| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 126, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 14.136896419046405, | |
| "learning_rate": 9.998445910004082e-06, | |
| "loss": 1.2293, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 9.414916379661909, | |
| "learning_rate": 9.993784606094612e-06, | |
| "loss": 0.8742, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 7.055571354439035, | |
| "learning_rate": 9.986018985905901e-06, | |
| "loss": 0.6603, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 6.0109678713542065, | |
| "learning_rate": 9.975153876827008e-06, | |
| "loss": 0.6402, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 6.286672268332456, | |
| "learning_rate": 9.961196033000862e-06, | |
| "loss": 0.5669, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 3.75907915574848, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.4859, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 2.7783434693295717, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.4665, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 3.5757346497904514, | |
| "learning_rate": 9.900862439242719e-06, | |
| "loss": 0.4592, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 3.0154588649031027, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.4289, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 2.9640179533460693, | |
| "learning_rate": 9.84538643114539e-06, | |
| "loss": 0.4375, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 2.4616695644526505, | |
| "learning_rate": 9.81312123475006e-06, | |
| "loss": 0.4332, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 2.2172107727937336, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.3862, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 2.0026334395442253, | |
| "learning_rate": 9.73963673083566e-06, | |
| "loss": 0.3375, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 2.008975524410557, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.3209, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 2.0020308436998615, | |
| "learning_rate": 9.654368743221022e-06, | |
| "loss": 0.3352, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 1.7951748955191769, | |
| "learning_rate": 9.60738105935204e-06, | |
| "loss": 0.324, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 1.9455517887105027, | |
| "learning_rate": 9.557529261558367e-06, | |
| "loss": 0.3292, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 1.9655040976499933, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.3404, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 2.257691645993752, | |
| "learning_rate": 9.449359044057344e-06, | |
| "loss": 0.3416, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 1.9178041719883492, | |
| "learning_rate": 9.391107866851143e-06, | |
| "loss": 0.3283, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.8376837225343723, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.3065, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 2.255376117326017, | |
| "learning_rate": 9.266454408160779e-06, | |
| "loss": 0.3898, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 2.3147995497368017, | |
| "learning_rate": 9.200129615753858e-06, | |
| "loss": 0.3566, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 1.9670946821467294, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.3397, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 1.9043148573056243, | |
| "learning_rate": 9.059690028579285e-06, | |
| "loss": 0.3146, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 2.0018857432631436, | |
| "learning_rate": 8.985662536114614e-06, | |
| "loss": 0.326, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 1.9805418049960362, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.3561, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 2.002055916467354, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.2896, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 1.9173208879793804, | |
| "learning_rate": 8.748906014838672e-06, | |
| "loss": 0.3481, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.7900189098270431, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.3038, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 1.6509929161585246, | |
| "learning_rate": 8.579334246298593e-06, | |
| "loss": 0.2826, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 1.671085946192089, | |
| "learning_rate": 8.491184090430365e-06, | |
| "loss": 0.2615, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 1.8602137091136728, | |
| "learning_rate": 8.400863688854598e-06, | |
| "loss": 0.2967, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 1.7196097127036853, | |
| "learning_rate": 8.308429187984298e-06, | |
| "loss": 0.2952, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 1.931814044902892, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.3372, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 2.1049189778926003, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.3404, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 1.8220307585532305, | |
| "learning_rate": 8.019022051627387e-06, | |
| "loss": 0.2965, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 2.1357017184024776, | |
| "learning_rate": 7.918718361173951e-06, | |
| "loss": 0.3163, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 1.9302715132373918, | |
| "learning_rate": 7.81660029031811e-06, | |
| "loss": 0.3218, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 1.6349542473923382, | |
| "learning_rate": 7.712731319328798e-06, | |
| "loss": 0.2527, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 2.300579309285893, | |
| "learning_rate": 7.607176016897491e-06, | |
| "loss": 0.3232, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 2.1075973294948263, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.367, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 1.8304036968165827, | |
| "learning_rate": 7.391269893106592e-06, | |
| "loss": 0.3094, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 1.9405966550411191, | |
| "learning_rate": 7.281053286765816e-06, | |
| "loss": 0.282, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.920794107387291, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.287, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 1.8746805079924518, | |
| "learning_rate": 7.056435515653059e-06, | |
| "loss": 0.3016, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 1.6795992682490004, | |
| "learning_rate": 6.942173981373474e-06, | |
| "loss": 0.2765, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.5560823579671665, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.2532, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 1.895658785400284, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.2892, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 1.8753314593622137, | |
| "learning_rate": 6.592433251258423e-06, | |
| "loss": 0.3301, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 1.829267141246674, | |
| "learning_rate": 6.473775872054522e-06, | |
| "loss": 0.2891, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 1.7420326737904375, | |
| "learning_rate": 6.354202340715027e-06, | |
| "loss": 0.2866, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 1.6023111714513605, | |
| "learning_rate": 6.233786988451468e-06, | |
| "loss": 0.2546, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 1.6958524220983555, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.263, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 1.641293642541498, | |
| "learning_rate": 5.990730715996989e-06, | |
| "loss": 0.2519, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.9737668057519353, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3114, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 1.6222206764245741, | |
| "learning_rate": 5.745211330880872e-06, | |
| "loss": 0.2611, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 1.6779148540502549, | |
| "learning_rate": 5.621718523237427e-06, | |
| "loss": 0.2749, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 1.5069533839875844, | |
| "learning_rate": 5.497839232979084e-06, | |
| "loss": 0.2395, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 1.8959358860602276, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.2893, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 1.523846974950363, | |
| "learning_rate": 5.249229428303486e-06, | |
| "loss": 0.2465, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 1.648360547715999, | |
| "learning_rate": 5.1246534586903655e-06, | |
| "loss": 0.2895, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7313601334897208, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2725, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 1.5974296941422694, | |
| "learning_rate": 4.875346541309637e-06, | |
| "loss": 0.2332, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 1.7754313088972902, | |
| "learning_rate": 4.750770571696514e-06, | |
| "loss": 0.2586, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 1.7376217331972978, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.2542, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 1.6766787084873536, | |
| "learning_rate": 4.502160767020918e-06, | |
| "loss": 0.2526, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 1.6040607155196303, | |
| "learning_rate": 4.3782814767625755e-06, | |
| "loss": 0.2305, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 1.76649027588737, | |
| "learning_rate": 4.254788669119127e-06, | |
| "loss": 0.2713, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 1.689340586990794, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2382, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 1.6502677192393955, | |
| "learning_rate": 4.009269284003014e-06, | |
| "loss": 0.226, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 1.6333958229898413, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.2289, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 1.7310975785907043, | |
| "learning_rate": 3.7662130115485317e-06, | |
| "loss": 0.2873, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 1.5468081102672981, | |
| "learning_rate": 3.6457976592849753e-06, | |
| "loss": 0.1974, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 1.5296817125485362, | |
| "learning_rate": 3.526224127945479e-06, | |
| "loss": 0.2013, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 1.6513883543662364, | |
| "learning_rate": 3.4075667487415785e-06, | |
| "loss": 0.2257, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 1.7657149068040885, | |
| "learning_rate": 3.289899283371657e-06, | |
| "loss": 0.2389, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 1.6542259947868383, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.2207, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 1.785500829184994, | |
| "learning_rate": 3.057826018626527e-06, | |
| "loss": 0.2409, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 1.624145624947224, | |
| "learning_rate": 2.9435644843469434e-06, | |
| "loss": 0.2361, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 1.552144673465774, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.2051, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 1.6259207954939752, | |
| "learning_rate": 2.718946713234185e-06, | |
| "loss": 0.2213, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 1.6311364982482741, | |
| "learning_rate": 2.608730106893411e-06, | |
| "loss": 0.2505, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.5946669632314558, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2077, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 1.6540736898757478, | |
| "learning_rate": 2.39282398310251e-06, | |
| "loss": 0.2231, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 1.5631800963608526, | |
| "learning_rate": 2.2872686806712037e-06, | |
| "loss": 0.2063, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 1.6790445903303324, | |
| "learning_rate": 2.1833997096818897e-06, | |
| "loss": 0.2275, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 1.71594970960584, | |
| "learning_rate": 2.081281638826052e-06, | |
| "loss": 0.2304, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 1.6467944353789714, | |
| "learning_rate": 1.980977948372612e-06, | |
| "loss": 0.2284, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.508481981490406, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.2133, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 1.6577136924947176, | |
| "learning_rate": 1.7860619515673034e-06, | |
| "loss": 0.2344, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 1.6358011219062036, | |
| "learning_rate": 1.6915708120157042e-06, | |
| "loss": 0.2326, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 1.5343887042100857, | |
| "learning_rate": 1.5991363111454023e-06, | |
| "loss": 0.2376, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 1.660583983348991, | |
| "learning_rate": 1.5088159095696365e-06, | |
| "loss": 0.225, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 1.6005130793524434, | |
| "learning_rate": 1.4206657537014078e-06, | |
| "loss": 0.2305, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 1.553147977120734, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.2288, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 1.5908104698304062, | |
| "learning_rate": 1.2510939851613285e-06, | |
| "loss": 0.2215, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 1.9504543471947708, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.278, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 1.704011286004108, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.2287, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 1.618234720661816, | |
| "learning_rate": 1.0143374638853892e-06, | |
| "loss": 0.2402, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 1.6475638575246008, | |
| "learning_rate": 9.403099714207175e-07, | |
| "loss": 0.2192, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 1.5958099241656356, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.2549, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 1.561819978797271, | |
| "learning_rate": 7.99870384246143e-07, | |
| "loss": 0.2111, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 1.7949770813372914, | |
| "learning_rate": 7.33545591839222e-07, | |
| "loss": 0.2376, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.698853452351741, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.2243, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 1.7776514869091975, | |
| "learning_rate": 6.088921331488568e-07, | |
| "loss": 0.2356, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 1.6125533182565601, | |
| "learning_rate": 5.506409559426573e-07, | |
| "loss": 0.2192, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 1.4580499888991223, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2011, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 1.44711042249179, | |
| "learning_rate": 4.4247073844163434e-07, | |
| "loss": 0.1919, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 1.5319692607578637, | |
| "learning_rate": 3.9261894064796136e-07, | |
| "loss": 0.2228, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 1.703198205141255, | |
| "learning_rate": 3.4563125677897936e-07, | |
| "loss": 0.2446, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 1.4382012783244584, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.2085, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 1.4938131300996766, | |
| "learning_rate": 2.6036326916434153e-07, | |
| "loss": 0.2043, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 1.5238283598819027, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.213, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 1.5482037294832478, | |
| "learning_rate": 1.8687876524993987e-07, | |
| "loss": 0.2136, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 1.480422561570781, | |
| "learning_rate": 1.5461356885461077e-07, | |
| "loss": 0.2242, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 1.4385375275853922, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.1934, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 1.398116868516386, | |
| "learning_rate": 9.913756075728088e-08, | |
| "loss": 0.1943, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 1.481535586423097, | |
| "learning_rate": 7.59612349389599e-08, | |
| "loss": 0.2096, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 1.474742617578431, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.207, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 1.3680015424187337, | |
| "learning_rate": 3.8803966999139686e-08, | |
| "loss": 0.1845, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 1.7090636442828981, | |
| "learning_rate": 2.4846123172992953e-08, | |
| "loss": 0.2299, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 1.4560355937269758, | |
| "learning_rate": 1.3981014094099354e-08, | |
| "loss": 0.1846, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 1.6521726152458496, | |
| "learning_rate": 6.215393905388278e-09, | |
| "loss": 0.2492, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 1.3715437105112576, | |
| "learning_rate": 1.5540899959187727e-09, | |
| "loss": 0.1846, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.411488207674534, | |
| "learning_rate": 0.0, | |
| "loss": 0.1976, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 126, | |
| "total_flos": 1772529647616.0, | |
| "train_loss": 0.2918981748913962, | |
| "train_runtime": 115.6265, | |
| "train_samples_per_second": 8.683, | |
| "train_steps_per_second": 1.09 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 126, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 70000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1772529647616.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |