |
| "learning_rate": 9.246231155778894e-08, |
| "loss": 0.9886, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 6.15625, |
| "learning_rate": 9.236180904522613e-08, |
| "loss": 0.9345, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 4.28125, |
| "learning_rate": 9.226130653266332e-08, |
| "loss": 0.9994, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 7.8125, |
| "learning_rate": 9.21608040201005e-08, |
| "loss": 0.8965, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 6.15625, |
| "learning_rate": 9.20603015075377e-08, |
| "loss": 0.9273, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 6.5, |
| "learning_rate": 9.195979899497488e-08, |
| "loss": 0.9655, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.21, |
| "grad_norm": 6.15625, |
| "learning_rate": 9.185929648241207e-08, |
| "loss": 0.9608, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 6.1875, |
| "learning_rate": 9.175879396984926e-08, |
| "loss": 0.9906, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 5.78125, |
| "learning_rate": 9.165829145728644e-08, |
| "loss": 0.9592, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 6.21875, |
| "learning_rate": 9.155778894472363e-08, |
| "loss": 0.9491, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 11.4375, |
| "learning_rate": 9.14572864321608e-08, |
| "loss": 0.9509, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 6.1875, |
| "learning_rate": 9.135678391959799e-08, |
| "loss": 0.9575, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 5.15625, |
| "learning_rate": 9.125628140703518e-08, |
| "loss": 0.9879, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 4.5, |
| "learning_rate": 9.115577889447237e-08, |
| "loss": 0.9281, |
| "step": 9300 |
| }, |
| { |
| "epoch": 0.23, |
| "grad_norm": 7.625, |
| "learning_rate": 9.105527638190955e-08, |
| "loss": 0.9937, |
| "step": 9400 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 6.21875, |
| "learning_rate": 9.095477386934674e-08, |
| "loss": 0.9812, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 4.625, |
| "learning_rate": 9.085427135678393e-08, |
| "loss": 0.9643, |
| "step": 9600 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 9.625, |
| "learning_rate": 9.075376884422111e-08, |
| "loss": 1.0037, |
| "step": 9700 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 4.15625, |
| "learning_rate": 9.06532663316583e-08, |
| "loss": 1.0041, |
| "step": 9800 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 4.40625, |
| "learning_rate": 9.055276381909548e-08, |
| "loss": 0.9367, |
| "step": 9900 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 4.8125, |
| "learning_rate": 9.045226130653266e-08, |
| "loss": 0.9733, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.25, |
| "eval_loss": 1.025413990020752, |
| "eval_runtime": 60.1, |
| "eval_samples_per_second": 16.639, |
| "eval_steps_per_second": 4.16, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 5.1875, |
| "learning_rate": 9.035175879396985e-08, |
| "loss": 0.9202, |
| "step": 10100 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 5.78125, |
| "learning_rate": 9.025125628140704e-08, |
| "loss": 0.9826, |
| "step": 10200 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 9.1875, |
| "learning_rate": 9.015075376884422e-08, |
| "loss": 0.9588, |
| "step": 10300 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 7.71875, |
| "learning_rate": 9.005025125628141e-08, |
| "loss": 0.9928, |
| "step": 10400 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 7.65625, |
| "learning_rate": 8.99497487437186e-08, |
| "loss": 0.9502, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 7.59375, |
| "learning_rate": 8.984924623115579e-08, |
| "loss": 0.9946, |
| "step": 10600 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 7.0, |
| "learning_rate": 8.974874371859297e-08, |
| "loss": 0.9375, |
| "step": 10700 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 6.53125, |
| "learning_rate": 8.964824120603016e-08, |
| "loss": 0.9633, |
| "step": 10800 |
| }, |
| { |
| "epoch": 0.27, |
| "grad_norm": 6.625, |
| "learning_rate": 8.954773869346733e-08, |
| "loss": 0.9744, |
| "step": 10900 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 4.09375, |
| "learning_rate": 8.944723618090452e-08, |
| "loss": 0.9602, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 6.5, |
| "learning_rate": 8.934673366834171e-08, |
| "loss": 0.96, |
| "step": 11100 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 7.0625, |
| "learning_rate": 8.92462311557789e-08, |
| "loss": 0.8934, |
| "step": 11200 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 8.1875, |
| "learning_rate": 8.914572864321608e-08, |
| "loss": 0.9041, |
| "step": 11300 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 6.03125, |
| "learning_rate": 8.904522613065327e-08, |
| "loss": 0.9754, |
| "step": 11400 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 11.5, |
| "learning_rate": 8.894472361809046e-08, |
| "loss": 0.9461, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 2.84375, |
| "learning_rate": 8.884422110552765e-08, |
| "loss": 0.9161, |
| "step": 11600 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 8.25, |
| "learning_rate": 8.874371859296483e-08, |
| "loss": 0.9703, |
| "step": 11700 |
| }, |
| { |
| "epoch": 0.29, |
| "grad_norm": 5.09375, |
| "learning_rate": 8.8643216080402e-08, |
| "loss": 1.0176, |
| "step": 11800 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 5.90625, |
| "learning_rate": 8.85427135678392e-08, |
| "loss": 0.9929, |
| "step": 11900 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 4.90625, |
| "learning_rate": 8.844221105527638e-08, |
| "loss": 1.0158, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 4.75, |
| "learning_rate": 8.834170854271357e-08, |
| "loss": 0.856, |
| "step": 12100 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 5.75, |
| "learning_rate": 8.824120603015076e-08, |
| "loss": 0.9228, |
| "step": 12200 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 6.5625, |
| "learning_rate": 8.814070351758794e-08, |
| "loss": 0.9731, |
| "step": 12300 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 6.3125, |
| "learning_rate": 8.804020100502513e-08, |
| "loss": 0.9674, |
| "step": 12400 |
| }, |
| { |
| "epoch": 0.31, |
| "grad_norm": 7.15625, |
| "learning_rate": 8.793969849246232e-08, |
| "loss": 0.9545, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 6.0, |
| "learning_rate": 8.78391959798995e-08, |
| "loss": 0.9233, |
| "step": 12600 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 11.8125, |
| "learning_rate": 8.773869346733669e-08, |
| "loss": 1.0133, |
| "step": 12700 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 4.5, |
| "learning_rate": 8.763819095477387e-08, |
| "loss": 0.9488, |
| "step": 12800 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 6.3125, |
| "learning_rate": 8.753768844221105e-08, |
| "loss": 0.9521, |
| "step": 12900 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 7.1875, |
| "learning_rate": 8.743718592964824e-08, |
| "loss": 0.9728, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 7.875, |
| "learning_rate": 8.733668341708543e-08, |
| "loss": 0.9632, |
| "step": 13100 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 6.3125, |
| "learning_rate": 8.723618090452261e-08, |
| "loss": 0.9788, |
| "step": 13200 |
| }, |
| { |
| "epoch": 0.33, |
| "grad_norm": 4.9375, |
| "learning_rate": 8.71356783919598e-08, |
| "loss": 0.9722, |
| "step": 13300 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 6.21875, |
| "learning_rate": 8.703517587939699e-08, |
| "loss": 0.9887, |
| "step": 13400 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 6.5, |
| "learning_rate": 8.693467336683418e-08, |
| "loss": 0.9568, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 7.65625, |
| "learning_rate": 8.683417085427136e-08, |
| "loss": 0.9804, |
| "step": 13600 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 6.125, |
| "learning_rate": 8.673366834170854e-08, |
| "loss": 0.9415, |
| "step": 13700 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 6.21875, |
| "learning_rate": 8.663316582914572e-08, |
| "loss": 0.9515, |
| "step": 13800 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 5.875, |
| "learning_rate": 8.653266331658291e-08, |
| "loss": 0.9264, |
| "step": 13900 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 7.25, |
| "learning_rate": 8.64321608040201e-08, |
| "loss": 0.9275, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 7.75, |
| "learning_rate": 8.633165829145729e-08, |
| "loss": 0.8601, |
| "step": 14100 |
| }, |
| { |
| "epoch": 0.35, |
| "grad_norm": 8.75, |
| "learning_rate": 8.623115577889447e-08, |
| "loss": 0.991, |
| "step": 14200 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 6.78125, |
| "learning_rate": 8.613065326633166e-08, |
| "loss": 0.9526, |
| "step": 14300 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 5.6875, |
| "learning_rate": 8.603015075376885e-08, |
| "loss": 1.0127, |
| "step": 14400 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 8.3125, |
| "learning_rate": 8.592964824120603e-08, |
| "loss": 0.9497, |
| "step": 14500 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 4.0, |
| "learning_rate": 8.582914572864322e-08, |
| "loss": 0.976, |
| "step": 14600 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 6.40625, |
| "learning_rate": 8.57286432160804e-08, |
| "loss": 0.915, |
| "step": 14700 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 4.875, |
| "learning_rate": 8.562814070351758e-08, |
| "loss": 0.9379, |
| "step": 14800 |
| }, |
| { |
| "epoch": 0.37, |
| "grad_norm": 6.03125, |
| "learning_rate": 8.552763819095477e-08, |
| "loss": 0.9729, |
| "step": 14900 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 5.9375, |
| "learning_rate": 8.542713567839196e-08, |
| "loss": 0.9725, |
| "step": 15000 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 7.3125, |
| "learning_rate": 8.532663316582914e-08, |
| "loss": 0.9677, |
| "step": 15100 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 5.71875, |
| "learning_rate": 8.522613065326633e-08, |
| "loss": 0.977, |
| "step": 15200 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 7.0, |
| "learning_rate": 8.512562814070352e-08, |
| "loss": 0.9989, |
| "step": 15300 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 5.78125, |
| "learning_rate": 8.502512562814071e-08, |
| "loss": 0.9648, |
| "step": 15400 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 13.5625, |
| "learning_rate": 8.49246231155779e-08, |
| "loss": 1.042, |
| "step": 15500 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 4.96875, |
| "learning_rate": 8.482412060301508e-08, |
| "loss": 0.9511, |
| "step": 15600 |
| }, |
| { |
| "epoch": 0.39, |
| "grad_norm": 5.15625, |
| "learning_rate": 8.472361809045226e-08, |
| "loss": 0.8953, |
| "step": 15700 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 8.875, |
| "learning_rate": 8.462311557788944e-08, |
| "loss": 0.968, |
| "step": 15800 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 7.71875, |
| "learning_rate": 8.452261306532663e-08, |
| "loss": 0.9539, |
| "step": 15900 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 7.28125, |
| "learning_rate": 8.442211055276382e-08, |
| "loss": 0.9637, |
| "step": 16000 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 7.625, |
| "learning_rate": 8.4321608040201e-08, |
| "loss": 0.8945, |
| "step": 16100 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 6.40625, |
| "learning_rate": 8.422110552763819e-08, |
| "loss": 0.9473, |
| "step": 16200 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 5.59375, |
| "learning_rate": 8.412060301507538e-08, |
| "loss": 0.9729, |
| "step": 16300 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 6.125, |
| "learning_rate": 8.402010050251257e-08, |
| "loss": 1.0077, |
| "step": 16400 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 7.5, |
| "learning_rate": 8.391959798994975e-08, |
| "loss": 1.012, |
| "step": 16500 |
| }, |
| { |
| "epoch": 0.41, |
| "grad_norm": 7.09375, |
| "learning_rate": 8.381909547738693e-08, |
| "loss": 1.0129, |
| "step": 16600 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 10.125, |
| "learning_rate": 8.371859296482411e-08, |
| "loss": 0.9403, |
| "step": 16700 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 7.125, |
| "learning_rate": 8.36180904522613e-08, |
| "loss": 0.9256, |
| "step": 16800 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 6.53125, |
| "learning_rate": 8.351758793969849e-08, |
| "loss": 1.0053, |
| "step": 16900 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 9.125, |
| "learning_rate": 8.341708542713568e-08, |
| "loss": 0.9406, |
| "step": 17000 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 9.5, |
| "learning_rate": 8.331658291457286e-08, |
| "loss": 0.9271, |
| "step": 17100 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 6.71875, |
| "learning_rate": 8.321608040201005e-08, |
| "loss": 0.9238, |
| "step": 17200 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 5.59375, |
| "learning_rate": 8.311557788944724e-08, |
| "loss": 0.9762, |
| "step": 17300 |
| }, |
| { |
| "epoch": 0.43, |
| "grad_norm": 7.15625, |
| "learning_rate": 8.301507537688442e-08, |
| "loss": 0.9427, |
| "step": 17400 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 6.09375, |
| "learning_rate": 8.291457286432161e-08, |
| "loss": 0.9835, |
| "step": 17500 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 5.9375, |
| "learning_rate": 8.281407035175879e-08, |
| "loss": 0.9718, |
| "step": 17600 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 6.03125, |
| "learning_rate": 8.271356783919597e-08, |
| "loss": 0.9812, |
| "step": 17700 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 6.40625, |
| "learning_rate": 8.261306532663316e-08, |
| "loss": 0.9781, |
| "step": 17800 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 5.0, |
| "learning_rate": 8.251256281407035e-08, |
| "loss": 0.9996, |
| "step": 17900 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 6.625, |
| "learning_rate": 8.241206030150753e-08, |
| "loss": 0.9603, |
| "step": 18000 |
| }, |
| { |
| "epoch": 0.45, |
| "grad_norm": 7.125, |
| "learning_rate": 8.231155778894472e-08, |
| "loss": 0.9493, |
| "step": 18100 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 4.53125, |
| "learning_rate": 8.221105527638191e-08, |
| "loss": 0.9126, |
| "step": 18200 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 6.0, |
| "learning_rate": 8.21105527638191e-08, |
| "loss": 0.9765, |
| "step": 18300 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 10.625, |
| "learning_rate": 8.201005025125628e-08, |
| "loss": 0.9804, |
| "step": 18400 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 5.875, |
| "learning_rate": 8.190954773869346e-08, |
| "loss": 1.0446, |
| "step": 18500 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 7.1875, |
| "learning_rate": 8.180904522613064e-08, |
| "loss": 0.9757, |
| "step": 18600 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 6.34375, |
| "learning_rate": 8.170854271356783e-08, |
| "loss": 1.0207, |
| "step": 18700 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 9.75, |
| "learning_rate": 8.160804020100503e-08, |
| "loss": 1.002, |
| "step": 18800 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 8.8125, |
| "learning_rate": 8.150753768844222e-08, |
| "loss": 0.9783, |
| "step": 18900 |
| }, |
| { |
| "epoch": 0.47, |
| "grad_norm": 7.125, |
| "learning_rate": 8.140703517587941e-08, |
| "loss": 0.9591, |
| "step": 19000 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 5.46875, |
| "learning_rate": 8.13065326633166e-08, |
| "loss": 0.9886, |
| "step": 19100 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 9.5, |
| "learning_rate": 8.120603015075378e-08, |
| "loss": 0.9721, |
| "step": 19200 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 4.59375, |
| "learning_rate": 8.110552763819096e-08, |
| "loss": 1.0216, |
| "step": 19300 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 6.8125, |
| "learning_rate": 8.100502512562814e-08, |
| "loss": 0.9477, |
| "step": 19400 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 7.25, |
| "learning_rate": 8.090452261306533e-08, |
| "loss": 0.9425, |
| "step": 19500 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 5.625, |
| "learning_rate": 8.080402010050252e-08, |
| "loss": 1.0231, |
| "step": 19600 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 5.25, |
| "learning_rate": 8.07035175879397e-08, |
| "loss": 0.9594, |
| "step": 19700 |
| }, |
| { |
| "epoch": 0.49, |
| "grad_norm": 4.46875, |
| "learning_rate": 8.060301507537689e-08, |
| "loss": 0.9909, |
| "step": 19800 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 14.0, |
| "learning_rate": 8.050251256281408e-08, |
| "loss": 0.9236, |
| "step": 19900 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 6.09375, |
| "learning_rate": 8.040201005025127e-08, |
| "loss": 1.0047, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.5, |
| "eval_loss": 1.0322586297988892, |
| "eval_runtime": 60.0446, |
| "eval_samples_per_second": 16.654, |
| "eval_steps_per_second": 4.164, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 7.875, |
| "learning_rate": 8.030150753768845e-08, |
| "loss": 1.0004, |
| "step": 20100 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 7.21875, |
| "learning_rate": 8.020100502512564e-08, |
| "loss": 1.0313, |
| "step": 20200 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 7.40625, |
| "learning_rate": 8.010050251256281e-08, |
| "loss": 0.9795, |
| "step": 20300 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 9.0625, |
| "learning_rate": 8e-08, |
| "loss": 1.0003, |
| "step": 20400 |
| }, |
| { |
| "epoch": 0.51, |
| "grad_norm": 10.8125, |
| "learning_rate": 7.989949748743719e-08, |
| "loss": 0.9945, |
| "step": 20500 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 7.96875, |
| "learning_rate": 7.979899497487438e-08, |
| "loss": 1.0274, |
| "step": 20600 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 5.71875, |
| "learning_rate": 7.969849246231156e-08, |
| "loss": 0.9918, |
| "step": 20700 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 5.0, |
| "learning_rate": 7.959798994974875e-08, |
| "loss": 0.9931, |
| "step": 20800 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 6.4375, |
| "learning_rate": 7.949748743718594e-08, |
| "loss": 1.0109, |
| "step": 20900 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 3.265625, |
| "learning_rate": 7.939698492462313e-08, |
| "loss": 0.9392, |
| "step": 21000 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 12.875, |
| "learning_rate": 7.929648241206031e-08, |
| "loss": 0.9646, |
| "step": 21100 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 6.65625, |
| "learning_rate": 7.919597989949749e-08, |
| "loss": 0.9456, |
| "step": 21200 |
| }, |
| { |
| "epoch": 0.53, |
| "grad_norm": 6.5, |
| "learning_rate": 7.909547738693467e-08, |
| "loss": 0.9637, |
| "step": 21300 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 6.84375, |
| "learning_rate": 7.899497487437186e-08, |
| "loss": 1.0326, |
| "step": 21400 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 10.5625, |
| "learning_rate": 7.889447236180905e-08, |
| "loss": 0.972, |
| "step": 21500 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 8.75, |
| "learning_rate": 7.879396984924624e-08, |
| "loss": 0.9322, |
| "step": 21600 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 6.84375, |
| "learning_rate": 7.869346733668342e-08, |
| "loss": 1.0037, |
| "step": 21700 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 7.46875, |
| "learning_rate": 7.859296482412061e-08, |
| "loss": 1.0428, |
| "step": 21800 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 5.46875, |
| "learning_rate": 7.84924623115578e-08, |
| "loss": 0.8998, |
| "step": 21900 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 4.875, |
| "learning_rate": 7.839195979899498e-08, |
| "loss": 1.0292, |
| "step": 22000 |
| }, |
| { |
| "epoch": 0.55, |
| "grad_norm": 6.0, |
| "learning_rate": 7.829145728643217e-08, |
| "loss": 0.9347, |
| "step": 22100 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 6.53125, |
| "learning_rate": 7.819095477386935e-08, |
| "loss": 0.9662, |
| "step": 22200 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 8.5625, |
| "learning_rate": 7.809045226130653e-08, |
| "loss": 0.9424, |
| "step": 22300 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 6.6875, |
| "learning_rate": 7.798994974874372e-08, |
| "loss": 0.972, |
| "step": 22400 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 5.59375, |
| "learning_rate": 7.788944723618091e-08, |
| "loss": 0.9037, |
| "step": 22500 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 10.25, |
| "learning_rate": 7.77889447236181e-08, |
| "loss": 0.9267, |
| "step": 22600 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 7.21875, |
| "learning_rate": 7.768844221105528e-08, |
| "loss": 0.941, |
| "step": 22700 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 5.46875, |
| "learning_rate": 7.758793969849247e-08, |
| "loss": 0.9825, |
| "step": 22800 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 7.53125, |
| "learning_rate": 7.748743718592966e-08, |
| "loss": 0.9415, |
| "step": 22900 |
| }, |
| { |
| "epoch": 0.57, |
| "grad_norm": 6.75, |
| "learning_rate": 7.738693467336684e-08, |
| "loss": 0.9295, |
| "step": 23000 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 5.71875, |
| "learning_rate": 7.728643216080402e-08, |
| "loss": 0.9659, |
| "step": 23100 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 6.46875, |
| "learning_rate": 7.71859296482412e-08, |
| "loss": 0.9845, |
| "step": 23200 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 8.625, |
| "learning_rate": 7.708542713567839e-08, |
| "loss": 0.9109, |
| "step": 23300 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 7.34375, |
| "learning_rate": 7.698492462311558e-08, |
| "loss": 0.9265, |
| "step": 23400 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 4.9375, |
| "learning_rate": 7.688442211055277e-08, |
| "loss": 0.8953, |
| "step": 23500 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 4.96875, |
| "learning_rate": 7.678391959798995e-08, |
| "loss": 0.8979, |
| "step": 23600 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 5.21875, |
| "learning_rate": 7.668341708542714e-08, |
| "loss": 0.9755, |
| "step": 23700 |
| }, |
| { |
| "epoch": 0.59, |
| "grad_norm": 5.875, |
| "learning_rate": 7.658291457286433e-08, |
| "loss": 0.9359, |
| "step": 23800 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 6.9375, |
| "learning_rate": 7.648241206030151e-08, |
| "loss": 0.982, |
| "step": 23900 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 8.375, |
| "learning_rate": 7.63819095477387e-08, |
| "loss": 0.9888, |
| "step": 24000 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 2.734375, |
| "learning_rate": 7.628140703517588e-08, |
| "loss": 0.9529, |
| "step": 24100 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 5.96875, |
| "learning_rate": 7.618090452261306e-08, |
| "loss": 0.9906, |
| "step": 24200 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 6.125, |
| "learning_rate": 7.608040201005025e-08, |
| "loss": 0.9653, |
| "step": 24300 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 7.78125, |
| "learning_rate": 7.597989949748744e-08, |
| "loss": 0.9666, |
| "step": 24400 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 6.15625, |
| "learning_rate": 7.587939698492462e-08, |
| "loss": 1.0016, |
| "step": 24500 |
| }, |
| { |
| "epoch": 0.61, |
| "grad_norm": 6.0, |
| "learning_rate": 7.577889447236181e-08, |
| "loss": 0.9509, |
| "step": 24600 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 6.59375, |
| "learning_rate": 7.5678391959799e-08, |
| "loss": 0.9285, |
| "step": 24700 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 8.1875, |
| "learning_rate": 7.557788944723619e-08, |
| "loss": 0.9239, |
| "step": 24800 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 7.3125, |
| "learning_rate": 7.547738693467337e-08, |
| "loss": 0.9311, |
| "step": 24900 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 5.3125, |
| "learning_rate": 7.537688442211056e-08, |
| "loss": 0.905, |
| "step": 25000 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 6.53125, |
| "learning_rate": 7.527638190954774e-08, |
| "loss": 0.9557, |
| "step": 25100 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 7.25, |
| "learning_rate": 7.517587939698492e-08, |
| "loss": 0.9289, |
| "step": 25200 |
| }, |
| { |
| "epoch": 0.63, |
| "grad_norm": 7.21875, |
| "learning_rate": 7.507537688442211e-08, |
| "loss": 0.9892, |
| "step": 25300 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 6.84375, |
| "learning_rate": 7.49748743718593e-08, |
| "loss": 0.9056, |
| "step": 25400 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 5.03125, |
| "learning_rate": 7.487437185929648e-08, |
| "loss": 0.9216, |
| "step": 25500 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 7.1875, |
| "learning_rate": 7.477386934673367e-08, |
| "loss": 1.0327, |
| "step": 25600 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 5.53125, |
| "learning_rate": 7.467336683417086e-08, |
| "loss": 0.9853, |
| "step": 25700 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 6.96875, |
| "learning_rate": 7.457286432160805e-08, |
| "loss": 0.9082, |
| "step": 25800 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 4.59375, |
| "learning_rate": 7.447236180904523e-08, |
| "loss": 0.967, |
| "step": 25900 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 4.3125, |
| "learning_rate": 7.437185929648241e-08, |
| "loss": 0.9624, |
| "step": 26000 |
| }, |
| { |
| "epoch": 0.65, |
| "grad_norm": 5.90625, |
| "learning_rate": 7.42713567839196e-08, |
| "loss": 0.9153, |
| "step": 26100 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 7.15625, |
| "learning_rate": 7.417085427135678e-08, |
| "loss": 0.9392, |
| "step": 26200 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 4.40625, |
| "learning_rate": 7.407035175879397e-08, |
| "loss": 0.9845, |
| "step": 26300 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 5.03125, |
| "learning_rate": 7.396984924623116e-08, |
| "loss": 0.9416, |
| "step": 26400 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 5.78125, |
| "learning_rate": 7.386934673366834e-08, |
| "loss": 1.0113, |
| "step": 26500 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 7.96875, |
| "learning_rate": 7.376884422110553e-08, |
| "loss": 0.9729, |
| "step": 26600 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 7.15625, |
| "learning_rate": 7.366834170854272e-08, |
| "loss": 0.9909, |
| "step": 26700 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 5.25, |
| "learning_rate": 7.35678391959799e-08, |
| "loss": 0.9247, |
| "step": 26800 |
| }, |
| { |
| "epoch": 0.67, |
| "grad_norm": 5.1875, |
| "learning_rate": 7.346733668341709e-08, |
| "loss": 0.977, |
| "step": 26900 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 7.125, |
| "learning_rate": 7.336683417085427e-08, |
| "loss": 0.9593, |
| "step": 27000 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 6.625, |
| "learning_rate": 7.326633165829145e-08, |
| "loss": 0.9694, |
| "step": 27100 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 4.375, |
| "learning_rate": 7.316582914572864e-08, |
| "loss": 0.9794, |
| "step": 27200 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 6.25, |
| "learning_rate": 7.306532663316583e-08, |
| "loss": 0.9583, |
| "step": 27300 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 4.3125, |
| "learning_rate": 7.296482412060301e-08, |
| "loss": 0.9174, |
| "step": 27400 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 8.125, |
| "learning_rate": 7.28643216080402e-08, |
| "loss": 0.9515, |
| "step": 27500 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 6.21875, |
| "learning_rate": 7.276381909547739e-08, |
| "loss": 0.9606, |
| "step": 27600 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 6.8125, |
| "learning_rate": 7.266331658291458e-08, |
| "loss": 1.0048, |
| "step": 27700 |
| }, |
| { |
| "epoch": 0.69, |
| "grad_norm": 4.4375, |
| "learning_rate": 7.256281407035176e-08, |
| "loss": 0.9901, |
| "step": 27800 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 4.875, |
| "learning_rate": 7.246231155778894e-08, |
| "loss": 0.9554, |
| "step": 27900 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 6.5625, |
| "learning_rate": 7.236180904522612e-08, |
| "loss": 0.9625, |
| "step": 28000 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 4.84375, |
| "learning_rate": 7.226130653266331e-08, |
| "loss": 0.9843, |
| "step": 28100 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 3.828125, |
| "learning_rate": 7.21608040201005e-08, |
| "loss": 0.9848, |
| "step": 28200 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 4.40625, |
| "learning_rate": 7.206030150753769e-08, |
| "loss": 0.9629, |
| "step": 28300 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 9.625, |
| "learning_rate": 7.195979899497487e-08, |
| "loss": 0.9431, |
| "step": 28400 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 8.6875, |
| "learning_rate": 7.185929648241206e-08, |
| "loss": 1.0016, |
| "step": 28500 |
| }, |
| { |
| "epoch": 0.71, |
| "grad_norm": 5.84375, |
| "learning_rate": 7.175879396984925e-08, |
| "loss": 1.0209, |
| "step": 28600 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 4.1875, |
| "learning_rate": 7.165829145728644e-08, |
| "loss": 1.0161, |
| "step": 28700 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 9.0, |
| "learning_rate": 7.155778894472362e-08, |
| "loss": 0.9406, |
| "step": 28800 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 4.5, |
| "learning_rate": 7.14572864321608e-08, |
| "loss": 0.9077, |
| "step": 28900 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 6.09375, |
| "learning_rate": 7.135678391959798e-08, |
| "loss": 0.9576, |
| "step": 29000 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 7.625, |
| "learning_rate": 7.125628140703517e-08, |
| "loss": 0.9693, |
| "step": 29100 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 5.96875, |
| "learning_rate": 7.115577889447236e-08, |
| "loss": 0.8866, |
| "step": 29200 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 6.78125, |
| "learning_rate": 7.105527638190955e-08, |
| "loss": 1.0084, |
| "step": 29300 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 6.25, |
| "learning_rate": 7.095477386934673e-08, |
| "loss": 0.9381, |
| "step": 29400 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 4.53125, |
| "learning_rate": 7.085427135678392e-08, |
| "loss": 0.9975, |
| "step": 29500 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 6.09375, |
| "learning_rate": 7.075376884422111e-08, |
| "loss": 0.9602, |
| "step": 29600 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 7.53125, |
| "learning_rate": 7.06532663316583e-08, |
| "loss": 0.941, |
| "step": 29700 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 5.75, |
| "learning_rate": 7.055276381909547e-08, |
| "loss": 1.0162, |
| "step": 29800 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 7.65625, |
| "learning_rate": 7.045226130653266e-08, |
| "loss": 0.9812, |
| "step": 29900 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 6.25, |
| "learning_rate": 7.035175879396984e-08, |
| "loss": 0.9399, |
| "step": 30000 |
| }, |
| { |
| "epoch": 0.75, |
| "eval_loss": 1.0255589485168457, |
| "eval_runtime": 60.091, |
| "eval_samples_per_second": 16.641, |
| "eval_steps_per_second": 4.16, |
| "step": 30000 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 5.90625, |
| "learning_rate": 7.025125628140703e-08, |
| "loss": 0.9587, |
| "step": 30100 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 5.71875, |
| "learning_rate": 7.015075376884422e-08, |
| "loss": 0.9292, |
| "step": 30200 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 5.71875, |
| "learning_rate": 7.00502512562814e-08, |
| "loss": 0.9821, |
| "step": 30300 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 5.03125, |
| "learning_rate": 6.994974874371859e-08, |
| "loss": 0.9572, |
| "step": 30400 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 7.96875, |
| "learning_rate": 6.984924623115578e-08, |
| "loss": 0.9498, |
| "step": 30500 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 4.375, |
| "learning_rate": 6.974874371859297e-08, |
| "loss": 0.8748, |
| "step": 30600 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 7.28125, |
| "learning_rate": 6.964824120603015e-08, |
| "loss": 0.9525, |
| "step": 30700 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 7.625, |
| "learning_rate": 6.954773869346733e-08, |
| "loss": 0.9893, |
| "step": 30800 |
| }, |
| { |
| "epoch": 0.77, |
| "grad_norm": 6.8125, |
| "learning_rate": 6.944723618090451e-08, |
| "loss": 0.9862, |
| "step": 30900 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 8.375, |
| "learning_rate": 6.93467336683417e-08, |
| "loss": 1.0285, |
| "step": 31000 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 7.96875, |
| "learning_rate": 6.924623115577889e-08, |
| "loss": 0.9803, |
| "step": 31100 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 6.40625, |
| "learning_rate": 6.914572864321608e-08, |
| "loss": 0.9453, |
| "step": 31200 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 4.1875, |
| "learning_rate": 6.904522613065328e-08, |
| "loss": 0.9543, |
| "step": 31300 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 6.40625, |
| "learning_rate": 6.894472361809046e-08, |
| "loss": 0.941, |
| "step": 31400 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 8.125, |
| "learning_rate": 6.884422110552765e-08, |
| "loss": 0.9058, |
| "step": 31500 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 8.625, |
| "learning_rate": 6.874371859296483e-08, |
| "loss": 0.9443, |
| "step": 31600 |
| }, |
| { |
| "epoch": 0.79, |
| "grad_norm": 7.84375, |
| "learning_rate": 6.864321608040201e-08, |
| "loss": 0.924, |
| "step": 31700 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 5.65625, |
| "learning_rate": 6.85427135678392e-08, |
| "loss": 0.9291, |
| "step": 31800 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 8.9375, |
| "learning_rate": 6.844221105527639e-08, |
| "loss": 0.9687, |
| "step": 31900 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 5.8125, |
| "learning_rate": 6.834170854271357e-08, |
| "loss": 0.9915, |
| "step": 32000 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 20.25, |
| "learning_rate": 6.824120603015076e-08, |
| "loss": 0.9376, |
| "step": 32100 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 11.0625, |
| "learning_rate": 6.814070351758795e-08, |
| "loss": 0.9131, |
| "step": 32200 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 7.53125, |
| "learning_rate": 6.804020100502514e-08, |
| "loss": 0.9248, |
| "step": 32300 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 6.375, |
| "learning_rate": 6.793969849246232e-08, |
| "loss": 0.9706, |
| "step": 32400 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 10.125, |
| "learning_rate": 6.78391959798995e-08, |
| "loss": 0.9587, |
| "step": 32500 |
| }, |
| { |
| "epoch": 0.81, |
| "grad_norm": 7.25, |
| "learning_rate": 6.773869346733668e-08, |
| "loss": 1.0018, |
| "step": 32600 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 6.78125, |
| "learning_rate": 6.763819095477387e-08, |
| "loss": 0.9087, |
| "step": 32700 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 5.6875, |
| "learning_rate": 6.753768844221106e-08, |
| "loss": 1.0081, |
| "step": 32800 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 7.4375, |
| "learning_rate": 6.743718592964825e-08, |
| "loss": 0.9636, |
| "step": 32900 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 5.1875, |
| "learning_rate": 6.733668341708543e-08, |
| "loss": 0.8546, |
| "step": 33000 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 6.0625, |
| "learning_rate": 6.723618090452262e-08, |
| "loss": 0.9708, |
| "step": 33100 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 5.15625, |
| "learning_rate": 6.713567839195981e-08, |
| "loss": 0.967, |
| "step": 33200 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 5.4375, |
| "learning_rate": 6.7035175879397e-08, |
| "loss": 0.9658, |
| "step": 33300 |
| }, |
| { |
| "epoch": 0.83, |
| "grad_norm": 7.65625, |
| "learning_rate": 6.693467336683418e-08, |
| "loss": 0.8818, |
| "step": 33400 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 3.578125, |
| "learning_rate": 6.683417085427136e-08, |
| "loss": 0.9882, |
| "step": 33500 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 5.75, |
| "learning_rate": 6.673366834170854e-08, |
| "loss": 0.9726, |
| "step": 33600 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 7.03125, |
| "learning_rate": 6.663316582914573e-08, |
| "loss": 0.9468, |
| "step": 33700 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 6.78125, |
| "learning_rate": 6.653266331658292e-08, |
| "loss": 0.9443, |
| "step": 33800 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 6.15625, |
| "learning_rate": 6.64321608040201e-08, |
| "loss": 1.0086, |
| "step": 33900 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 6.5625, |
| "learning_rate": 6.633165829145729e-08, |
| "loss": 0.9398, |
| "step": 34000 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 7.5, |
| "learning_rate": 6.623115577889448e-08, |
| "loss": 0.9407, |
| "step": 34100 |
| }, |
| { |
| "epoch": 0.85, |
| "grad_norm": 4.9375, |
| "learning_rate": 6.613065326633167e-08, |
| "loss": 0.9193, |
| "step": 34200 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 6.875, |
| "learning_rate": 6.603015075376885e-08, |
| "loss": 0.918, |
| "step": 34300 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 6.28125, |
| "learning_rate": 6.592964824120603e-08, |
| "loss": 0.9085, |
| "step": 34400 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 11.5625, |
| "learning_rate": 6.582914572864322e-08, |
| "loss": 0.9345, |
| "step": 34500 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 9.1875, |
| "learning_rate": 6.57286432160804e-08, |
| "loss": 1.0002, |
| "step": 34600 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 6.3125, |
| "learning_rate": 6.562814070351759e-08, |
| "loss": 0.9694, |
| "step": 34700 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 9.75, |
| "learning_rate": 6.552763819095478e-08, |
| "loss": 0.9525, |
| "step": 34800 |
| }, |
| { |
| "epoch": 0.87, |
| "grad_norm": 6.0, |
| "learning_rate": 6.542713567839196e-08, |
| "loss": 0.9514, |
| "step": 34900 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 7.8125, |
| "learning_rate": 6.532663316582915e-08, |
| "loss": 0.9394, |
| "step": 35000 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 7.78125, |
| "learning_rate": 6.522613065326634e-08, |
| "loss": 0.9558, |
| "step": 35100 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 6.9375, |
| "learning_rate": 6.512562814070353e-08, |
| "loss": 0.9644, |
| "step": 35200 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 5.46875, |
| "learning_rate": 6.502512562814071e-08, |
| "loss": 0.9093, |
| "step": 35300 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 8.6875, |
| "learning_rate": 6.492462311557789e-08, |
| "loss": 0.9008, |
| "step": 35400 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 10.0625, |
| "learning_rate": 6.482412060301507e-08, |
| "loss": 0.9712, |
| "step": 35500 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 8.1875, |
| "learning_rate": 6.472361809045226e-08, |
| "loss": 0.9111, |
| "step": 35600 |
| }, |
| { |
| "epoch": 0.89, |
| "grad_norm": 8.3125, |
| "learning_rate": 6.462311557788945e-08, |
| "loss": 0.9768, |
| "step": 35700 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 4.03125, |
| "learning_rate": 6.452261306532664e-08, |
| "loss": 0.9247, |
| "step": 35800 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 5.09375, |
| "learning_rate": 6.442211055276382e-08, |
| "loss": 0.9598, |
| "step": 35900 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 4.75, |
| "learning_rate": 6.432160804020101e-08, |
| "loss": 0.9472, |
| "step": 36000 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 7.5, |
| "learning_rate": 6.42211055276382e-08, |
| "loss": 0.976, |
| "step": 36100 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 6.46875, |
| "learning_rate": 6.412060301507538e-08, |
| "loss": 0.9229, |
| "step": 36200 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 6.4375, |
| "learning_rate": 6.402010050251257e-08, |
| "loss": 0.916, |
| "step": 36300 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 7.90625, |
| "learning_rate": 6.391959798994975e-08, |
| "loss": 0.9456, |
| "step": 36400 |
| }, |
| { |
| "epoch": 0.91, |
| "grad_norm": 9.0, |
| "learning_rate": 6.381909547738693e-08, |
| "loss": 0.952, |
| "step": 36500 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 7.09375, |
| "learning_rate": 6.371859296482412e-08, |
| "loss": 0.9805, |
| "step": 36600 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 7.46875, |
| "learning_rate": 6.361809045226131e-08, |
| "loss": 0.9899, |
| "step": 36700 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 5.125, |
| "learning_rate": 6.35175879396985e-08, |
| "loss": 0.9339, |
| "step": 36800 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 9.5625, |
| "learning_rate": 6.341708542713568e-08, |
| "loss": 0.9171, |
| "step": 36900 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 8.6875, |
| "learning_rate": 6.331658291457287e-08, |
| "loss": 0.9853, |
| "step": 37000 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 5.5, |
| "learning_rate": 6.321608040201006e-08, |
| "loss": 0.9881, |
| "step": 37100 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 6.78125, |
| "learning_rate": 6.311557788944724e-08, |
| "loss": 0.9987, |
| "step": 37200 |
| }, |
| { |
| "epoch": 0.93, |
| "grad_norm": 4.59375, |
| "learning_rate": 6.301507537688442e-08, |
| "loss": 1.0295, |
| "step": 37300 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 5.9375, |
| "learning_rate": 6.29145728643216e-08, |
| "loss": 0.9171, |
| "step": 37400 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 6.3125, |
| "learning_rate": 6.281407035175879e-08, |
| "loss": 0.9337, |
| "step": 37500 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 5.21875, |
| "learning_rate": 6.271356783919598e-08, |
| "loss": 0.9479, |
| "step": 37600 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 5.59375, |
| "learning_rate": 6.261306532663317e-08, |
| "loss": 0.9544, |
| "step": 37700 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 6.46875, |
| "learning_rate": 6.251256281407035e-08, |
| "loss": 0.9719, |
| "step": 37800 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 5.15625, |
| "learning_rate": 6.241206030150754e-08, |
| "loss": 0.9901, |
| "step": 37900 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 6.28125, |
| "learning_rate": 6.231155778894473e-08, |
| "loss": 0.9377, |
| "step": 38000 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 6.875, |
| "learning_rate": 6.221105527638192e-08, |
| "loss": 0.9521, |
| "step": 38100 |
| }, |
| { |
| "epoch": 0.95, |
| "grad_norm": 5.9375, |
| "learning_rate": 6.21105527638191e-08, |
| "loss": 0.9793, |
| "step": 38200 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 12.875, |
| "learning_rate": 6.201005025125628e-08, |
| "loss": 1.0009, |
| "step": 38300 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 5.75, |
| "learning_rate": 6.190954773869346e-08, |
| "loss": 0.9451, |
| "step": 38400 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 6.46875, |
| "learning_rate": 6.180904522613065e-08, |
| "loss": 0.9397, |
| "step": 38500 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 4.84375, |
| "learning_rate": 6.170854271356784e-08, |
| "loss": 0.9011, |
| "step": 38600 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 5.59375, |
| "learning_rate": 6.160804020100503e-08, |
| "loss": 0.963, |
| "step": 38700 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 9.375, |
| "learning_rate": 6.150753768844221e-08, |
| "loss": 0.9328, |
| "step": 38800 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 6.90625, |
| "learning_rate": 6.14070351758794e-08, |
| "loss": 0.9225, |
| "step": 38900 |
| }, |
| { |
| "epoch": 0.97, |
| "grad_norm": 6.71875, |
| "learning_rate": 6.130653266331659e-08, |
| "loss": 0.9526, |
| "step": 39000 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 4.53125, |
| "learning_rate": 6.120603015075377e-08, |
| "loss": 0.9599, |
| "step": 39100 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 6.1875, |
| "learning_rate": 6.110552763819095e-08, |
| "loss": 0.939, |
| "step": 39200 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 7.5625, |
| "learning_rate": 6.100502512562814e-08, |
| "loss": 0.8991, |
| "step": 39300 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 7.6875, |
| "learning_rate": 6.090452261306532e-08, |
| "loss": 0.9519, |
| "step": 39400 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 6.46875, |
| "learning_rate": 6.080402010050251e-08, |
| "loss": 0.9196, |
| "step": 39500 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 5.59375, |
| "learning_rate": 6.07035175879397e-08, |
| "loss": 0.9775, |
| "step": 39600 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 9.125, |
| "learning_rate": 6.060301507537688e-08, |
| "loss": 0.9111, |
| "step": 39700 |
| }, |
| { |
| "epoch": 0.99, |
| "grad_norm": 8.5, |
| "learning_rate": 6.050251256281407e-08, |
| "loss": 0.9614, |
| "step": 39800 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 3.90625, |
| "learning_rate": 6.040201005025126e-08, |
| "loss": 0.9762, |
| "step": 39900 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 6.125, |
| "learning_rate": 6.030150753768845e-08, |
| "loss": 0.8843, |
| "step": 40000 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 1.0208626985549927, |
| "eval_runtime": 60.0329, |
| "eval_samples_per_second": 16.658, |
| "eval_steps_per_second": 4.164, |
| "step": 40000 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 8.5625, |
| "learning_rate": 6.020100502512563e-08, |
| "loss": 1.0214, |
| "step": 40100 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 13.0625, |
| "learning_rate": 6.010050251256281e-08, |
| "loss": 0.9954, |
| "step": 40200 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 6.90625, |
| "learning_rate": 6e-08, |
| "loss": 0.9723, |
| "step": 40300 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 4.125, |
| "learning_rate": 5.989949748743718e-08, |
| "loss": 0.9679, |
| "step": 40400 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 6.3125, |
| "learning_rate": 5.979899497487437e-08, |
| "loss": 0.9618, |
| "step": 40500 |
| }, |
| { |
| "epoch": 1.01, |
| "grad_norm": 7.5625, |
| "learning_rate": 5.969849246231156e-08, |
| "loss": 0.9921, |
| "step": 40600 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 5.09375, |
| "learning_rate": 5.9597989949748744e-08, |
| "loss": 0.9605, |
| "step": 40700 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 5.53125, |
| "learning_rate": 5.949748743718593e-08, |
| "loss": 0.9766, |
| "step": 40800 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 11.5625, |
| "learning_rate": 5.939698492462311e-08, |
| "loss": 0.963, |
| "step": 40900 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 6.8125, |
| "learning_rate": 5.92964824120603e-08, |
| "loss": 0.9791, |
| "step": 41000 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 4.84375, |
| "learning_rate": 5.9195979899497486e-08, |
| "loss": 0.9815, |
| "step": 41100 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 4.59375, |
| "learning_rate": 5.909547738693467e-08, |
| "loss": 0.9632, |
| "step": 41200 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 5.84375, |
| "learning_rate": 5.899497487437186e-08, |
| "loss": 0.9738, |
| "step": 41300 |
| }, |
| { |
| "epoch": 1.03, |
| "grad_norm": 13.125, |
| "learning_rate": 5.889447236180904e-08, |
| "loss": 0.956, |
| "step": 41400 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 6.9375, |
| "learning_rate": 5.879396984924623e-08, |
| "loss": 0.9617, |
| "step": 41500 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 6.34375, |
| "learning_rate": 5.8693467336683415e-08, |
| "loss": 0.9972, |
| "step": 41600 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 9.1875, |
| "learning_rate": 5.85929648241206e-08, |
| "loss": 0.9284, |
| "step": 41700 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 6.0, |
| "learning_rate": 5.849246231155778e-08, |
| "loss": 0.982, |
| "step": 41800 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 5.65625, |
| "learning_rate": 5.839195979899497e-08, |
| "loss": 0.9013, |
| "step": 41900 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 6.21875, |
| "learning_rate": 5.829145728643216e-08, |
| "loss": 0.994, |
| "step": 42000 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 4.6875, |
| "learning_rate": 5.8190954773869345e-08, |
| "loss": 0.9819, |
| "step": 42100 |
| }, |
| { |
| "epoch": 1.05, |
| "grad_norm": 7.09375, |
| "learning_rate": 5.809045226130653e-08, |
| "loss": 0.961, |
| "step": 42200 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 11.625, |
| "learning_rate": 5.798994974874371e-08, |
| "loss": 0.9869, |
| "step": 42300 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 7.625, |
| "learning_rate": 5.78894472361809e-08, |
| "loss": 1.0082, |
| "step": 42400 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 9.5, |
| "learning_rate": 5.778894472361809e-08, |
| "loss": 0.9392, |
| "step": 42500 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 6.34375, |
| "learning_rate": 5.7688442211055274e-08, |
| "loss": 0.9234, |
| "step": 42600 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 5.4375, |
| "learning_rate": 5.758793969849246e-08, |
| "loss": 0.978, |
| "step": 42700 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 8.25, |
| "learning_rate": 5.748743718592964e-08, |
| "loss": 0.9698, |
| "step": 42800 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 8.1875, |
| "learning_rate": 5.738693467336683e-08, |
| "loss": 0.9794, |
| "step": 42900 |
| }, |
| { |
| "epoch": 1.07, |
| "grad_norm": 7.625, |
| "learning_rate": 5.7286432160804017e-08, |
| "loss": 0.9655, |
| "step": 43000 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 8.375, |
| "learning_rate": 5.7185929648241204e-08, |
| "loss": 0.9743, |
| "step": 43100 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 4.375, |
| "learning_rate": 5.708542713567839e-08, |
| "loss": 0.9515, |
| "step": 43200 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 6.96875, |
| "learning_rate": 5.698492462311557e-08, |
| "loss": 1.0257, |
| "step": 43300 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 8.5625, |
| "learning_rate": 5.688442211055276e-08, |
| "loss": 0.9759, |
| "step": 43400 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 5.90625, |
| "learning_rate": 5.6783919597989946e-08, |
| "loss": 0.9961, |
| "step": 43500 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 7.5, |
| "learning_rate": 5.6683417085427133e-08, |
| "loss": 0.9859, |
| "step": 43600 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 6.90625, |
| "learning_rate": 5.658291457286432e-08, |
| "loss": 0.986, |
| "step": 43700 |
| }, |
| { |
| "epoch": 1.09, |
| "grad_norm": 11.75, |
| "learning_rate": 5.6482412060301514e-08, |
| "loss": 0.9469, |
| "step": 43800 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 6.5, |
| "learning_rate": 5.63819095477387e-08, |
| "loss": 0.9295, |
| "step": 43900 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 6.71875, |
| "learning_rate": 5.628140703517589e-08, |
| "loss": 1.0052, |
| "step": 44000 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 5.40625, |
| "learning_rate": 5.618090452261307e-08, |
| "loss": 1.018, |
| "step": 44100 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 8.625, |
| "learning_rate": 5.6080402010050257e-08, |
| "loss": 0.9492, |
| "step": 44200 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 8.3125, |
| "learning_rate": 5.5979899497487444e-08, |
| "loss": 0.9821, |
| "step": 44300 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 6.34375, |
| "learning_rate": 5.587939698492463e-08, |
| "loss": 0.9606, |
| "step": 44400 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 6.15625, |
| "learning_rate": 5.577889447236182e-08, |
| "loss": 0.9933, |
| "step": 44500 |
| }, |
| { |
| "epoch": 1.11, |
| "grad_norm": 8.0625, |
| "learning_rate": 5.5678391959799e-08, |
| "loss": 0.9813, |
| "step": 44600 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 7.09375, |
| "learning_rate": 5.5577889447236186e-08, |
| "loss": 1.0051, |
| "step": 44700 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 9.5625, |
| "learning_rate": 5.5477386934673373e-08, |
| "loss": 0.9277, |
| "step": 44800 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 10.6875, |
| "learning_rate": 5.537688442211056e-08, |
| "loss": 0.9432, |
| "step": 44900 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 4.84375, |
| "learning_rate": 5.527638190954774e-08, |
| "loss": 0.9998, |
| "step": 45000 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 7.15625, |
| "learning_rate": 5.517587939698493e-08, |
| "loss": 1.0521, |
| "step": 45100 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 6.0625, |
| "learning_rate": 5.5075376884422116e-08, |
| "loss": 1.0189, |
| "step": 45200 |
| }, |
| { |
| "epoch": 1.13, |
| "grad_norm": 6.5625, |
| "learning_rate": 5.49748743718593e-08, |
| "loss": 0.97, |
| "step": 45300 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 6.03125, |
| "learning_rate": 5.487437185929649e-08, |
| "loss": 0.9906, |
| "step": 45400 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 5.1875, |
| "learning_rate": 5.477386934673367e-08, |
| "loss": 0.9627, |
| "step": 45500 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 5.0, |
| "learning_rate": 5.467336683417086e-08, |
| "loss": 0.9621, |
| "step": 45600 |
| }, |
| { |
| "epoch": 1.14, |
| "grad_norm": 9.0, |
| "learning_rate": 5.4572864321608045e-08, |
| "loss": 0.9875, |
| "step": 45700 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 8.25, |
| "learning_rate": 5.447236180904523e-08, |
| "loss": 0.9537, |
| "step": 45800 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 5.125, |
| "learning_rate": 5.437185929648242e-08, |
| "loss": 0.938, |
| "step": 45900 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 10.25, |
| "learning_rate": 5.42713567839196e-08, |
| "loss": 0.9981, |
| "step": 46000 |
| }, |
| { |
| "epoch": 1.15, |
| "grad_norm": 7.5, |
| "learning_rate": 5.417085427135679e-08, |
| "loss": 0.9967, |
| "step": 46100 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 6.75, |
| "learning_rate": 5.4070351758793975e-08, |
| "loss": 0.95, |
| "step": 46200 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 5.5625, |
| "learning_rate": 5.396984924623116e-08, |
| "loss": 0.9748, |
| "step": 46300 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 8.875, |
| "learning_rate": 5.386934673366835e-08, |
| "loss": 0.9643, |
| "step": 46400 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 10.6875, |
| "learning_rate": 5.376884422110553e-08, |
| "loss": 0.9516, |
| "step": 46500 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 5.9375, |
| "learning_rate": 5.366834170854272e-08, |
| "loss": 0.9789, |
| "step": 46600 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 5.59375, |
| "learning_rate": 5.3567839195979904e-08, |
| "loss": 0.9469, |
| "step": 46700 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 9.5, |
| "learning_rate": 5.346733668341709e-08, |
| "loss": 0.9931, |
| "step": 46800 |
| }, |
| { |
| "epoch": 1.17, |
| "grad_norm": 8.0, |
| "learning_rate": 5.336683417085427e-08, |
| "loss": 0.9775, |
| "step": 46900 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 6.6875, |
| "learning_rate": 5.326633165829146e-08, |
| "loss": 0.9502, |
| "step": 47000 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 6.96875, |
| "learning_rate": 5.3165829145728646e-08, |
| "loss": 0.9498, |
| "step": 47100 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 4.96875, |
| "learning_rate": 5.3065326633165834e-08, |
| "loss": 0.9383, |
| "step": 47200 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 4.875, |
| "learning_rate": 5.296482412060302e-08, |
| "loss": 0.9807, |
| "step": 47300 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 7.84375, |
| "learning_rate": 5.28643216080402e-08, |
| "loss": 0.9163, |
| "step": 47400 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 5.53125, |
| "learning_rate": 5.276381909547739e-08, |
| "loss": 0.9943, |
| "step": 47500 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 7.03125, |
| "learning_rate": 5.2663316582914576e-08, |
| "loss": 0.9654, |
| "step": 47600 |
| }, |
| { |
| "epoch": 1.19, |
| "grad_norm": 7.40625, |
| "learning_rate": 5.256281407035176e-08, |
| "loss": 0.9992, |
| "step": 47700 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 4.1875, |
| "learning_rate": 5.246231155778895e-08, |
| "loss": 0.9386, |
| "step": 47800 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 8.3125, |
| "learning_rate": 5.236180904522613e-08, |
| "loss": 0.9572, |
| "step": 47900 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 5.71875, |
| "learning_rate": 5.226130653266332e-08, |
| "loss": 0.9433, |
| "step": 48000 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 5.46875, |
| "learning_rate": 5.2160804020100505e-08, |
| "loss": 0.9092, |
| "step": 48100 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 7.46875, |
| "learning_rate": 5.206030150753769e-08, |
| "loss": 0.9815, |
| "step": 48200 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 8.0625, |
| "learning_rate": 5.195979899497488e-08, |
| "loss": 0.9606, |
| "step": 48300 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 6.40625, |
| "learning_rate": 5.185929648241206e-08, |
| "loss": 1.0209, |
| "step": 48400 |
| }, |
| { |
| "epoch": 1.21, |
| "grad_norm": 5.90625, |
| "learning_rate": 5.175879396984925e-08, |
| "loss": 0.9392, |
| "step": 48500 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 12.1875, |
| "learning_rate": 5.1658291457286435e-08, |
| "loss": 0.9491, |
| "step": 48600 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 6.6875, |
| "learning_rate": 5.155778894472362e-08, |
| "loss": 0.9291, |
| "step": 48700 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 7.0, |
| "learning_rate": 5.145728643216081e-08, |
| "loss": 0.97, |
| "step": 48800 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 5.03125, |
| "learning_rate": 5.135678391959799e-08, |
| "loss": 0.9782, |
| "step": 48900 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 6.5, |
| "learning_rate": 5.125628140703518e-08, |
| "loss": 0.9603, |
| "step": 49000 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 7.15625, |
| "learning_rate": 5.1155778894472365e-08, |
| "loss": 0.9882, |
| "step": 49100 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 10.5, |
| "learning_rate": 5.105527638190955e-08, |
| "loss": 0.9778, |
| "step": 49200 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 6.40625, |
| "learning_rate": 5.095477386934673e-08, |
| "loss": 0.9749, |
| "step": 49300 |
| }, |
| { |
| "epoch": 1.23, |
| "grad_norm": 5.0625, |
| "learning_rate": 5.085427135678392e-08, |
| "loss": 0.9316, |
| "step": 49400 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 5.6875, |
| "learning_rate": 5.075376884422111e-08, |
| "loss": 0.9238, |
| "step": 49500 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 6.1875, |
| "learning_rate": 5.0653266331658294e-08, |
| "loss": 0.9979, |
| "step": 49600 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 7.96875, |
| "learning_rate": 5.055276381909548e-08, |
| "loss": 0.936, |
| "step": 49700 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 6.65625, |
| "learning_rate": 5.045226130653266e-08, |
| "loss": 0.9642, |
| "step": 49800 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 7.9375, |
| "learning_rate": 5.035175879396985e-08, |
| "loss": 0.9363, |
| "step": 49900 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 4.28125, |
| "learning_rate": 5.0251256281407036e-08, |
| "loss": 0.9519, |
| "step": 50000 |
| }, |
| { |
| "epoch": 1.25, |
| "eval_loss": 1.057003140449524, |
| "eval_runtime": 60.0619, |
| "eval_samples_per_second": 16.649, |
| "eval_steps_per_second": 4.162, |
| "step": 50000 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 7.8125, |
| "learning_rate": 5.0150753768844224e-08, |
| "loss": 0.962, |
| "step": 50100 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 5.34375, |
| "learning_rate": 5.005025125628141e-08, |
| "loss": 0.9802, |
| "step": 50200 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 6.40625, |
| "learning_rate": 4.994974874371859e-08, |
| "loss": 0.9832, |
| "step": 50300 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 7.9375, |
| "learning_rate": 4.984924623115578e-08, |
| "loss": 0.9983, |
| "step": 50400 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 7.28125, |
| "learning_rate": 4.9748743718592966e-08, |
| "loss": 0.98, |
| "step": 50500 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 4.34375, |
| "learning_rate": 4.964824120603015e-08, |
| "loss": 1.0038, |
| "step": 50600 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 5.625, |
| "learning_rate": 4.954773869346734e-08, |
| "loss": 0.9949, |
| "step": 50700 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 3.671875, |
| "learning_rate": 4.944723618090452e-08, |
| "loss": 0.9518, |
| "step": 50800 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 5.9375, |
| "learning_rate": 4.934673366834171e-08, |
| "loss": 0.9327, |
| "step": 50900 |
| }, |
| { |
| "epoch": 1.27, |
| "grad_norm": 7.71875, |
| "learning_rate": 4.9246231155778895e-08, |
| "loss": 0.9931, |
| "step": 51000 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 11.0, |
| "learning_rate": 4.914572864321608e-08, |
| "loss": 0.981, |
| "step": 51100 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 7.25, |
| "learning_rate": 4.904522613065326e-08, |
| "loss": 0.9865, |
| "step": 51200 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 7.0, |
| "learning_rate": 4.894472361809045e-08, |
| "loss": 0.9617, |
| "step": 51300 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 5.96875, |
| "learning_rate": 4.884422110552764e-08, |
| "loss": 0.9184, |
| "step": 51400 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 4.125, |
| "learning_rate": 4.8743718592964825e-08, |
| "loss": 0.9074, |
| "step": 51500 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 7.03125, |
| "learning_rate": 4.864321608040201e-08, |
| "loss": 0.9717, |
| "step": 51600 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 8.5625, |
| "learning_rate": 4.854271356783919e-08, |
| "loss": 0.964, |
| "step": 51700 |
| }, |
| { |
| "epoch": 1.29, |
| "grad_norm": 3.671875, |
| "learning_rate": 4.844221105527638e-08, |
| "loss": 0.9795, |
| "step": 51800 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 7.375, |
| "learning_rate": 4.834170854271357e-08, |
| "loss": 0.9763, |
| "step": 51900 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 5.9375, |
| "learning_rate": 4.8241206030150754e-08, |
| "loss": 0.931, |
| "step": 52000 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 8.375, |
| "learning_rate": 4.814070351758794e-08, |
| "loss": 0.9088, |
| "step": 52100 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 6.5, |
| "learning_rate": 4.804020100502512e-08, |
| "loss": 0.9834, |
| "step": 52200 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 7.96875, |
| "learning_rate": 4.793969849246231e-08, |
| "loss": 0.9953, |
| "step": 52300 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 6.25, |
| "learning_rate": 4.7839195979899497e-08, |
| "loss": 1.0036, |
| "step": 52400 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 8.625, |
| "learning_rate": 4.7738693467336684e-08, |
| "loss": 0.8907, |
| "step": 52500 |
| }, |
| { |
| "epoch": 1.31, |
| "grad_norm": 10.75, |
| "learning_rate": 4.763819095477387e-08, |
| "loss": 0.9583, |
| "step": 52600 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 8.1875, |
| "learning_rate": 4.753768844221105e-08, |
| "loss": 0.977, |
| "step": 52700 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 7.96875, |
| "learning_rate": 4.743718592964824e-08, |
| "loss": 0.9456, |
| "step": 52800 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 8.9375, |
| "learning_rate": 4.7336683417085426e-08, |
| "loss": 0.9952, |
| "step": 52900 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 5.8125, |
| "learning_rate": 4.7236180904522613e-08, |
| "loss": 0.9916, |
| "step": 53000 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 8.9375, |
| "learning_rate": 4.71356783919598e-08, |
| "loss": 1.0113, |
| "step": 53100 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 7.15625, |
| "learning_rate": 4.703517587939699e-08, |
| "loss": 0.8996, |
| "step": 53200 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 6.0, |
| "learning_rate": 4.6934673366834175e-08, |
| "loss": 0.9295, |
| "step": 53300 |
| }, |
| { |
| "epoch": 1.33, |
| "grad_norm": 8.625, |
| "learning_rate": 4.683417085427136e-08, |
| "loss": 0.9844, |
| "step": 53400 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 4.9375, |
| "learning_rate": 4.673366834170855e-08, |
| "loss": 1.0053, |
| "step": 53500 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 6.8125, |
| "learning_rate": 4.663316582914573e-08, |
| "loss": 0.9091, |
| "step": 53600 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 9.5, |
| "learning_rate": 4.653266331658292e-08, |
| "loss": 0.9683, |
| "step": 53700 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 6.0625, |
| "learning_rate": 4.6432160804020105e-08, |
| "loss": 0.9715, |
| "step": 53800 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 8.0625, |
| "learning_rate": 4.633165829145729e-08, |
| "loss": 0.9531, |
| "step": 53900 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 6.0, |
| "learning_rate": 4.623115577889447e-08, |
| "loss": 0.956, |
| "step": 54000 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 7.03125, |
| "learning_rate": 4.613065326633166e-08, |
| "loss": 0.9348, |
| "step": 54100 |
| }, |
| { |
| "epoch": 1.35, |
| "grad_norm": 9.1875, |
| "learning_rate": 4.603015075376885e-08, |
| "loss": 0.9732, |
| "step": 54200 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 8.375, |
| "learning_rate": 4.5929648241206034e-08, |
| "loss": 0.9213, |
| "step": 54300 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 4.125, |
| "learning_rate": 4.582914572864322e-08, |
| "loss": 0.9606, |
| "step": 54400 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 6.6875, |
| "learning_rate": 4.57286432160804e-08, |
| "loss": 0.9223, |
| "step": 54500 |
| }, |
| { |
| "epoch": 1.36, |
| "grad_norm": 5.21875, |
| "learning_rate": 4.562814070351759e-08, |
| "loss": 0.9549, |
| "step": 54600 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 5.9375, |
| "learning_rate": 4.5527638190954776e-08, |
| "loss": 0.9934, |
| "step": 54700 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 5.6875, |
| "learning_rate": 4.5427135678391964e-08, |
| "loss": 0.962, |
| "step": 54800 |
| }, |
| { |
| "epoch": 1.37, |
| "grad_norm": 8.5, |
| "learning_rate": 4.532663316582915e-08, |
| "loss": 1.0155, |
| "step": 54900 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 7.0625, |
| "learning_rate": 4.522613065326633e-08, |
| "loss": 0.964, |
| "step": 55000 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 5.40625, |
| "learning_rate": 4.512562814070352e-08, |
| "loss": 0.956, |
| "step": 55100 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 5.84375, |
| "learning_rate": 4.5025125628140706e-08, |
| "loss": 1.0025, |
| "step": 55200 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 6.0625, |
| "learning_rate": 4.492462311557789e-08, |
| "loss": 0.9457, |
| "step": 55300 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 5.40625, |
| "learning_rate": 4.482412060301508e-08, |
| "loss": 0.9152, |
| "step": 55400 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 3.8125, |
| "learning_rate": 4.472361809045226e-08, |
| "loss": 1.0083, |
| "step": 55500 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 5.53125, |
| "learning_rate": 4.462311557788945e-08, |
| "loss": 0.937, |
| "step": 55600 |
| }, |
| { |
| "epoch": 1.39, |
| "grad_norm": 6.65625, |
| "learning_rate": 4.4522613065326635e-08, |
| "loss": 0.9472, |
| "step": 55700 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 5.0, |
| "learning_rate": 4.442211055276382e-08, |
| "loss": 1.013, |
| "step": 55800 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 7.46875, |
| "learning_rate": 4.4321608040201e-08, |
| "loss": 0.9848, |
| "step": 55900 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 3.4375, |
| "learning_rate": 4.422110552763819e-08, |
| "loss": 0.9673, |
| "step": 56000 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 5.8125, |
| "learning_rate": 4.412060301507538e-08, |
| "loss": 0.9647, |
| "step": 56100 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 7.34375, |
| "learning_rate": 4.4020100502512565e-08, |
| "loss": 0.9551, |
| "step": 56200 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 6.625, |
| "learning_rate": 4.391959798994975e-08, |
| "loss": 1.0394, |
| "step": 56300 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 9.25, |
| "learning_rate": 4.381909547738693e-08, |
| "loss": 0.9819, |
| "step": 56400 |
| }, |
| { |
| "epoch": 1.41, |
| "grad_norm": 6.4375, |
| "learning_rate": 4.371859296482412e-08, |
| "loss": 1.0049, |
| "step": 56500 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 7.8125, |
| "learning_rate": 4.361809045226131e-08, |
| "loss": 0.9815, |
| "step": 56600 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 6.375, |
| "learning_rate": 4.3517587939698494e-08, |
| "loss": 0.9417, |
| "step": 56700 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 5.9375, |
| "learning_rate": 4.341708542713568e-08, |
| "loss": 0.9617, |
| "step": 56800 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 6.96875, |
| "learning_rate": 4.331658291457286e-08, |
| "loss": 0.9694, |
| "step": 56900 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 4.40625, |
| "learning_rate": 4.321608040201005e-08, |
| "loss": 0.9392, |
| "step": 57000 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 2.28125, |
| "learning_rate": 4.3115577889447237e-08, |
| "loss": 0.939, |
| "step": 57100 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 6.78125, |
| "learning_rate": 4.3015075376884424e-08, |
| "loss": 0.9567, |
| "step": 57200 |
| }, |
| { |
| "epoch": 1.43, |
| "grad_norm": 6.90625, |
| "learning_rate": 4.291457286432161e-08, |
| "loss": 0.9173, |
| "step": 57300 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 7.6875, |
| "learning_rate": 4.281407035175879e-08, |
| "loss": 0.9893, |
| "step": 57400 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 10.0, |
| "learning_rate": 4.271356783919598e-08, |
| "loss": 0.9424, |
| "step": 57500 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 5.21875, |
| "learning_rate": 4.2613065326633166e-08, |
| "loss": 0.9104, |
| "step": 57600 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 6.59375, |
| "learning_rate": 4.2512562814070353e-08, |
| "loss": 1.004, |
| "step": 57700 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 5.21875, |
| "learning_rate": 4.241206030150754e-08, |
| "loss": 0.975, |
| "step": 57800 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 7.71875, |
| "learning_rate": 4.231155778894472e-08, |
| "loss": 0.9487, |
| "step": 57900 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 5.625, |
| "learning_rate": 4.221105527638191e-08, |
| "loss": 0.973, |
| "step": 58000 |
| }, |
| { |
| "epoch": 1.45, |
| "grad_norm": 7.6875, |
| "learning_rate": 4.2110552763819096e-08, |
| "loss": 0.9233, |
| "step": 58100 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 6.0, |
| "learning_rate": 4.201005025125628e-08, |
| "loss": 1.0018, |
| "step": 58200 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 7.34375, |
| "learning_rate": 4.1909547738693463e-08, |
| "loss": 0.9226, |
| "step": 58300 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 4.65625, |
| "learning_rate": 4.180904522613065e-08, |
| "loss": 0.9206, |
| "step": 58400 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 13.6875, |
| "learning_rate": 4.170854271356784e-08, |
| "loss": 1.0329, |
| "step": 58500 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 6.6875, |
| "learning_rate": 4.1608040201005025e-08, |
| "loss": 0.9998, |
| "step": 58600 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 6.71875, |
| "learning_rate": 4.150753768844221e-08, |
| "loss": 0.9444, |
| "step": 58700 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 6.90625, |
| "learning_rate": 4.140703517587939e-08, |
| "loss": 0.9434, |
| "step": 58800 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 5.5625, |
| "learning_rate": 4.130653266331658e-08, |
| "loss": 0.8801, |
| "step": 58900 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 5.21875, |
| "learning_rate": 4.120603015075377e-08, |
| "loss": 0.9564, |
| "step": 59000 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 5.375, |
| "learning_rate": 4.1105527638190955e-08, |
| "loss": 0.9745, |
| "step": 59100 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 6.875, |
| "learning_rate": 4.100502512562814e-08, |
| "loss": 0.9509, |
| "step": 59200 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 5.96875, |
| "learning_rate": 4.090452261306532e-08, |
| "loss": 0.8941, |
| "step": 59300 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 6.75, |
| "learning_rate": 4.0804020100502516e-08, |
| "loss": 0.9393, |
| "step": 59400 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 5.375, |
| "learning_rate": 4.0703517587939704e-08, |
| "loss": 0.924, |
| "step": 59500 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 4.84375, |
| "learning_rate": 4.060301507537689e-08, |
| "loss": 0.9013, |
| "step": 59600 |
| }, |
| { |
| "epoch": 1.49, |
| "grad_norm": 6.125, |
| "learning_rate": 4.050251256281407e-08, |
| "loss": 0.951, |
| "step": 59700 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 6.5, |
| "learning_rate": 4.040201005025126e-08, |
| "loss": 0.9097, |
| "step": 59800 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 4.0625, |
| "learning_rate": 4.0301507537688446e-08, |
| "loss": 0.9846, |
| "step": 59900 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 4.3125, |
| "learning_rate": 4.020100502512563e-08, |
| "loss": 1.001, |
| "step": 60000 |
| }, |
| { |
| "epoch": 1.5, |
| "eval_loss": 1.0297139883041382, |
| "eval_runtime": 60.0085, |
| "eval_samples_per_second": 16.664, |
| "eval_steps_per_second": 4.166, |
| "step": 60000 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 5.65625, |
| "learning_rate": 4.010050251256282e-08, |
| "loss": 0.9566, |
| "step": 60100 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 5.75, |
| "learning_rate": 4e-08, |
| "loss": 0.9717, |
| "step": 60200 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 3.28125, |
| "learning_rate": 3.989949748743719e-08, |
| "loss": 0.9496, |
| "step": 60300 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 8.625, |
| "learning_rate": 3.9798994974874375e-08, |
| "loss": 0.9947, |
| "step": 60400 |
| }, |
| { |
| "epoch": 1.51, |
| "grad_norm": 6.71875, |
| "learning_rate": 3.969849246231156e-08, |
| "loss": 0.9912, |
| "step": 60500 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 9.875, |
| "learning_rate": 3.959798994974874e-08, |
| "loss": 0.9177, |
| "step": 60600 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 8.5, |
| "learning_rate": 3.949748743718593e-08, |
| "loss": 1.026, |
| "step": 60700 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 6.375, |
| "learning_rate": 3.939698492462312e-08, |
| "loss": 0.9714, |
| "step": 60800 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 6.8125, |
| "learning_rate": 3.9296482412060305e-08, |
| "loss": 0.9195, |
| "step": 60900 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 4.46875, |
| "learning_rate": 3.919597989949749e-08, |
| "loss": 0.9541, |
| "step": 61000 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 6.09375, |
| "learning_rate": 3.909547738693467e-08, |
| "loss": 0.9479, |
| "step": 61100 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 4.03125, |
| "learning_rate": 3.899497487437186e-08, |
| "loss": 0.9016, |
| "step": 61200 |
| }, |
| { |
| "epoch": 1.53, |
| "grad_norm": 4.5625, |
| "learning_rate": 3.889447236180905e-08, |
| "loss": 1.0165, |
| "step": 61300 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 5.875, |
| "learning_rate": 3.8793969849246234e-08, |
| "loss": 0.9437, |
| "step": 61400 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 5.90625, |
| "learning_rate": 3.869346733668342e-08, |
| "loss": 0.9593, |
| "step": 61500 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 6.28125, |
| "learning_rate": 3.85929648241206e-08, |
| "loss": 0.9451, |
| "step": 61600 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 8.5, |
| "learning_rate": 3.849246231155779e-08, |
| "loss": 0.9821, |
| "step": 61700 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 4.15625, |
| "learning_rate": 3.8391959798994977e-08, |
| "loss": 0.9196, |
| "step": 61800 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 6.65625, |
| "learning_rate": 3.8291457286432164e-08, |
| "loss": 0.9607, |
| "step": 61900 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 7.0, |
| "learning_rate": 3.819095477386935e-08, |
| "loss": 0.9695, |
| "step": 62000 |
| }, |
| { |
| "epoch": 1.55, |
| "grad_norm": 8.4375, |
| "learning_rate": 3.809045226130653e-08, |
| "loss": 0.9509, |
| "step": 62100 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 8.3125, |
| "learning_rate": 3.798994974874372e-08, |
| "loss": 0.9513, |
| "step": 62200 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 7.875, |
| "learning_rate": 3.7889447236180906e-08, |
| "loss": 0.9426, |
| "step": 62300 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 7.40625, |
| "learning_rate": 3.778894472361809e-08, |
| "loss": 0.9296, |
| "step": 62400 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 6.53125, |
| "learning_rate": 3.768844221105528e-08, |
| "loss": 0.9573, |
| "step": 62500 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 6.625, |
| "learning_rate": 3.758793969849246e-08, |
| "loss": 0.9514, |
| "step": 62600 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 5.75, |
| "learning_rate": 3.748743718592965e-08, |
| "loss": 0.9511, |
| "step": 62700 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 6.09375, |
| "learning_rate": 3.7386934673366836e-08, |
| "loss": 0.9011, |
| "step": 62800 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 7.15625, |
| "learning_rate": 3.728643216080402e-08, |
| "loss": 0.9259, |
| "step": 62900 |
| }, |
| { |
| "epoch": 1.57, |
| "grad_norm": 4.53125, |
| "learning_rate": 3.7185929648241203e-08, |
| "loss": 0.9448, |
| "step": 63000 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 5.53125, |
| "learning_rate": 3.708542713567839e-08, |
| "loss": 0.9775, |
| "step": 63100 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 8.25, |
| "learning_rate": 3.698492462311558e-08, |
| "loss": 0.9529, |
| "step": 63200 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 5.84375, |
| "learning_rate": 3.6884422110552765e-08, |
| "loss": 0.9213, |
| "step": 63300 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 5.03125, |
| "learning_rate": 3.678391959798995e-08, |
| "loss": 0.9599, |
| "step": 63400 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 4.9375, |
| "learning_rate": 3.668341708542713e-08, |
| "loss": 0.9787, |
| "step": 63500 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 11.4375, |
| "learning_rate": 3.658291457286432e-08, |
| "loss": 0.9525, |
| "step": 63600 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 5.3125, |
| "learning_rate": 3.648241206030151e-08, |
| "loss": 0.9541, |
| "step": 63700 |
| }, |
| { |
| "epoch": 1.59, |
| "grad_norm": 7.0625, |
| "learning_rate": 3.6381909547738695e-08, |
| "loss": 1.0157, |
| "step": 63800 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 6.8125, |
| "learning_rate": 3.628140703517588e-08, |
| "loss": 0.9937, |
| "step": 63900 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 7.21875, |
| "learning_rate": 3.618090452261306e-08, |
| "loss": 1.0143, |
| "step": 64000 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 5.625, |
| "learning_rate": 3.608040201005025e-08, |
| "loss": 0.9805, |
| "step": 64100 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 6.875, |
| "learning_rate": 3.597989949748744e-08, |
| "loss": 0.9366, |
| "step": 64200 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 4.875, |
| "learning_rate": 3.5879396984924624e-08, |
| "loss": 0.981, |
| "step": 64300 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 9.25, |
| "learning_rate": 3.577889447236181e-08, |
| "loss": 0.9633, |
| "step": 64400 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 5.625, |
| "learning_rate": 3.567839195979899e-08, |
| "loss": 0.9892, |
| "step": 64500 |
| }, |
| { |
| "epoch": 1.61, |
| "grad_norm": 6.21875, |
| "learning_rate": 3.557788944723618e-08, |
| "loss": 0.9169, |
| "step": 64600 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 6.75, |
| "learning_rate": 3.5477386934673366e-08, |
| "loss": 1.0164, |
| "step": 64700 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 7.28125, |
| "learning_rate": 3.5376884422110554e-08, |
| "loss": 0.9512, |
| "step": 64800 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 9.1875, |
| "learning_rate": 3.5276381909547734e-08, |
| "loss": 0.9667, |
| "step": 64900 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 3.984375, |
| "learning_rate": 3.517587939698492e-08, |
| "loss": 0.9844, |
| "step": 65000 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 6.6875, |
| "learning_rate": 3.507537688442211e-08, |
| "loss": 0.9412, |
| "step": 65100 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 5.90625, |
| "learning_rate": 3.4974874371859296e-08, |
| "loss": 0.9439, |
| "step": 65200 |
| }, |
| { |
| "epoch": 1.63, |
| "grad_norm": 6.875, |
| "learning_rate": 3.487437185929648e-08, |
| "loss": 0.9834, |
| "step": 65300 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 7.5, |
| "learning_rate": 3.4773869346733664e-08, |
| "loss": 0.8961, |
| "step": 65400 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 6.6875, |
| "learning_rate": 3.467336683417085e-08, |
| "loss": 0.9536, |
| "step": 65500 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 6.46875, |
| "learning_rate": 3.457286432160804e-08, |
| "loss": 0.9665, |
| "step": 65600 |
| }, |
| { |
| "epoch": 1.64, |
| "grad_norm": 5.40625, |
| "learning_rate": 3.447236180904523e-08, |
| "loss": 0.9056, |
| "step": 65700 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 6.375, |
| "learning_rate": 3.437185929648241e-08, |
| "loss": 0.9048, |
| "step": 65800 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 7.46875, |
| "learning_rate": 3.42713567839196e-08, |
| "loss": 0.9802, |
| "step": 65900 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 5.875, |
| "learning_rate": 3.417085427135679e-08, |
| "loss": 0.9669, |
| "step": 66000 |
| }, |
| { |
| "epoch": 1.65, |
| "grad_norm": 6.75, |
| "learning_rate": 3.4070351758793974e-08, |
| "loss": 0.9424, |
| "step": 66100 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 7.0, |
| "learning_rate": 3.396984924623116e-08, |
| "loss": 0.9306, |
| "step": 66200 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 6.1875, |
| "learning_rate": 3.386934673366834e-08, |
| "loss": 0.9764, |
| "step": 66300 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 9.4375, |
| "learning_rate": 3.376884422110553e-08, |
| "loss": 1.0015, |
| "step": 66400 |
| }, |
| { |
| "epoch": 1.66, |
| "grad_norm": 7.8125, |
| "learning_rate": 3.3668341708542717e-08, |
| "loss": 0.9376, |
| "step": 66500 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 4.71875, |
| "learning_rate": 3.3567839195979904e-08, |
| "loss": 0.9352, |
| "step": 66600 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 5.90625, |
| "learning_rate": 3.346733668341709e-08, |
| "loss": 0.895, |
| "step": 66700 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 6.21875, |
| "learning_rate": 3.336683417085427e-08, |
| "loss": 0.9482, |
| "step": 66800 |
| }, |
| { |
| "epoch": 1.67, |
| "grad_norm": 3.859375, |
| "learning_rate": 3.326633165829146e-08, |
| "loss": 1.0106, |
| "step": 66900 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 4.90625, |
| "learning_rate": 3.3165829145728646e-08, |
| "loss": 0.9115, |
| "step": 67000 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 5.03125, |
| "learning_rate": 3.306532663316583e-08, |
| "loss": 0.9384, |
| "step": 67100 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 8.375, |
| "learning_rate": 3.2964824120603014e-08, |
| "loss": 0.9182, |
| "step": 67200 |
| }, |
| { |
| "epoch": 1.68, |
| "grad_norm": 4.9375, |
| "learning_rate": 3.28643216080402e-08, |
| "loss": 0.9324, |
| "step": 67300 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 5.96875, |
| "learning_rate": 3.276381909547739e-08, |
| "loss": 0.9554, |
| "step": 67400 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 3.46875, |
| "learning_rate": 3.2663316582914576e-08, |
| "loss": 0.9322, |
| "step": 67500 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 5.75, |
| "learning_rate": 3.256281407035176e-08, |
| "loss": 1.0046, |
| "step": 67600 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 7.625, |
| "learning_rate": 3.2462311557788943e-08, |
| "loss": 0.9466, |
| "step": 67700 |
| }, |
| { |
| "epoch": 1.69, |
| "grad_norm": 4.46875, |
| "learning_rate": 3.236180904522613e-08, |
| "loss": 0.9562, |
| "step": 67800 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 8.9375, |
| "learning_rate": 3.226130653266332e-08, |
| "loss": 0.9675, |
| "step": 67900 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 5.3125, |
| "learning_rate": 3.2160804020100505e-08, |
| "loss": 0.9889, |
| "step": 68000 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 7.1875, |
| "learning_rate": 3.206030150753769e-08, |
| "loss": 0.9679, |
| "step": 68100 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 5.0625, |
| "learning_rate": 3.195979899497487e-08, |
| "loss": 0.9824, |
| "step": 68200 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 9.9375, |
| "learning_rate": 3.185929648241206e-08, |
| "loss": 0.9237, |
| "step": 68300 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 6.28125, |
| "learning_rate": 3.175879396984925e-08, |
| "loss": 0.9529, |
| "step": 68400 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 5.96875, |
| "learning_rate": 3.1658291457286435e-08, |
| "loss": 0.9364, |
| "step": 68500 |
| }, |
| { |
| "epoch": 1.71, |
| "grad_norm": 4.84375, |
| "learning_rate": 3.155778894472362e-08, |
| "loss": 0.9004, |
| "step": 68600 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 4.3125, |
| "learning_rate": 3.14572864321608e-08, |
| "loss": 0.9296, |
| "step": 68700 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 5.0625, |
| "learning_rate": 3.135678391959799e-08, |
| "loss": 0.9691, |
| "step": 68800 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 6.4375, |
| "learning_rate": 3.125628140703518e-08, |
| "loss": 0.9076, |
| "step": 68900 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 5.34375, |
| "learning_rate": 3.1155778894472364e-08, |
| "loss": 0.8654, |
| "step": 69000 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 6.5625, |
| "learning_rate": 3.105527638190955e-08, |
| "loss": 0.9213, |
| "step": 69100 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 8.375, |
| "learning_rate": 3.095477386934673e-08, |
| "loss": 0.9055, |
| "step": 69200 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 7.53125, |
| "learning_rate": 3.085427135678392e-08, |
| "loss": 0.9347, |
| "step": 69300 |
| }, |
| { |
| "epoch": 1.73, |
| "grad_norm": 5.9375, |
| "learning_rate": 3.0753768844221106e-08, |
| "loss": 1.0156, |
| "step": 69400 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 6.15625, |
| "learning_rate": 3.0653266331658294e-08, |
| "loss": 0.9217, |
| "step": 69500 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 6.6875, |
| "learning_rate": 3.0552763819095474e-08, |
| "loss": 0.9123, |
| "step": 69600 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 10.8125, |
| "learning_rate": 3.045226130653266e-08, |
| "loss": 0.9781, |
| "step": 69700 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 4.96875, |
| "learning_rate": 3.035175879396985e-08, |
| "loss": 0.9449, |
| "step": 69800 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 4.75, |
| "learning_rate": 3.0251256281407036e-08, |
| "loss": 1.0215, |
| "step": 69900 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 6.21875, |
| "learning_rate": 3.015075376884422e-08, |
| "loss": 0.9357, |
| "step": 70000 |
| }, |
| { |
| "epoch": 1.75, |
| "eval_loss": 0.9999986290931702, |
| "eval_runtime": 60.0527, |
| "eval_samples_per_second": 16.652, |
| "eval_steps_per_second": 4.163, |
| "step": 70000 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 7.78125, |
| "learning_rate": 3.0050251256281404e-08, |
| "loss": 0.9616, |
| "step": 70100 |
| }, |
| { |
| "epoch": 1.75, |
| "grad_norm": 5.46875, |
| "learning_rate": 2.994974874371859e-08, |
| "loss": 0.9282, |
| "step": 70200 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 7.0625, |
| "learning_rate": 2.984924623115578e-08, |
| "loss": 0.9819, |
| "step": 70300 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 11.6875, |
| "learning_rate": 2.9748743718592965e-08, |
| "loss": 1.0126, |
| "step": 70400 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 6.53125, |
| "learning_rate": 2.964824120603015e-08, |
| "loss": 0.952, |
| "step": 70500 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 6.71875, |
| "learning_rate": 2.9547738693467337e-08, |
| "loss": 0.9167, |
| "step": 70600 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 4.875, |
| "learning_rate": 2.944723618090452e-08, |
| "loss": 0.959, |
| "step": 70700 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 6.78125, |
| "learning_rate": 2.9346733668341708e-08, |
| "loss": 0.9443, |
| "step": 70800 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 6.125, |
| "learning_rate": 2.924623115577889e-08, |
| "loss": 0.9041, |
| "step": 70900 |
| }, |
| { |
| "epoch": 1.77, |
| "grad_norm": 10.25, |
| "learning_rate": 2.914572864321608e-08, |
| "loss": 0.9468, |
| "step": 71000 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 5.65625, |
| "learning_rate": 2.9045226130653266e-08, |
| "loss": 0.9685, |
| "step": 71100 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 9.0625, |
| "learning_rate": 2.894472361809045e-08, |
| "loss": 0.9723, |
| "step": 71200 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 10.1875, |
| "learning_rate": 2.8844221105527637e-08, |
| "loss": 0.9101, |
| "step": 71300 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 7.0625, |
| "learning_rate": 2.874371859296482e-08, |
| "loss": 0.9806, |
| "step": 71400 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 11.4375, |
| "learning_rate": 2.8643216080402008e-08, |
| "loss": 1.0012, |
| "step": 71500 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 6.6875, |
| "learning_rate": 2.8542713567839196e-08, |
| "loss": 0.9899, |
| "step": 71600 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 6.6875, |
| "learning_rate": 2.844221105527638e-08, |
| "loss": 0.9397, |
| "step": 71700 |
| }, |
| { |
| "epoch": 1.79, |
| "grad_norm": 5.46875, |
| "learning_rate": 2.8341708542713567e-08, |
| "loss": 0.982, |
| "step": 71800 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 6.9375, |
| "learning_rate": 2.8241206030150757e-08, |
| "loss": 0.9602, |
| "step": 71900 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 6.90625, |
| "learning_rate": 2.8140703517587944e-08, |
| "loss": 0.9691, |
| "step": 72000 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 5.375, |
| "learning_rate": 2.8040201005025128e-08, |
| "loss": 0.9466, |
| "step": 72100 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 7.375, |
| "learning_rate": 2.7939698492462316e-08, |
| "loss": 1.0198, |
| "step": 72200 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 6.5625, |
| "learning_rate": 2.78391959798995e-08, |
| "loss": 0.9378, |
| "step": 72300 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 5.34375, |
| "learning_rate": 2.7738693467336687e-08, |
| "loss": 0.9785, |
| "step": 72400 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 5.1875, |
| "learning_rate": 2.763819095477387e-08, |
| "loss": 0.9804, |
| "step": 72500 |
| }, |
| { |
| "epoch": 1.81, |
| "grad_norm": 9.3125, |
| "learning_rate": 2.7537688442211058e-08, |
| "loss": 0.9293, |
| "step": 72600 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 5.71875, |
| "learning_rate": 2.7437185929648245e-08, |
| "loss": 0.8982, |
| "step": 72700 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 7.34375, |
| "learning_rate": 2.733668341708543e-08, |
| "loss": 0.9553, |
| "step": 72800 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 5.6875, |
| "learning_rate": 2.7236180904522616e-08, |
| "loss": 0.9614, |
| "step": 72900 |
| }, |
| { |
| "epoch": 1.82, |
| "grad_norm": 5.9375, |
| "learning_rate": 2.71356783919598e-08, |
| "loss": 0.9744, |
| "step": 73000 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 4.40625, |
| "learning_rate": 2.7035175879396987e-08, |
| "loss": 0.9284, |
| "step": 73100 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 4.40625, |
| "learning_rate": 2.6934673366834175e-08, |
| "loss": 0.9075, |
| "step": 73200 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 8.1875, |
| "learning_rate": 2.683417085427136e-08, |
| "loss": 0.9448, |
| "step": 73300 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 4.9375, |
| "learning_rate": 2.6733668341708546e-08, |
| "loss": 0.9215, |
| "step": 73400 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 5.46875, |
| "learning_rate": 2.663316582914573e-08, |
| "loss": 0.9815, |
| "step": 73500 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 3.90625, |
| "learning_rate": 2.6532663316582917e-08, |
| "loss": 0.9832, |
| "step": 73600 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 8.625, |
| "learning_rate": 2.64321608040201e-08, |
| "loss": 0.9929, |
| "step": 73700 |
| }, |
| { |
| "epoch": 1.84, |
| "grad_norm": 10.125, |
| "learning_rate": 2.6331658291457288e-08, |
| "loss": 0.9937, |
| "step": 73800 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 6.40625, |
| "learning_rate": 2.6231155778894475e-08, |
| "loss": 0.9443, |
| "step": 73900 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 6.3125, |
| "learning_rate": 2.613065326633166e-08, |
| "loss": 0.9822, |
| "step": 74000 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 6.65625, |
| "learning_rate": 2.6030150753768846e-08, |
| "loss": 0.9814, |
| "step": 74100 |
| }, |
| { |
| "epoch": 1.85, |
| "grad_norm": 8.125, |
| "learning_rate": 2.592964824120603e-08, |
| "loss": 0.9854, |
| "step": 74200 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 7.40625, |
| "learning_rate": 2.5829145728643218e-08, |
| "loss": 0.9702, |
| "step": 74300 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 9.25, |
| "learning_rate": 2.5728643216080405e-08, |
| "loss": 1.0065, |
| "step": 74400 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 12.0625, |
| "learning_rate": 2.562814070351759e-08, |
| "loss": 0.9264, |
| "step": 74500 |
| }, |
| { |
| "epoch": 1.86, |
| "grad_norm": 7.0625, |
| "learning_rate": 2.5527638190954776e-08, |
| "loss": 0.9612, |
| "step": 74600 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 5.125, |
| "learning_rate": 2.542713567839196e-08, |
| "loss": 0.9377, |
| "step": 74700 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 4.6875, |
| "learning_rate": 2.5326633165829147e-08, |
| "loss": 0.9855, |
| "step": 74800 |
| }, |
| { |
| "epoch": 1.87, |
| "grad_norm": 7.46875, |
| "learning_rate": 2.522613065326633e-08, |
| "loss": 0.9767, |
| "step": 74900 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 8.375, |
| "learning_rate": 2.5125628140703518e-08, |
| "loss": 0.913, |
| "step": 75000 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 5.78125, |
| "learning_rate": 2.5025125628140705e-08, |
| "loss": 0.957, |
| "step": 75100 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 7.71875, |
| "learning_rate": 2.492462311557789e-08, |
| "loss": 0.9567, |
| "step": 75200 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 5.0, |
| "learning_rate": 2.4824120603015077e-08, |
| "loss": 0.9379, |
| "step": 75300 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 6.5625, |
| "learning_rate": 2.472361809045226e-08, |
| "loss": 0.978, |
| "step": 75400 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 5.625, |
| "learning_rate": 2.4623115577889448e-08, |
| "loss": 0.9688, |
| "step": 75500 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 6.71875, |
| "learning_rate": 2.452261306532663e-08, |
| "loss": 0.9691, |
| "step": 75600 |
| }, |
| { |
| "epoch": 1.89, |
| "grad_norm": 5.09375, |
| "learning_rate": 2.442211055276382e-08, |
| "loss": 0.9414, |
| "step": 75700 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 9.6875, |
| "learning_rate": 2.4321608040201006e-08, |
| "loss": 0.8994, |
| "step": 75800 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 6.5625, |
| "learning_rate": 2.422110552763819e-08, |
| "loss": 0.8903, |
| "step": 75900 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 6.65625, |
| "learning_rate": 2.4120603015075377e-08, |
| "loss": 0.9246, |
| "step": 76000 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 5.90625, |
| "learning_rate": 2.402010050251256e-08, |
| "loss": 0.9351, |
| "step": 76100 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 6.3125, |
| "learning_rate": 2.3919597989949748e-08, |
| "loss": 0.9671, |
| "step": 76200 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 5.53125, |
| "learning_rate": 2.3819095477386936e-08, |
| "loss": 0.9706, |
| "step": 76300 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 5.5625, |
| "learning_rate": 2.371859296482412e-08, |
| "loss": 0.9419, |
| "step": 76400 |
| }, |
| { |
| "epoch": 1.91, |
| "grad_norm": 5.09375, |
| "learning_rate": 2.3618090452261307e-08, |
| "loss": 0.9739, |
| "step": 76500 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 5.375, |
| "learning_rate": 2.3517587939698494e-08, |
| "loss": 0.9722, |
| "step": 76600 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 5.71875, |
| "learning_rate": 2.341708542713568e-08, |
| "loss": 0.9212, |
| "step": 76700 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 5.8125, |
| "learning_rate": 2.3316582914572865e-08, |
| "loss": 0.8863, |
| "step": 76800 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 9.0, |
| "learning_rate": 2.3216080402010052e-08, |
| "loss": 0.9425, |
| "step": 76900 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 6.65625, |
| "learning_rate": 2.3115577889447236e-08, |
| "loss": 0.9497, |
| "step": 77000 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 5.28125, |
| "learning_rate": 2.3015075376884423e-08, |
| "loss": 0.933, |
| "step": 77100 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 11.875, |
| "learning_rate": 2.291457286432161e-08, |
| "loss": 0.9661, |
| "step": 77200 |
| }, |
| { |
| "epoch": 1.93, |
| "grad_norm": 7.875, |
| "learning_rate": 2.2814070351758795e-08, |
| "loss": 0.9015, |
| "step": 77300 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 6.59375, |
| "learning_rate": 2.2713567839195982e-08, |
| "loss": 0.9095, |
| "step": 77400 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 7.21875, |
| "learning_rate": 2.2613065326633166e-08, |
| "loss": 0.9207, |
| "step": 77500 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 4.375, |
| "learning_rate": 2.2512562814070353e-08, |
| "loss": 0.9315, |
| "step": 77600 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 11.0, |
| "learning_rate": 2.241206030150754e-08, |
| "loss": 0.9018, |
| "step": 77700 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 8.4375, |
| "learning_rate": 2.2311557788944724e-08, |
| "loss": 0.9298, |
| "step": 77800 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 6.65625, |
| "learning_rate": 2.221105527638191e-08, |
| "loss": 0.9506, |
| "step": 77900 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 9.3125, |
| "learning_rate": 2.2110552763819095e-08, |
| "loss": 0.9076, |
| "step": 78000 |
| }, |
| { |
| "epoch": 1.95, |
| "grad_norm": 7.5625, |
| "learning_rate": 2.2010050251256282e-08, |
| "loss": 1.0066, |
| "step": 78100 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 5.78125, |
| "learning_rate": 2.1909547738693466e-08, |
| "loss": 0.972, |
| "step": 78200 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 5.25, |
| "learning_rate": 2.1809045226130654e-08, |
| "loss": 0.9311, |
| "step": 78300 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 4.84375, |
| "learning_rate": 2.170854271356784e-08, |
| "loss": 0.9604, |
| "step": 78400 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 2.796875, |
| "learning_rate": 2.1608040201005025e-08, |
| "loss": 0.8835, |
| "step": 78500 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 7.15625, |
| "learning_rate": 2.1507537688442212e-08, |
| "loss": 0.9863, |
| "step": 78600 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 6.28125, |
| "learning_rate": 2.1407035175879396e-08, |
| "loss": 0.9428, |
| "step": 78700 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 3.921875, |
| "learning_rate": 2.1306532663316583e-08, |
| "loss": 0.9335, |
| "step": 78800 |
| }, |
| { |
| "epoch": 1.97, |
| "grad_norm": 5.03125, |
| "learning_rate": 2.120603015075377e-08, |
| "loss": 1.0154, |
| "step": 78900 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 6.15625, |
| "learning_rate": 2.1105527638190954e-08, |
| "loss": 0.9613, |
| "step": 79000 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 5.21875, |
| "learning_rate": 2.100502512562814e-08, |
| "loss": 0.9645, |
| "step": 79100 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 9.875, |
| "learning_rate": 2.0904522613065325e-08, |
| "loss": 0.9791, |
| "step": 79200 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 10.3125, |
| "learning_rate": 2.0804020100502513e-08, |
| "loss": 0.9765, |
| "step": 79300 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 6.28125, |
| "learning_rate": 2.0703517587939696e-08, |
| "loss": 0.9877, |
| "step": 79400 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 6.71875, |
| "learning_rate": 2.0603015075376884e-08, |
| "loss": 0.9844, |
| "step": 79500 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 7.21875, |
| "learning_rate": 2.050251256281407e-08, |
| "loss": 0.9206, |
| "step": 79600 |
| }, |
| { |
| "epoch": 1.99, |
| "grad_norm": 7.34375, |
| "learning_rate": 2.0402010050251258e-08, |
| "loss": 0.9705, |
| "step": 79700 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 6.9375, |
| "learning_rate": 2.0301507537688445e-08, |
| "loss": 0.972, |
| "step": 79800 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 4.5625, |
| "learning_rate": 2.020100502512563e-08, |
| "loss": 0.9654, |
| "step": 79900 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 7.0625, |
| "learning_rate": 2.0100502512562817e-08, |
| "loss": 0.9442, |
| "step": 80000 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_loss": 1.016454815864563, |
| "eval_runtime": 60.0142, |
| "eval_samples_per_second": 16.663, |
| "eval_steps_per_second": 4.166, |
| "step": 80000 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 6.28125, |
| "learning_rate": 2e-08, |
| "loss": 0.9476, |
| "step": 80100 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 5.0625, |
| "learning_rate": 1.9899497487437188e-08, |
| "loss": 0.9094, |
| "step": 80200 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 6.1875, |
| "learning_rate": 1.979899497487437e-08, |
| "loss": 0.955, |
| "step": 80300 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 6.125, |
| "learning_rate": 1.969849246231156e-08, |
| "loss": 0.9935, |
| "step": 80400 |
| }, |
| { |
| "epoch": 2.01, |
| "grad_norm": 8.125, |
| "learning_rate": 1.9597989949748746e-08, |
| "loss": 0.9063, |
| "step": 80500 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 6.5625, |
| "learning_rate": 1.949748743718593e-08, |
| "loss": 0.9381, |
| "step": 80600 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 10.125, |
| "learning_rate": 1.9396984924623117e-08, |
| "loss": 0.9802, |
| "step": 80700 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 5.5, |
| "learning_rate": 1.92964824120603e-08, |
| "loss": 0.9377, |
| "step": 80800 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 6.09375, |
| "learning_rate": 1.9195979899497488e-08, |
| "loss": 0.8899, |
| "step": 80900 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 5.75, |
| "learning_rate": 1.9095477386934676e-08, |
| "loss": 0.8906, |
| "step": 81000 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 4.90625, |
| "learning_rate": 1.899497487437186e-08, |
| "loss": 0.9292, |
| "step": 81100 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 9.8125, |
| "learning_rate": 1.8894472361809047e-08, |
| "loss": 0.8881, |
| "step": 81200 |
| }, |
| { |
| "epoch": 2.03, |
| "grad_norm": 7.3125, |
| "learning_rate": 1.879396984924623e-08, |
| "loss": 0.9303, |
| "step": 81300 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 4.25, |
| "learning_rate": 1.8693467336683418e-08, |
| "loss": 1.0181, |
| "step": 81400 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 4.9375, |
| "learning_rate": 1.8592964824120602e-08, |
| "loss": 0.9946, |
| "step": 81500 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 5.59375, |
| "learning_rate": 1.849246231155779e-08, |
| "loss": 0.9528, |
| "step": 81600 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 6.75, |
| "learning_rate": 1.8391959798994976e-08, |
| "loss": 0.9546, |
| "step": 81700 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 3.875, |
| "learning_rate": 1.829145728643216e-08, |
| "loss": 0.9268, |
| "step": 81800 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 4.34375, |
| "learning_rate": 1.8190954773869347e-08, |
| "loss": 0.9261, |
| "step": 81900 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 10.9375, |
| "learning_rate": 1.809045226130653e-08, |
| "loss": 1.0338, |
| "step": 82000 |
| }, |
| { |
| "epoch": 2.05, |
| "grad_norm": 6.21875, |
| "learning_rate": 1.798994974874372e-08, |
| "loss": 1.0056, |
| "step": 82100 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 4.15625, |
| "learning_rate": 1.7889447236180906e-08, |
| "loss": 0.9034, |
| "step": 82200 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 9.9375, |
| "learning_rate": 1.778894472361809e-08, |
| "loss": 0.9524, |
| "step": 82300 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 8.625, |
| "learning_rate": 1.7688442211055277e-08, |
| "loss": 0.9035, |
| "step": 82400 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 5.78125, |
| "learning_rate": 1.758793969849246e-08, |
| "loss": 0.9823, |
| "step": 82500 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 6.5, |
| "learning_rate": 1.7487437185929648e-08, |
| "loss": 0.9044, |
| "step": 82600 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 4.9375, |
| "learning_rate": 1.7386934673366832e-08, |
| "loss": 0.8919, |
| "step": 82700 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 4.625, |
| "learning_rate": 1.728643216080402e-08, |
| "loss": 0.9323, |
| "step": 82800 |
| }, |
| { |
| "epoch": 2.07, |
| "grad_norm": 4.875, |
| "learning_rate": 1.7185929648241206e-08, |
| "loss": 0.9792, |
| "step": 82900 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 6.03125, |
| "learning_rate": 1.7085427135678394e-08, |
| "loss": 0.9269, |
| "step": 83000 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 4.5, |
| "learning_rate": 1.698492462311558e-08, |
| "loss": 0.9527, |
| "step": 83100 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 6.5625, |
| "learning_rate": 1.6884422110552765e-08, |
| "loss": 0.9316, |
| "step": 83200 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 7.09375, |
| "learning_rate": 1.6783919597989952e-08, |
| "loss": 0.8942, |
| "step": 83300 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 5.34375, |
| "learning_rate": 1.6683417085427136e-08, |
| "loss": 0.9274, |
| "step": 83400 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 5.25, |
| "learning_rate": 1.6582914572864323e-08, |
| "loss": 0.9712, |
| "step": 83500 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 7.96875, |
| "learning_rate": 1.6482412060301507e-08, |
| "loss": 0.931, |
| "step": 83600 |
| }, |
| { |
| "epoch": 2.09, |
| "grad_norm": 4.125, |
| "learning_rate": 1.6381909547738694e-08, |
| "loss": 0.8989, |
| "step": 83700 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 6.4375, |
| "learning_rate": 1.628140703517588e-08, |
| "loss": 0.9929, |
| "step": 83800 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 7.65625, |
| "learning_rate": 1.6180904522613065e-08, |
| "loss": 0.9448, |
| "step": 83900 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 7.40625, |
| "learning_rate": 1.6080402010050253e-08, |
| "loss": 0.9364, |
| "step": 84000 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 7.59375, |
| "learning_rate": 1.5979899497487436e-08, |
| "loss": 0.9086, |
| "step": 84100 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 5.71875, |
| "learning_rate": 1.5879396984924624e-08, |
| "loss": 0.9385, |
| "step": 84200 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 5.375, |
| "learning_rate": 1.577889447236181e-08, |
| "loss": 0.9125, |
| "step": 84300 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 6.5, |
| "learning_rate": 1.5678391959798995e-08, |
| "loss": 0.9316, |
| "step": 84400 |
| }, |
| { |
| "epoch": 2.11, |
| "grad_norm": 5.28125, |
| "learning_rate": 1.5577889447236182e-08, |
| "loss": 0.9358, |
| "step": 84500 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 7.625, |
| "learning_rate": 1.5477386934673366e-08, |
| "loss": 0.9307, |
| "step": 84600 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 5.96875, |
| "learning_rate": 1.5376884422110553e-08, |
| "loss": 0.913, |
| "step": 84700 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 6.71875, |
| "learning_rate": 1.5276381909547737e-08, |
| "loss": 0.8759, |
| "step": 84800 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 6.03125, |
| "learning_rate": 1.5175879396984924e-08, |
| "loss": 0.9576, |
| "step": 84900 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 7.28125, |
| "learning_rate": 1.507537688442211e-08, |
| "loss": 0.9269, |
| "step": 85000 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 5.625, |
| "learning_rate": 1.4974874371859295e-08, |
| "loss": 0.9792, |
| "step": 85100 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 6.21875, |
| "learning_rate": 1.4874371859296483e-08, |
| "loss": 0.9723, |
| "step": 85200 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 6.71875, |
| "learning_rate": 1.4773869346733668e-08, |
| "loss": 0.9035, |
| "step": 85300 |
| }, |
| { |
| "epoch": 2.13, |
| "grad_norm": 6.5625, |
| "learning_rate": 1.4673366834170854e-08, |
| "loss": 0.9527, |
| "step": 85400 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 5.75, |
| "learning_rate": 1.457286432160804e-08, |
| "loss": 0.9798, |
| "step": 85500 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 5.3125, |
| "learning_rate": 1.4472361809045225e-08, |
| "loss": 0.9221, |
| "step": 85600 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 6.9375, |
| "learning_rate": 1.437185929648241e-08, |
| "loss": 0.9874, |
| "step": 85700 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 8.1875, |
| "learning_rate": 1.4271356783919598e-08, |
| "loss": 0.93, |
| "step": 85800 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 5.6875, |
| "learning_rate": 1.4170854271356783e-08, |
| "loss": 0.8976, |
| "step": 85900 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 9.1875, |
| "learning_rate": 1.4070351758793972e-08, |
| "loss": 0.9174, |
| "step": 86000 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 6.96875, |
| "learning_rate": 1.3969849246231158e-08, |
| "loss": 0.9544, |
| "step": 86100 |
| }, |
| { |
| "epoch": 2.15, |
| "grad_norm": 8.875, |
| "learning_rate": 1.3869346733668343e-08, |
| "loss": 0.9221, |
| "step": 86200 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 7.21875, |
| "learning_rate": 1.3768844221105529e-08, |
| "loss": 0.972, |
| "step": 86300 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 7.1875, |
| "learning_rate": 1.3668341708542714e-08, |
| "loss": 0.9527, |
| "step": 86400 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 6.46875, |
| "learning_rate": 1.35678391959799e-08, |
| "loss": 0.9785, |
| "step": 86500 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 5.53125, |
| "learning_rate": 1.3467336683417087e-08, |
| "loss": 0.9369, |
| "step": 86600 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 5.21875, |
| "learning_rate": 1.3366834170854273e-08, |
| "loss": 0.9371, |
| "step": 86700 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 5.5625, |
| "learning_rate": 1.3266331658291458e-08, |
| "loss": 0.9412, |
| "step": 86800 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 6.8125, |
| "learning_rate": 1.3165829145728644e-08, |
| "loss": 0.9501, |
| "step": 86900 |
| }, |
| { |
| "epoch": 2.17, |
| "grad_norm": 6.75, |
| "learning_rate": 1.306532663316583e-08, |
| "loss": 0.8898, |
| "step": 87000 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 5.84375, |
| "learning_rate": 1.2964824120603015e-08, |
| "loss": 0.9686, |
| "step": 87100 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 7.53125, |
| "learning_rate": 1.2864321608040202e-08, |
| "loss": 0.8871, |
| "step": 87200 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 5.25, |
| "learning_rate": 1.2763819095477388e-08, |
| "loss": 0.9987, |
| "step": 87300 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 8.0, |
| "learning_rate": 1.2663316582914574e-08, |
| "loss": 0.9803, |
| "step": 87400 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 4.1875, |
| "learning_rate": 1.2562814070351759e-08, |
| "loss": 0.8711, |
| "step": 87500 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 4.71875, |
| "learning_rate": 1.2462311557788945e-08, |
| "loss": 0.985, |
| "step": 87600 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 5.53125, |
| "learning_rate": 1.236180904522613e-08, |
| "loss": 1.0226, |
| "step": 87700 |
| }, |
| { |
| "epoch": 2.19, |
| "grad_norm": 6.5, |
| "learning_rate": 1.2261306532663316e-08, |
| "loss": 0.9716, |
| "step": 87800 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 7.71875, |
| "learning_rate": 1.2160804020100503e-08, |
| "loss": 0.8701, |
| "step": 87900 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 5.4375, |
| "learning_rate": 1.2060301507537689e-08, |
| "loss": 0.9632, |
| "step": 88000 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 7.0, |
| "learning_rate": 1.1959798994974874e-08, |
| "loss": 0.9337, |
| "step": 88100 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 5.65625, |
| "learning_rate": 1.185929648241206e-08, |
| "loss": 0.9862, |
| "step": 88200 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 4.125, |
| "learning_rate": 1.1758793969849247e-08, |
| "loss": 0.9775, |
| "step": 88300 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 6.34375, |
| "learning_rate": 1.1658291457286433e-08, |
| "loss": 0.9627, |
| "step": 88400 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 9.0625, |
| "learning_rate": 1.1557788944723618e-08, |
| "loss": 0.9698, |
| "step": 88500 |
| }, |
| { |
| "epoch": 2.21, |
| "grad_norm": 4.625, |
| "learning_rate": 1.1457286432160805e-08, |
| "loss": 0.9363, |
| "step": 88600 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 7.4375, |
| "learning_rate": 1.1356783919597991e-08, |
| "loss": 0.9233, |
| "step": 88700 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 6.71875, |
| "learning_rate": 1.1256281407035176e-08, |
| "loss": 0.9362, |
| "step": 88800 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 7.9375, |
| "learning_rate": 1.1155778894472362e-08, |
| "loss": 0.9636, |
| "step": 88900 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 6.3125, |
| "learning_rate": 1.1055276381909548e-08, |
| "loss": 0.9665, |
| "step": 89000 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 6.0625, |
| "learning_rate": 1.0954773869346733e-08, |
| "loss": 0.9535, |
| "step": 89100 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 6.90625, |
| "learning_rate": 1.085427135678392e-08, |
| "loss": 0.9221, |
| "step": 89200 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 5.9375, |
| "learning_rate": 1.0753768844221106e-08, |
| "loss": 0.9877, |
| "step": 89300 |
| }, |
| { |
| "epoch": 2.23, |
| "grad_norm": 5.09375, |
| "learning_rate": 1.0653266331658292e-08, |
| "loss": 0.9351, |
| "step": 89400 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 5.25, |
| "learning_rate": 1.0552763819095477e-08, |
| "loss": 0.9609, |
| "step": 89500 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 4.71875, |
| "learning_rate": 1.0452261306532663e-08, |
| "loss": 1.0024, |
| "step": 89600 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 7.0625, |
| "learning_rate": 1.0351758793969848e-08, |
| "loss": 0.905, |
| "step": 89700 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 6.34375, |
| "learning_rate": 1.0251256281407035e-08, |
| "loss": 0.9278, |
| "step": 89800 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 3.5, |
| "learning_rate": 1.0150753768844223e-08, |
| "loss": 0.9558, |
| "step": 89900 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 5.15625, |
| "learning_rate": 1.0050251256281408e-08, |
| "loss": 0.8837, |
| "step": 90000 |
| }, |
| { |
| "epoch": 2.25, |
| "eval_loss": 1.0259228944778442, |
| "eval_runtime": 60.132, |
| "eval_samples_per_second": 16.63, |
| "eval_steps_per_second": 4.158, |
| "step": 90000 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 7.28125, |
| "learning_rate": 9.949748743718594e-09, |
| "loss": 0.952, |
| "step": 90100 |
| }, |
| { |
| "epoch": 2.25, |
| "grad_norm": 5.90625, |
| "learning_rate": 9.84924623115578e-09, |
| "loss": 0.9146, |
| "step": 90200 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 7.375, |
| "learning_rate": 9.748743718592965e-09, |
| "loss": 0.9258, |
| "step": 90300 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 5.78125, |
| "learning_rate": 9.64824120603015e-09, |
| "loss": 0.929, |
| "step": 90400 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 5.09375, |
| "learning_rate": 9.547738693467338e-09, |
| "loss": 0.9446, |
| "step": 90500 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 7.03125, |
| "learning_rate": 9.447236180904523e-09, |
| "loss": 0.9348, |
| "step": 90600 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 5.4375, |
| "learning_rate": 9.346733668341709e-09, |
| "loss": 0.949, |
| "step": 90700 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 6.65625, |
| "learning_rate": 9.246231155778894e-09, |
| "loss": 0.9297, |
| "step": 90800 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 5.0625, |
| "learning_rate": 9.14572864321608e-09, |
| "loss": 0.987, |
| "step": 90900 |
| }, |
| { |
| "epoch": 2.27, |
| "grad_norm": 5.9375, |
| "learning_rate": 9.045226130653266e-09, |
| "loss": 0.9568, |
| "step": 91000 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 8.1875, |
| "learning_rate": 8.944723618090453e-09, |
| "loss": 0.961, |
| "step": 91100 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 8.125, |
| "learning_rate": 8.844221105527638e-09, |
| "loss": 0.996, |
| "step": 91200 |
| }, |
| { |
| "epoch": 2.28, |
| "grad_norm": 8.1875, |
| "learning_rate": 8.743718592964824e-09, |
| "loss": 0.9385, |
| "step": 91300 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 5.96875, |
| "learning_rate": 8.64321608040201e-09, |
| "loss": 0.9266, |
| "step": 91400 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 5.3125, |
| "learning_rate": 8.542713567839197e-09, |
| "loss": 0.9684, |
| "step": 91500 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 5.28125, |
| "learning_rate": 8.442211055276382e-09, |
| "loss": 0.9791, |
| "step": 91600 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 5.84375, |
| "learning_rate": 8.341708542713568e-09, |
| "loss": 0.9238, |
| "step": 91700 |
| }, |
| { |
| "epoch": 2.29, |
| "grad_norm": 4.84375, |
| "learning_rate": 8.241206030150753e-09, |
| "loss": 0.9493, |
| "step": 91800 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 5.625, |
| "learning_rate": 8.14070351758794e-09, |
| "loss": 0.96, |
| "step": 91900 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 5.25, |
| "learning_rate": 8.040201005025126e-09, |
| "loss": 0.9639, |
| "step": 92000 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 6.59375, |
| "learning_rate": 7.939698492462312e-09, |
| "loss": 0.9787, |
| "step": 92100 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 8.6875, |
| "learning_rate": 7.839195979899497e-09, |
| "loss": 1.002, |
| "step": 92200 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 7.34375, |
| "learning_rate": 7.738693467336683e-09, |
| "loss": 0.9659, |
| "step": 92300 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 7.0, |
| "learning_rate": 7.638190954773869e-09, |
| "loss": 0.926, |
| "step": 92400 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 6.28125, |
| "learning_rate": 7.537688442211056e-09, |
| "loss": 1.0108, |
| "step": 92500 |
| }, |
| { |
| "epoch": 2.31, |
| "grad_norm": 5.09375, |
| "learning_rate": 7.437185929648241e-09, |
| "loss": 0.9151, |
| "step": 92600 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 7.71875, |
| "learning_rate": 7.336683417085427e-09, |
| "loss": 0.9009, |
| "step": 92700 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 11.5625, |
| "learning_rate": 7.2361809045226125e-09, |
| "loss": 0.9579, |
| "step": 92800 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 6.78125, |
| "learning_rate": 7.135678391959799e-09, |
| "loss": 0.975, |
| "step": 92900 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 6.84375, |
| "learning_rate": 7.035175879396986e-09, |
| "loss": 1.0061, |
| "step": 93000 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 4.625, |
| "learning_rate": 6.934673366834172e-09, |
| "loss": 0.9624, |
| "step": 93100 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 8.1875, |
| "learning_rate": 6.834170854271357e-09, |
| "loss": 0.9731, |
| "step": 93200 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 9.375, |
| "learning_rate": 6.733668341708544e-09, |
| "loss": 0.952, |
| "step": 93300 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 8.1875, |
| "learning_rate": 6.633165829145729e-09, |
| "loss": 0.9204, |
| "step": 93400 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 6.6875, |
| "learning_rate": 6.532663316582915e-09, |
| "loss": 0.9009, |
| "step": 93500 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 4.96875, |
| "learning_rate": 6.432160804020101e-09, |
| "loss": 1.0315, |
| "step": 93600 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 5.5, |
| "learning_rate": 6.331658291457287e-09, |
| "loss": 0.972, |
| "step": 93700 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 3.328125, |
| "learning_rate": 6.231155778894472e-09, |
| "loss": 0.9263, |
| "step": 93800 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 6.8125, |
| "learning_rate": 6.130653266331658e-09, |
| "loss": 0.9461, |
| "step": 93900 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 6.15625, |
| "learning_rate": 6.030150753768844e-09, |
| "loss": 0.9521, |
| "step": 94000 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 5.4375, |
| "learning_rate": 5.92964824120603e-09, |
| "loss": 1.0075, |
| "step": 94100 |
| }, |
| { |
| "epoch": 2.35, |
| "grad_norm": 6.15625, |
| "learning_rate": 5.829145728643216e-09, |
| "loss": 1.0315, |
| "step": 94200 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 5.09375, |
| "learning_rate": 5.728643216080403e-09, |
| "loss": 0.9098, |
| "step": 94300 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 6.5, |
| "learning_rate": 5.628140703517588e-09, |
| "loss": 1.005, |
| "step": 94400 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 4.71875, |
| "learning_rate": 5.527638190954774e-09, |
| "loss": 0.9818, |
| "step": 94500 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 5.125, |
| "learning_rate": 5.42713567839196e-09, |
| "loss": 0.9769, |
| "step": 94600 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 4.875, |
| "learning_rate": 5.326633165829146e-09, |
| "loss": 0.9299, |
| "step": 94700 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 9.0625, |
| "learning_rate": 5.226130653266331e-09, |
| "loss": 0.9481, |
| "step": 94800 |
| }, |
| { |
| "epoch": 2.37, |
| "grad_norm": 6.9375, |
| "learning_rate": 5.125628140703518e-09, |
| "loss": 0.9804, |
| "step": 94900 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 5.625, |
| "learning_rate": 5.025125628140704e-09, |
| "loss": 0.965, |
| "step": 95000 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 5.75, |
| "learning_rate": 4.92462311557789e-09, |
| "loss": 0.926, |
| "step": 95100 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 7.125, |
| "learning_rate": 4.824120603015075e-09, |
| "loss": 0.9498, |
| "step": 95200 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 4.78125, |
| "learning_rate": 4.723618090452262e-09, |
| "loss": 0.9698, |
| "step": 95300 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 7.8125, |
| "learning_rate": 4.623115577889447e-09, |
| "loss": 0.988, |
| "step": 95400 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 6.0625, |
| "learning_rate": 4.522613065326633e-09, |
| "loss": 0.9923, |
| "step": 95500 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 7.59375, |
| "learning_rate": 4.422110552763819e-09, |
| "loss": 0.9814, |
| "step": 95600 |
| }, |
| { |
| "epoch": 2.39, |
| "grad_norm": 6.53125, |
| "learning_rate": 4.321608040201005e-09, |
| "loss": 0.9604, |
| "step": 95700 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 6.25, |
| "learning_rate": 4.221105527638191e-09, |
| "loss": 1.0047, |
| "step": 95800 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 8.0, |
| "learning_rate": 4.120603015075377e-09, |
| "loss": 1.0077, |
| "step": 95900 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 5.96875, |
| "learning_rate": 4.020100502512563e-09, |
| "loss": 0.899, |
| "step": 96000 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 4.875, |
| "learning_rate": 3.919597989949749e-09, |
| "loss": 1.0192, |
| "step": 96100 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 5.0, |
| "learning_rate": 3.819095477386934e-09, |
| "loss": 0.9151, |
| "step": 96200 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 5.34375, |
| "learning_rate": 3.7185929648241207e-09, |
| "loss": 0.9238, |
| "step": 96300 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 5.25, |
| "learning_rate": 3.6180904522613062e-09, |
| "loss": 0.9298, |
| "step": 96400 |
| }, |
| { |
| "epoch": 2.41, |
| "grad_norm": 6.65625, |
| "learning_rate": 3.517587939698493e-09, |
| "loss": 0.9592, |
| "step": 96500 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 4.59375, |
| "learning_rate": 3.4170854271356786e-09, |
| "loss": 0.9678, |
| "step": 96600 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 6.75, |
| "learning_rate": 3.3165829145728646e-09, |
| "loss": 0.9092, |
| "step": 96700 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 5.28125, |
| "learning_rate": 3.2160804020100506e-09, |
| "loss": 0.9392, |
| "step": 96800 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 5.34375, |
| "learning_rate": 3.115577889447236e-09, |
| "loss": 0.9643, |
| "step": 96900 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 5.625, |
| "learning_rate": 3.015075376884422e-09, |
| "loss": 0.9835, |
| "step": 97000 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 6.5, |
| "learning_rate": 2.914572864321608e-09, |
| "loss": 0.8944, |
| "step": 97100 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 5.5625, |
| "learning_rate": 2.814070351758794e-09, |
| "loss": 1.0095, |
| "step": 97200 |
| }, |
| { |
| "epoch": 2.43, |
| "grad_norm": 10.0, |
| "learning_rate": 2.71356783919598e-09, |
| "loss": 0.9684, |
| "step": 97300 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 6.59375, |
| "learning_rate": 2.6130653266331657e-09, |
| "loss": 0.9435, |
| "step": 97400 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 5.0, |
| "learning_rate": 2.512562814070352e-09, |
| "loss": 0.9796, |
| "step": 97500 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 5.84375, |
| "learning_rate": 2.4120603015075376e-09, |
| "loss": 0.9515, |
| "step": 97600 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 7.09375, |
| "learning_rate": 2.3115577889447236e-09, |
| "loss": 0.9792, |
| "step": 97700 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 5.28125, |
| "learning_rate": 2.2110552763819096e-09, |
| "loss": 0.931, |
| "step": 97800 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 6.40625, |
| "learning_rate": 2.1105527638190956e-09, |
| "loss": 0.9806, |
| "step": 97900 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 3.9375, |
| "learning_rate": 2.0100502512562816e-09, |
| "loss": 0.9843, |
| "step": 98000 |
| }, |
| { |
| "epoch": 2.45, |
| "grad_norm": 5.5, |
| "learning_rate": 1.909547738693467e-09, |
| "loss": 0.9771, |
| "step": 98100 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 5.8125, |
| "learning_rate": 1.8090452261306531e-09, |
| "loss": 0.9804, |
| "step": 98200 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 7.09375, |
| "learning_rate": 1.7085427135678393e-09, |
| "loss": 0.939, |
| "step": 98300 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 7.78125, |
| "learning_rate": 1.6080402010050253e-09, |
| "loss": 0.9246, |
| "step": 98400 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 10.5, |
| "learning_rate": 1.507537688442211e-09, |
| "loss": 0.9398, |
| "step": 98500 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 4.5, |
| "learning_rate": 1.407035175879397e-09, |
| "loss": 0.9882, |
| "step": 98600 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 9.6875, |
| "learning_rate": 1.3065326633165828e-09, |
| "loss": 1.0052, |
| "step": 98700 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 5.96875, |
| "learning_rate": 1.2060301507537688e-09, |
| "loss": 0.9882, |
| "step": 98800 |
| }, |
| { |
| "epoch": 2.47, |
| "grad_norm": 4.34375, |
| "learning_rate": 1.1055276381909548e-09, |
| "loss": 0.9811, |
| "step": 98900 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 5.3125, |
| "learning_rate": 1.0050251256281408e-09, |
| "loss": 0.9998, |
| "step": 99000 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 6.0625, |
| "learning_rate": 9.045226130653266e-10, |
| "loss": 0.9428, |
| "step": 99100 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 6.3125, |
| "learning_rate": 8.040201005025126e-10, |
| "loss": 1.0201, |
| "step": 99200 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 6.6875, |
| "learning_rate": 7.035175879396985e-10, |
| "loss": 0.9774, |
| "step": 99300 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 8.625, |
| "learning_rate": 6.030150753768844e-10, |
| "loss": 1.0282, |
| "step": 99400 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 6.125, |
| "learning_rate": 5.025125628140704e-10, |
| "loss": 0.9233, |
| "step": 99500 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 10.25, |
| "learning_rate": 4.020100502512563e-10, |
| "loss": 0.9259, |
| "step": 99600 |
| }, |
| { |
| "epoch": 2.49, |
| "grad_norm": 6.125, |
| "learning_rate": 3.015075376884422e-10, |
| "loss": 0.974, |
| "step": 99700 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 7.40625, |
| "learning_rate": 2.0100502512562816e-10, |
| "loss": 0.9634, |
| "step": 99800 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 7.65625, |
| "learning_rate": 1.0050251256281408e-10, |
| "loss": 0.98, |
| "step": 99900 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 4.84375, |
| "learning_rate": 0.0, |
| "loss": 0.9418, |
| "step": 100000 |
| }, |
| { |
| "epoch": 2.5, |
| "eval_loss": 1.0053296089172363, |
| "eval_runtime": 60.0342, |
| "eval_samples_per_second": 16.657, |
| "eval_steps_per_second": 4.164, |
| "step": 100000 |
| } |
| ], |
| "logging_steps": 100, |
| "max_steps": 100000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 10000, |
| "total_flos": 1.5733698330624e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
