| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.3439955711364746, | |
| "learning_rate": 0.0001, | |
| "loss": 5.4154, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.3439955711364746, | |
| "learning_rate": 0.0002, | |
| "loss": 5.4154, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.64585542678833, | |
| "learning_rate": 0.00019795918367346938, | |
| "loss": 5.1115, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.641846656799316, | |
| "learning_rate": 0.0001959183673469388, | |
| "loss": 4.5815, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 10.534232139587402, | |
| "learning_rate": 0.00019387755102040816, | |
| "loss": 4.1388, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.00019387755102040816, | |
| "loss": 3.7196, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 13.133545875549316, | |
| "learning_rate": 0.00019183673469387756, | |
| "loss": 3.7196, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 15.314064025878906, | |
| "learning_rate": 0.00018979591836734697, | |
| "loss": 3.2758, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 16.992698669433594, | |
| "learning_rate": 0.00018775510204081634, | |
| "loss": 2.8057, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 18.105308532714844, | |
| "learning_rate": 0.00018571428571428572, | |
| "loss": 2.3037, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 18.624324798583984, | |
| "learning_rate": 0.00018367346938775512, | |
| "loss": 1.7698, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 18.554916381835938, | |
| "learning_rate": 0.0001816326530612245, | |
| "loss": 1.2219, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 17.93878746032715, | |
| "learning_rate": 0.0001795918367346939, | |
| "loss": 0.6767, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 2.917646884918213, | |
| "learning_rate": 0.00017755102040816327, | |
| "loss": 0.1873, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.2019948959350586, | |
| "learning_rate": 0.00017551020408163265, | |
| "loss": 0.1488, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 3.1802170276641846, | |
| "learning_rate": 0.00017346938775510205, | |
| "loss": 0.1134, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.5858300924301147, | |
| "learning_rate": 0.00017142857142857143, | |
| "loss": 0.0832, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.5080211758613586, | |
| "learning_rate": 0.00016938775510204083, | |
| "loss": 0.0701, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.5199990272521973, | |
| "learning_rate": 0.00016734693877551023, | |
| "loss": 0.0623, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.5190378427505493, | |
| "learning_rate": 0.0001653061224489796, | |
| "loss": 0.0552, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.5189256072044373, | |
| "learning_rate": 0.00016326530612244898, | |
| "loss": 0.0456, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.5494657754898071, | |
| "learning_rate": 0.00016122448979591838, | |
| "loss": 0.0356, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.5577398538589478, | |
| "learning_rate": 0.00015918367346938776, | |
| "loss": 0.024, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.39652666449546814, | |
| "learning_rate": 0.00015714285714285716, | |
| "loss": 0.0128, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.12145466357469559, | |
| "learning_rate": 0.00015510204081632654, | |
| "loss": 0.0064, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.16227224469184875, | |
| "learning_rate": 0.0001530612244897959, | |
| "loss": 0.0058, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.010005966760218143, | |
| "learning_rate": 0.0001510204081632653, | |
| "loss": 0.0046, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.11974961310625076, | |
| "learning_rate": 0.00014897959183673472, | |
| "loss": 0.0053, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.026845891028642654, | |
| "learning_rate": 0.0001469387755102041, | |
| "loss": 0.0046, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.09766196459531784, | |
| "learning_rate": 0.0001448979591836735, | |
| "loss": 0.0053, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.05061652511358261, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.0046, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.0490519218146801, | |
| "learning_rate": 0.00014081632653061224, | |
| "loss": 0.0048, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.06446336209774017, | |
| "learning_rate": 0.00013877551020408165, | |
| "loss": 0.0049, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.007198956795036793, | |
| "learning_rate": 0.00013673469387755102, | |
| "loss": 0.0045, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.06024309992790222, | |
| "learning_rate": 0.0001346938775510204, | |
| "loss": 0.0048, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 0.04099668189883232, | |
| "learning_rate": 0.0001326530612244898, | |
| "loss": 0.0048, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "grad_norm": 0.030607614666223526, | |
| "learning_rate": 0.00013061224489795917, | |
| "loss": 0.0047, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 0.04647354409098625, | |
| "learning_rate": 0.00012857142857142858, | |
| "loss": 0.0048, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "grad_norm": 0.013468354940414429, | |
| "learning_rate": 0.00012653061224489798, | |
| "loss": 0.0047, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.02858574502170086, | |
| "learning_rate": 0.00012448979591836735, | |
| "loss": 0.0045, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "grad_norm": 0.039958544075489044, | |
| "learning_rate": 0.00012244897959183676, | |
| "loss": 0.0048, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 0.003950281068682671, | |
| "learning_rate": 0.00012040816326530613, | |
| "loss": 0.0047, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "grad_norm": 0.029744163155555725, | |
| "learning_rate": 0.00011836734693877552, | |
| "loss": 0.0047, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 0.029589517042040825, | |
| "learning_rate": 0.0001163265306122449, | |
| "loss": 0.0047, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 0.0025897289160639048, | |
| "learning_rate": 0.00011428571428571428, | |
| "loss": 0.0049, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "grad_norm": 0.03239087387919426, | |
| "learning_rate": 0.00011224489795918367, | |
| "loss": 0.005, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "grad_norm": 0.021254172548651695, | |
| "learning_rate": 0.00011020408163265306, | |
| "loss": 0.0047, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 0.01296665333211422, | |
| "learning_rate": 0.00010816326530612246, | |
| "loss": 0.0047, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "grad_norm": 0.012932351790368557, | |
| "learning_rate": 0.00010612244897959185, | |
| "loss": 0.0047, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.0024836526717990637, | |
| "learning_rate": 0.00010408163265306123, | |
| "loss": 0.0049, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "grad_norm": 0.012770796194672585, | |
| "learning_rate": 0.00010204081632653062, | |
| "loss": 0.0047, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "grad_norm": 0.010142244398593903, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0045, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "grad_norm": 0.020766526460647583, | |
| "learning_rate": 9.79591836734694e-05, | |
| "loss": 0.0047, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "grad_norm": 0.0039475164376199245, | |
| "learning_rate": 9.591836734693878e-05, | |
| "loss": 0.0047, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "grad_norm": 0.012760567478835583, | |
| "learning_rate": 9.387755102040817e-05, | |
| "loss": 0.0047, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "grad_norm": 0.028441807255148888, | |
| "learning_rate": 9.183673469387756e-05, | |
| "loss": 0.0047, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 57.0, | |
| "grad_norm": 0.004045899026095867, | |
| "learning_rate": 8.979591836734695e-05, | |
| "loss": 0.0047, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "grad_norm": 0.020622774958610535, | |
| "learning_rate": 8.775510204081632e-05, | |
| "loss": 0.0047, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 59.0, | |
| "grad_norm": 0.010014132596552372, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.0045, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 0.0038325833156704903, | |
| "learning_rate": 8.367346938775511e-05, | |
| "loss": 0.0047, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "grad_norm": 0.01254805363714695, | |
| "learning_rate": 8.163265306122449e-05, | |
| "loss": 0.0047, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "grad_norm": 0.012542281299829483, | |
| "learning_rate": 7.959183673469388e-05, | |
| "loss": 0.0047, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 63.0, | |
| "grad_norm": 0.006473184563219547, | |
| "learning_rate": 7.755102040816327e-05, | |
| "loss": 0.0045, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "grad_norm": 0.014362330548465252, | |
| "learning_rate": 7.551020408163266e-05, | |
| "loss": 0.0049, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "grad_norm": 0.0204475000500679, | |
| "learning_rate": 7.346938775510205e-05, | |
| "loss": 0.0047, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "grad_norm": 0.0023198979906737804, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.0049, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 67.0, | |
| "grad_norm": 0.018235381692647934, | |
| "learning_rate": 6.938775510204082e-05, | |
| "loss": 0.0049, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "grad_norm": 0.012397863902151585, | |
| "learning_rate": 6.73469387755102e-05, | |
| "loss": 0.0047, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 69.0, | |
| "grad_norm": 0.003842939855530858, | |
| "learning_rate": 6.530612244897959e-05, | |
| "loss": 0.0047, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 0.003821918275207281, | |
| "learning_rate": 6.326530612244899e-05, | |
| "loss": 0.0047, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 71.0, | |
| "grad_norm": 0.0038236272521317005, | |
| "learning_rate": 6.122448979591838e-05, | |
| "loss": 0.0047, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "grad_norm": 0.009961229749023914, | |
| "learning_rate": 5.918367346938776e-05, | |
| "loss": 0.0045, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 73.0, | |
| "grad_norm": 0.0037804129533469677, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.0047, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "grad_norm": 0.003801505547016859, | |
| "learning_rate": 5.510204081632653e-05, | |
| "loss": 0.0047, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "grad_norm": 0.012337015941739082, | |
| "learning_rate": 5.3061224489795926e-05, | |
| "loss": 0.0047, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "grad_norm": 0.012310854159295559, | |
| "learning_rate": 5.102040816326531e-05, | |
| "loss": 0.0047, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 77.0, | |
| "grad_norm": 0.0063883536495268345, | |
| "learning_rate": 4.89795918367347e-05, | |
| "loss": 0.0045, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "grad_norm": 0.0036704791709780693, | |
| "learning_rate": 4.6938775510204086e-05, | |
| "loss": 0.0047, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 79.0, | |
| "grad_norm": 0.0038380566984415054, | |
| "learning_rate": 4.4897959183673474e-05, | |
| "loss": 0.0047, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 0.013979257084429264, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.0049, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 81.0, | |
| "grad_norm": 0.003754157805815339, | |
| "learning_rate": 4.0816326530612245e-05, | |
| "loss": 0.0047, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "grad_norm": 0.003731819801032543, | |
| "learning_rate": 3.8775510204081634e-05, | |
| "loss": 0.0047, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 83.0, | |
| "grad_norm": 0.003740091575309634, | |
| "learning_rate": 3.673469387755102e-05, | |
| "loss": 0.0047, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "grad_norm": 0.002331367926672101, | |
| "learning_rate": 3.469387755102041e-05, | |
| "loss": 0.0049, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "grad_norm": 0.006460592150688171, | |
| "learning_rate": 3.265306122448979e-05, | |
| "loss": 0.0045, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 86.0, | |
| "grad_norm": 0.01226158905774355, | |
| "learning_rate": 3.061224489795919e-05, | |
| "loss": 0.0047, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 87.0, | |
| "grad_norm": 0.012233145534992218, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.0047, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 88.0, | |
| "grad_norm": 0.002268604002892971, | |
| "learning_rate": 2.6530612244897963e-05, | |
| "loss": 0.0049, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 89.0, | |
| "grad_norm": 0.009719762951135635, | |
| "learning_rate": 2.448979591836735e-05, | |
| "loss": 0.0045, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 0.003743007080629468, | |
| "learning_rate": 2.2448979591836737e-05, | |
| "loss": 0.0047, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 91.0, | |
| "grad_norm": 0.00371920526959002, | |
| "learning_rate": 2.0408163265306123e-05, | |
| "loss": 0.0047, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 92.0, | |
| "grad_norm": 0.0036711846478283405, | |
| "learning_rate": 1.836734693877551e-05, | |
| "loss": 0.0047, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 93.0, | |
| "grad_norm": 0.020141731947660446, | |
| "learning_rate": 1.6326530612244897e-05, | |
| "loss": 0.0047, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "grad_norm": 0.0037132452707737684, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.0047, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "grad_norm": 0.0063994950614869595, | |
| "learning_rate": 1.2244897959183674e-05, | |
| "loss": 0.0045, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "grad_norm": 0.0037392491940408945, | |
| "learning_rate": 1.0204081632653061e-05, | |
| "loss": 0.0047, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 97.0, | |
| "grad_norm": 0.003673990024253726, | |
| "learning_rate": 8.163265306122448e-06, | |
| "loss": 0.0047, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 98.0, | |
| "grad_norm": 0.012367010116577148, | |
| "learning_rate": 6.122448979591837e-06, | |
| "loss": 0.0047, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 99.0, | |
| "grad_norm": 0.002525837393477559, | |
| "learning_rate": 4.081632653061224e-06, | |
| "loss": 0.0049, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 0.012382498942315578, | |
| "learning_rate": 2.040816326530612e-06, | |
| "loss": 0.0047, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 713859830784000.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |