| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5, | |
| "eval_steps": 2000, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9697519540786743, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1629, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.3848124742507935, | |
| "learning_rate": 9.898989898989898e-07, | |
| "loss": 0.14, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9986572265625, | |
| "learning_rate": 9.797979797979797e-07, | |
| "loss": 0.1354, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.11438798904418945, | |
| "learning_rate": 9.696969696969698e-07, | |
| "loss": 0.1182, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8548241257667542, | |
| "learning_rate": 9.595959595959596e-07, | |
| "loss": 0.1192, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.5312464237213135, | |
| "learning_rate": 9.494949494949495e-07, | |
| "loss": 0.0997, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9692059755325317, | |
| "learning_rate": 9.393939393939395e-07, | |
| "loss": 0.102, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.42864611744880676, | |
| "learning_rate": 9.292929292929292e-07, | |
| "loss": 0.0901, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.852543830871582, | |
| "learning_rate": 9.191919191919192e-07, | |
| "loss": 0.0937, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.5718303322792053, | |
| "learning_rate": 9.09090909090909e-07, | |
| "loss": 0.093, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9396565556526184, | |
| "learning_rate": 8.98989898989899e-07, | |
| "loss": 0.0892, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.08157779276371002, | |
| "learning_rate": 8.888888888888888e-07, | |
| "loss": 0.0934, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8076322078704834, | |
| "learning_rate": 8.787878787878787e-07, | |
| "loss": 0.0725, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.5076119899749756, | |
| "learning_rate": 8.686868686868687e-07, | |
| "loss": 0.0835, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1567238569259644, | |
| "learning_rate": 8.585858585858586e-07, | |
| "loss": 0.0747, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.6817927956581116, | |
| "learning_rate": 8.484848484848484e-07, | |
| "loss": 0.0903, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.6467050313949585, | |
| "learning_rate": 8.383838383838383e-07, | |
| "loss": 0.0721, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.8435570001602173, | |
| "learning_rate": 8.282828282828283e-07, | |
| "loss": 0.0847, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.6265794634819031, | |
| "learning_rate": 8.181818181818182e-07, | |
| "loss": 0.0687, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.360060453414917, | |
| "learning_rate": 8.08080808080808e-07, | |
| "loss": 0.0748, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 0.06745574623346329, | |
| "eval_runtime": 304.7718, | |
| "eval_samples_per_second": 3.281, | |
| "eval_steps_per_second": 0.82, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.5074820518493652, | |
| "learning_rate": 7.97979797979798e-07, | |
| "loss": 0.0699, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3634223937988281, | |
| "learning_rate": 7.878787878787878e-07, | |
| "loss": 0.0682, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 3.0763113498687744, | |
| "learning_rate": 7.777777777777778e-07, | |
| "loss": 0.0786, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.5437021255493164, | |
| "learning_rate": 7.676767676767675e-07, | |
| "loss": 0.065, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.5890172719955444, | |
| "learning_rate": 7.575757575757575e-07, | |
| "loss": 0.0772, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9904383420944214, | |
| "learning_rate": 7.474747474747475e-07, | |
| "loss": 0.0735, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9233341813087463, | |
| "learning_rate": 7.373737373737373e-07, | |
| "loss": 0.0695, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.10603007674217224, | |
| "learning_rate": 7.272727272727272e-07, | |
| "loss": 0.0595, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9929779767990112, | |
| "learning_rate": 7.171717171717171e-07, | |
| "loss": 0.0595, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.4711184799671173, | |
| "learning_rate": 7.07070707070707e-07, | |
| "loss": 0.0606, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.5627967715263367, | |
| "learning_rate": 6.96969696969697e-07, | |
| "loss": 0.0713, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7645086646080017, | |
| "learning_rate": 6.868686868686868e-07, | |
| "loss": 0.0672, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8071433901786804, | |
| "learning_rate": 6.767676767676767e-07, | |
| "loss": 0.0809, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.281545877456665, | |
| "learning_rate": 6.666666666666666e-07, | |
| "loss": 0.0577, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.15431547164917, | |
| "learning_rate": 6.565656565656566e-07, | |
| "loss": 0.0613, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.42522451281547546, | |
| "learning_rate": 6.464646464646465e-07, | |
| "loss": 0.0531, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2078278064727783, | |
| "learning_rate": 6.363636363636363e-07, | |
| "loss": 0.0692, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.207512855529785, | |
| "learning_rate": 6.262626262626263e-07, | |
| "loss": 0.0508, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.696768045425415, | |
| "learning_rate": 6.161616161616161e-07, | |
| "loss": 0.0655, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.6295761466026306, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 0.0577, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.057661667466163635, | |
| "eval_runtime": 304.4519, | |
| "eval_samples_per_second": 3.285, | |
| "eval_steps_per_second": 0.821, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.7527456283569336, | |
| "learning_rate": 5.959595959595959e-07, | |
| "loss": 0.0575, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.4716615676879883, | |
| "learning_rate": 5.858585858585858e-07, | |
| "loss": 0.0768, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.5344067215919495, | |
| "learning_rate": 5.757575757575758e-07, | |
| "loss": 0.0571, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.9244055151939392, | |
| "learning_rate": 5.656565656565657e-07, | |
| "loss": 0.0563, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.16805100440979, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 0.0629, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.6229370832443237, | |
| "learning_rate": 5.454545454545454e-07, | |
| "loss": 0.0641, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.9978136420249939, | |
| "learning_rate": 5.353535353535354e-07, | |
| "loss": 0.0642, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.4647519886493683, | |
| "learning_rate": 5.252525252525253e-07, | |
| "loss": 0.0661, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.3375582695007324, | |
| "learning_rate": 5.152525252525253e-07, | |
| "loss": 0.0595, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.08847852051258087, | |
| "learning_rate": 5.051515151515151e-07, | |
| "loss": 0.0461, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.0537497997283936, | |
| "learning_rate": 4.95050505050505e-07, | |
| "loss": 0.0494, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9269624948501587, | |
| "learning_rate": 4.849494949494949e-07, | |
| "loss": 0.0714, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.7937563061714172, | |
| "learning_rate": 4.748484848484848e-07, | |
| "loss": 0.065, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.4173548221588135, | |
| "learning_rate": 4.6474747474747473e-07, | |
| "loss": 0.0496, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.37847477197647095, | |
| "learning_rate": 4.546464646464646e-07, | |
| "loss": 0.0651, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.7063392400741577, | |
| "learning_rate": 4.445454545454545e-07, | |
| "loss": 0.0562, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.6490029096603394, | |
| "learning_rate": 4.344444444444444e-07, | |
| "loss": 0.0695, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8921324610710144, | |
| "learning_rate": 4.2434343434343434e-07, | |
| "loss": 0.0733, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.5432973504066467, | |
| "learning_rate": 4.142424242424242e-07, | |
| "loss": 0.0589, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.017520183697342873, | |
| "learning_rate": 4.041414141414141e-07, | |
| "loss": 0.0541, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_loss": 0.06119654327630997, | |
| "eval_runtime": 305.9355, | |
| "eval_samples_per_second": 3.269, | |
| "eval_steps_per_second": 0.817, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.68011212348938, | |
| "learning_rate": 3.94040404040404e-07, | |
| "loss": 0.0568, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.3149752616882324, | |
| "learning_rate": 3.839393939393939e-07, | |
| "loss": 0.056, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.04929787665605545, | |
| "learning_rate": 3.738383838383838e-07, | |
| "loss": 0.053, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7595096826553345, | |
| "learning_rate": 3.6373737373737373e-07, | |
| "loss": 0.0651, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.08241419494152069, | |
| "learning_rate": 3.536363636363636e-07, | |
| "loss": 0.0729, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.7712632417678833, | |
| "learning_rate": 3.435353535353535e-07, | |
| "loss": 0.0533, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.8940716981887817, | |
| "learning_rate": 3.3343434343434343e-07, | |
| "loss": 0.0583, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.07956444472074509, | |
| "learning_rate": 3.233333333333333e-07, | |
| "loss": 0.058, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.6199010610580444, | |
| "learning_rate": 3.1323232323232326e-07, | |
| "loss": 0.0568, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.5326449275016785, | |
| "learning_rate": 3.031313131313131e-07, | |
| "loss": 0.0525, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.6375436186790466, | |
| "learning_rate": 2.9303030303030304e-07, | |
| "loss": 0.0523, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.2404210567474365, | |
| "learning_rate": 2.829292929292929e-07, | |
| "loss": 0.0589, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.2247310876846313, | |
| "learning_rate": 2.728282828282828e-07, | |
| "loss": 0.0586, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.5559460520744324, | |
| "learning_rate": 2.6272727272727273e-07, | |
| "loss": 0.0578, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.2506352961063385, | |
| "learning_rate": 2.5262626262626265e-07, | |
| "loss": 0.0555, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.7911235094070435, | |
| "learning_rate": 2.425252525252525e-07, | |
| "loss": 0.0641, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.21588227152824402, | |
| "learning_rate": 2.324242424242424e-07, | |
| "loss": 0.068, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.5807000994682312, | |
| "learning_rate": 2.223232323232323e-07, | |
| "loss": 0.061, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.298220157623291, | |
| "learning_rate": 2.122222222222222e-07, | |
| "loss": 0.0594, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.0665823221206665, | |
| "learning_rate": 2.021212121212121e-07, | |
| "loss": 0.052, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.050024934113025665, | |
| "eval_runtime": 304.3337, | |
| "eval_samples_per_second": 3.286, | |
| "eval_steps_per_second": 0.821, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.76180499792099, | |
| "learning_rate": 1.92020202020202e-07, | |
| "loss": 0.0611, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.549780547618866, | |
| "learning_rate": 1.8191919191919193e-07, | |
| "loss": 0.0612, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 2.0332558155059814, | |
| "learning_rate": 1.7181818181818182e-07, | |
| "loss": 0.0545, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.8247748017311096, | |
| "learning_rate": 1.617171717171717e-07, | |
| "loss": 0.0616, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.0830433368682861, | |
| "learning_rate": 1.5161616161616162e-07, | |
| "loss": 0.0553, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.06752818822860718, | |
| "learning_rate": 1.4151515151515151e-07, | |
| "loss": 0.044, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.645291805267334, | |
| "learning_rate": 1.3141414141414143e-07, | |
| "loss": 0.049, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.5109922289848328, | |
| "learning_rate": 1.2141414141414142e-07, | |
| "loss": 0.0597, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.7064566612243652, | |
| "learning_rate": 1.1131313131313131e-07, | |
| "loss": 0.0731, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.9522804617881775, | |
| "learning_rate": 1.0121212121212121e-07, | |
| "loss": 0.0641, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.425570011138916, | |
| "learning_rate": 9.111111111111112e-08, | |
| "loss": 0.0531, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.16694176197052, | |
| "learning_rate": 8.1010101010101e-08, | |
| "loss": 0.0517, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.6792075634002686, | |
| "learning_rate": 7.090909090909091e-08, | |
| "loss": 0.0644, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.2597424983978271, | |
| "learning_rate": 6.080808080808081e-08, | |
| "loss": 0.0528, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 4.67294454574585, | |
| "learning_rate": 5.0707070707070707e-08, | |
| "loss": 0.0559, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.676389455795288, | |
| "learning_rate": 4.06060606060606e-08, | |
| "loss": 0.0589, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.2427666187286377, | |
| "learning_rate": 3.0505050505050505e-08, | |
| "loss": 0.0645, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.02416997402906418, | |
| "learning_rate": 2.0404040404040402e-08, | |
| "loss": 0.0453, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 2.5054471492767334, | |
| "learning_rate": 1.0303030303030303e-08, | |
| "loss": 0.0491, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6195088028907776, | |
| "learning_rate": 2.0202020202020202e-10, | |
| "loss": 0.054, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.059650979936122894, | |
| "eval_runtime": 304.5847, | |
| "eval_samples_per_second": 3.283, | |
| "eval_steps_per_second": 0.821, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "total_flos": 8.1602751234048e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |