| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9754098360655736, | |
| "eval_steps": 500, | |
| "global_step": 183, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 38.715505104327306, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 2.7935, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 43.92734901516536, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 2.8618, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 40.50339510154553, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.8164, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 28.51481486426967, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 2.672, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 17.557085161970114, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.526, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 9.833831469561684, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.4556, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 5.8460959136595845, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 2.1935, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 5.207230159450673, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 2.3815, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 7.281692741856846, | |
| "learning_rate": 6e-06, | |
| "loss": 2.311, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 4.085861410961766, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.3801, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 3.1648108898931473, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 2.2303, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 2.733124694129773, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.2489, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.21138211382113822, | |
| "grad_norm": 2.601048953766184, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 2.3216, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 2.799670635370881, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 2.3405, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 2.2889145851702035, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2761, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 2.074354640710373, | |
| "learning_rate": 9.999529497453782e-06, | |
| "loss": 2.2498, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2764227642276423, | |
| "grad_norm": 1.8496605844704992, | |
| "learning_rate": 9.998118078364186e-06, | |
| "loss": 2.1694, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 1.657414876254685, | |
| "learning_rate": 9.99576600836172e-06, | |
| "loss": 2.1922, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3089430894308943, | |
| "grad_norm": 1.6874201712764785, | |
| "learning_rate": 9.992473730108354e-06, | |
| "loss": 2.2212, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 1.6218362871051897, | |
| "learning_rate": 9.988241863214212e-06, | |
| "loss": 2.2361, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.34146341463414637, | |
| "grad_norm": 1.7125149417249366, | |
| "learning_rate": 9.98307120412095e-06, | |
| "loss": 2.0696, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 1.748991995811329, | |
| "learning_rate": 9.976962725951878e-06, | |
| "loss": 2.3047, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.37398373983739835, | |
| "grad_norm": 1.635597001502544, | |
| "learning_rate": 9.969917578328808e-06, | |
| "loss": 2.1672, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3902439024390244, | |
| "grad_norm": 1.5126187316968263, | |
| "learning_rate": 9.961937087155697e-06, | |
| "loss": 2.3152, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4065040650406504, | |
| "grad_norm": 1.5822580849972034, | |
| "learning_rate": 9.953022754369115e-06, | |
| "loss": 2.2176, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.42276422764227645, | |
| "grad_norm": 1.5777453977666092, | |
| "learning_rate": 9.943176257655567e-06, | |
| "loss": 2.3213, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.43902439024390244, | |
| "grad_norm": 1.4094336427074805, | |
| "learning_rate": 9.932399450135765e-06, | |
| "loss": 2.1472, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.45528455284552843, | |
| "grad_norm": 1.4386799640865944, | |
| "learning_rate": 9.920694360015864e-06, | |
| "loss": 2.2464, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4715447154471545, | |
| "grad_norm": 1.5398404589418597, | |
| "learning_rate": 9.908063190205739e-06, | |
| "loss": 2.2179, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 1.4686769996035434, | |
| "learning_rate": 9.894508317904418e-06, | |
| "loss": 2.1055, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5040650406504065, | |
| "grad_norm": 1.5185344171327764, | |
| "learning_rate": 9.880032294152673e-06, | |
| "loss": 2.4138, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5203252032520326, | |
| "grad_norm": 1.4788353106580088, | |
| "learning_rate": 9.864637843352916e-06, | |
| "loss": 2.1975, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5365853658536586, | |
| "grad_norm": 1.5482980039161824, | |
| "learning_rate": 9.848327862756466e-06, | |
| "loss": 2.2368, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5528455284552846, | |
| "grad_norm": 1.5529583533392486, | |
| "learning_rate": 9.831105421918287e-06, | |
| "loss": 2.244, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5691056910569106, | |
| "grad_norm": 1.2347439521124879, | |
| "learning_rate": 9.812973762119282e-06, | |
| "loss": 2.1969, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.7342145031832543, | |
| "learning_rate": 9.793936295756292e-06, | |
| "loss": 2.1326, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6016260162601627, | |
| "grad_norm": 1.6027037759927296, | |
| "learning_rate": 9.773996605699876e-06, | |
| "loss": 2.2234, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6178861788617886, | |
| "grad_norm": 1.5818719102225083, | |
| "learning_rate": 9.753158444620013e-06, | |
| "loss": 2.2304, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6341463414634146, | |
| "grad_norm": 1.5993004636102213, | |
| "learning_rate": 9.73142573427984e-06, | |
| "loss": 2.0879, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6504065040650406, | |
| "grad_norm": 1.4999201972757639, | |
| "learning_rate": 9.70880256479758e-06, | |
| "loss": 2.3389, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.5972052352215045, | |
| "learning_rate": 9.685293193876766e-06, | |
| "loss": 2.1184, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6829268292682927, | |
| "grad_norm": 1.5799871790968651, | |
| "learning_rate": 9.660902046004954e-06, | |
| "loss": 2.2381, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6991869918699187, | |
| "grad_norm": 1.7655935713426523, | |
| "learning_rate": 9.635633711621014e-06, | |
| "loss": 1.9785, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7154471544715447, | |
| "grad_norm": 1.4455569662489212, | |
| "learning_rate": 9.60949294625121e-06, | |
| "loss": 2.2598, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7317073170731707, | |
| "grad_norm": 1.6438734541523634, | |
| "learning_rate": 9.582484669614212e-06, | |
| "loss": 2.2254, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7479674796747967, | |
| "grad_norm": 1.5826415829816778, | |
| "learning_rate": 9.554613964695189e-06, | |
| "loss": 2.3195, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7642276422764228, | |
| "grad_norm": 1.5129794859111765, | |
| "learning_rate": 9.525886076789195e-06, | |
| "loss": 2.309, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7804878048780488, | |
| "grad_norm": 1.4425514088997948, | |
| "learning_rate": 9.496306412513989e-06, | |
| "loss": 2.1234, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7967479674796748, | |
| "grad_norm": 1.4940088273936272, | |
| "learning_rate": 9.465880538792519e-06, | |
| "loss": 2.2481, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8130081300813008, | |
| "grad_norm": 1.4327286196072573, | |
| "learning_rate": 9.434614181805203e-06, | |
| "loss": 2.18, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8292682926829268, | |
| "grad_norm": 1.431778831080833, | |
| "learning_rate": 9.402513225912273e-06, | |
| "loss": 2.1464, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8455284552845529, | |
| "grad_norm": 1.4448422926978557, | |
| "learning_rate": 9.369583712546322e-06, | |
| "loss": 2.1412, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8617886178861789, | |
| "grad_norm": 1.4679959607329234, | |
| "learning_rate": 9.335831839075303e-06, | |
| "loss": 2.0886, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8780487804878049, | |
| "grad_norm": 1.4297931147920537, | |
| "learning_rate": 9.30126395763618e-06, | |
| "loss": 2.2567, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8943089430894309, | |
| "grad_norm": 1.3468034919950198, | |
| "learning_rate": 9.265886573939448e-06, | |
| "loss": 2.2071, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9105691056910569, | |
| "grad_norm": 1.4936153334385893, | |
| "learning_rate": 9.229706346044749e-06, | |
| "loss": 2.2185, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.926829268292683, | |
| "grad_norm": 1.6789184395018877, | |
| "learning_rate": 9.19273008310782e-06, | |
| "loss": 2.2014, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.943089430894309, | |
| "grad_norm": 2.328661298153007, | |
| "learning_rate": 9.154964744099006e-06, | |
| "loss": 2.2311, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.959349593495935, | |
| "grad_norm": 1.6031494617421091, | |
| "learning_rate": 9.116417436493574e-06, | |
| "loss": 2.2817, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 1.378022632096063, | |
| "learning_rate": 9.077095414934076e-06, | |
| "loss": 2.3173, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.991869918699187, | |
| "grad_norm": 1.378014833579439, | |
| "learning_rate": 9.037006079865017e-06, | |
| "loss": 2.1865, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.008130081300813, | |
| "grad_norm": 1.777721048156252, | |
| "learning_rate": 8.996156976140088e-06, | |
| "loss": 1.9477, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.024390243902439, | |
| "grad_norm": 1.9572083586736355, | |
| "learning_rate": 8.95455579160221e-06, | |
| "loss": 1.9606, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.040650406504065, | |
| "grad_norm": 1.83840626983557, | |
| "learning_rate": 8.91221035563669e-06, | |
| "loss": 1.9641, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.056910569105691, | |
| "grad_norm": 1.7482663924396353, | |
| "learning_rate": 8.869128637697702e-06, | |
| "loss": 1.912, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0731707317073171, | |
| "grad_norm": 2.436293071917533, | |
| "learning_rate": 8.82531874580844e-06, | |
| "loss": 2.0053, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.089430894308943, | |
| "grad_norm": 1.7774996740932676, | |
| "learning_rate": 8.780788925035178e-06, | |
| "loss": 1.8775, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1056910569105691, | |
| "grad_norm": 1.5469357572162499, | |
| "learning_rate": 8.735547555935538e-06, | |
| "loss": 1.8116, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.1219512195121952, | |
| "grad_norm": 1.9514287859645285, | |
| "learning_rate": 8.689603152981262e-06, | |
| "loss": 1.9458, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1382113821138211, | |
| "grad_norm": 2.16053582553305, | |
| "learning_rate": 8.642964362955781e-06, | |
| "loss": 1.8903, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1544715447154472, | |
| "grad_norm": 1.8295345259876294, | |
| "learning_rate": 8.59563996332688e-06, | |
| "loss": 1.9097, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.170731707317073, | |
| "grad_norm": 1.6192771141787508, | |
| "learning_rate": 8.547638860594765e-06, | |
| "loss": 1.9404, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1869918699186992, | |
| "grad_norm": 1.732769202153044, | |
| "learning_rate": 8.498970088615861e-06, | |
| "loss": 1.9072, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.203252032520325, | |
| "grad_norm": 1.7124937892640733, | |
| "learning_rate": 8.449642806902623e-06, | |
| "loss": 1.9361, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2195121951219512, | |
| "grad_norm": 1.4632968865855904, | |
| "learning_rate": 8.399666298899706e-06, | |
| "loss": 1.9064, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2357723577235773, | |
| "grad_norm": 1.57738250567877, | |
| "learning_rate": 8.349049970236822e-06, | |
| "loss": 1.8884, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2520325203252032, | |
| "grad_norm": 1.6860131175954445, | |
| "learning_rate": 8.29780334695857e-06, | |
| "loss": 1.8003, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2682926829268293, | |
| "grad_norm": 1.6737455231202842, | |
| "learning_rate": 8.245936073731654e-06, | |
| "loss": 1.9188, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2845528455284554, | |
| "grad_norm": 1.5244480150455564, | |
| "learning_rate": 8.193457912029713e-06, | |
| "loss": 1.9428, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.3008130081300813, | |
| "grad_norm": 2.3077062591013546, | |
| "learning_rate": 8.140378738296233e-06, | |
| "loss": 1.9652, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3170731707317074, | |
| "grad_norm": 2.0246897260170433, | |
| "learning_rate": 8.086708542085769e-06, | |
| "loss": 1.8873, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.6727268288560573, | |
| "learning_rate": 8.032457424183909e-06, | |
| "loss": 1.8492, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3495934959349594, | |
| "grad_norm": 1.5569674481989706, | |
| "learning_rate": 7.977635594706298e-06, | |
| "loss": 1.9254, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3658536585365852, | |
| "grad_norm": 1.6023097703448475, | |
| "learning_rate": 7.922253371177081e-06, | |
| "loss": 1.9188, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3821138211382114, | |
| "grad_norm": 1.579712800121583, | |
| "learning_rate": 7.866321176587129e-06, | |
| "loss": 1.7658, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3983739837398375, | |
| "grad_norm": 1.5575689401980928, | |
| "learning_rate": 7.809849537432432e-06, | |
| "loss": 1.8844, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.4146341463414633, | |
| "grad_norm": 2.0996103231960186, | |
| "learning_rate": 7.752849081732993e-06, | |
| "loss": 1.8144, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4308943089430894, | |
| "grad_norm": 2.256462070639042, | |
| "learning_rate": 7.695330537032629e-06, | |
| "loss": 1.9585, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.4471544715447155, | |
| "grad_norm": 1.5391039524045877, | |
| "learning_rate": 7.637304728380036e-06, | |
| "loss": 1.7978, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4634146341463414, | |
| "grad_norm": 1.6288407781749783, | |
| "learning_rate": 7.578782576291501e-06, | |
| "loss": 1.8906, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4796747967479675, | |
| "grad_norm": 1.4226382530812676, | |
| "learning_rate": 7.51977509469565e-06, | |
| "loss": 1.8687, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4959349593495934, | |
| "grad_norm": 1.5704917583247462, | |
| "learning_rate": 7.460293388860616e-06, | |
| "loss": 2.0512, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.5121951219512195, | |
| "grad_norm": 1.4787223094009658, | |
| "learning_rate": 7.400348653304022e-06, | |
| "loss": 1.9736, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.5284552845528454, | |
| "grad_norm": 1.4275720991608627, | |
| "learning_rate": 7.3399521696861505e-06, | |
| "loss": 1.9977, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.5447154471544715, | |
| "grad_norm": 1.4098433082701363, | |
| "learning_rate": 7.2791153046867344e-06, | |
| "loss": 1.8225, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5609756097560976, | |
| "grad_norm": 4.67217439880156, | |
| "learning_rate": 7.217849507865724e-06, | |
| "loss": 1.9615, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5772357723577235, | |
| "grad_norm": 1.519886113544732, | |
| "learning_rate": 7.156166309508482e-06, | |
| "loss": 1.8566, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5934959349593496, | |
| "grad_norm": 1.4312222151537652, | |
| "learning_rate": 7.094077318455762e-06, | |
| "loss": 1.8472, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.6097560975609757, | |
| "grad_norm": 1.595194011963869, | |
| "learning_rate": 7.031594219918916e-06, | |
| "loss": 1.8246, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.6260162601626016, | |
| "grad_norm": 1.5531618087372638, | |
| "learning_rate": 6.96872877328073e-06, | |
| "loss": 1.9809, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6422764227642277, | |
| "grad_norm": 1.4743001623349186, | |
| "learning_rate": 6.905492809882286e-06, | |
| "loss": 2.0107, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6585365853658538, | |
| "grad_norm": 1.9754475724007412, | |
| "learning_rate": 6.841898230796302e-06, | |
| "loss": 1.8143, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6747967479674797, | |
| "grad_norm": 2.2741340273630635, | |
| "learning_rate": 6.777957004587332e-06, | |
| "loss": 1.7824, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6910569105691056, | |
| "grad_norm": 1.6564200369629691, | |
| "learning_rate": 6.713681165059271e-06, | |
| "loss": 1.8844, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.7073170731707317, | |
| "grad_norm": 1.5809400530395674, | |
| "learning_rate": 6.6490828089905854e-06, | |
| "loss": 1.8789, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.7235772357723578, | |
| "grad_norm": 1.4746817698203465, | |
| "learning_rate": 6.584174093857676e-06, | |
| "loss": 1.9045, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.7398373983739837, | |
| "grad_norm": 1.353250857647505, | |
| "learning_rate": 6.5189672355468415e-06, | |
| "loss": 1.8118, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7560975609756098, | |
| "grad_norm": 1.4850655925014062, | |
| "learning_rate": 6.453474506055228e-06, | |
| "loss": 1.8122, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7723577235772359, | |
| "grad_norm": 1.8401558539962772, | |
| "learning_rate": 6.387708231181229e-06, | |
| "loss": 1.7482, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7886178861788617, | |
| "grad_norm": 1.5200607844016307, | |
| "learning_rate": 6.3216807882047585e-06, | |
| "loss": 1.9692, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8048780487804879, | |
| "grad_norm": 1.5040703158854816, | |
| "learning_rate": 6.255404603557833e-06, | |
| "loss": 1.8885, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.821138211382114, | |
| "grad_norm": 1.5595167294836927, | |
| "learning_rate": 6.188892150485904e-06, | |
| "loss": 1.8763, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.8373983739837398, | |
| "grad_norm": 1.5145771953515301, | |
| "learning_rate": 6.122155946700381e-06, | |
| "loss": 2.0202, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8536585365853657, | |
| "grad_norm": 1.3682270101469567, | |
| "learning_rate": 6.0552085520227875e-06, | |
| "loss": 1.9047, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8699186991869918, | |
| "grad_norm": 1.6370392819430508, | |
| "learning_rate": 5.988062566020987e-06, | |
| "loss": 1.9071, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.886178861788618, | |
| "grad_norm": 1.5064391288691565, | |
| "learning_rate": 5.920730625637934e-06, | |
| "loss": 1.9622, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.9024390243902438, | |
| "grad_norm": 1.5117346102674705, | |
| "learning_rate": 5.853225402813381e-06, | |
| "loss": 1.8889, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.91869918699187, | |
| "grad_norm": 1.5330246513699148, | |
| "learning_rate": 5.785559602099019e-06, | |
| "loss": 1.8971, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.934959349593496, | |
| "grad_norm": 1.446396984889757, | |
| "learning_rate": 5.7177459582674595e-06, | |
| "loss": 1.8328, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.951219512195122, | |
| "grad_norm": 1.495856639357874, | |
| "learning_rate": 5.649797233915539e-06, | |
| "loss": 1.8684, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.967479674796748, | |
| "grad_norm": 1.4499233339195112, | |
| "learning_rate": 5.5817262170623865e-06, | |
| "loss": 1.8167, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9837398373983741, | |
| "grad_norm": 2.7226474462829744, | |
| "learning_rate": 5.513545718742702e-06, | |
| "loss": 1.9347, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6284827727818851, | |
| "learning_rate": 5.4452685705957084e-06, | |
| "loss": 1.9544, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.0081967213114753, | |
| "grad_norm": 2.196083883493474, | |
| "learning_rate": 5.376907622450229e-06, | |
| "loss": 1.726, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.0245901639344264, | |
| "grad_norm": 2.5371696562778325, | |
| "learning_rate": 5.308475739906329e-06, | |
| "loss": 1.6243, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.040983606557377, | |
| "grad_norm": 2.268797600455164, | |
| "learning_rate": 5.2399858019140005e-06, | |
| "loss": 1.6506, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.057377049180328, | |
| "grad_norm": 2.189268946289623, | |
| "learning_rate": 5.171450698349329e-06, | |
| "loss": 1.5677, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.0737704918032787, | |
| "grad_norm": 3.330757242919671, | |
| "learning_rate": 5.102883327588608e-06, | |
| "loss": 1.5414, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.0901639344262297, | |
| "grad_norm": 4.975856342017294, | |
| "learning_rate": 5.034296594080849e-06, | |
| "loss": 1.5934, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.1065573770491803, | |
| "grad_norm": 2.6202824821832644, | |
| "learning_rate": 4.965703405919154e-06, | |
| "loss": 1.5114, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.122950819672131, | |
| "grad_norm": 2.145467593529425, | |
| "learning_rate": 4.897116672411395e-06, | |
| "loss": 1.6081, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.139344262295082, | |
| "grad_norm": 2.2918703289964597, | |
| "learning_rate": 4.828549301650673e-06, | |
| "loss": 1.6128, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.1557377049180326, | |
| "grad_norm": 2.2020919561070307, | |
| "learning_rate": 4.760014198086001e-06, | |
| "loss": 1.5385, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.1721311475409837, | |
| "grad_norm": 1.9897624139230765, | |
| "learning_rate": 4.691524260093672e-06, | |
| "loss": 1.5731, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1885245901639343, | |
| "grad_norm": 2.1542550052135665, | |
| "learning_rate": 4.623092377549772e-06, | |
| "loss": 1.671, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.2049180327868854, | |
| "grad_norm": 2.2842158477910512, | |
| "learning_rate": 4.554731429404293e-06, | |
| "loss": 1.614, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.221311475409836, | |
| "grad_norm": 2.1239054707901452, | |
| "learning_rate": 4.4864542812573e-06, | |
| "loss": 1.6136, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.237704918032787, | |
| "grad_norm": 2.0393208350755008, | |
| "learning_rate": 4.4182737829376135e-06, | |
| "loss": 1.6467, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.2540983606557377, | |
| "grad_norm": 2.0314018092300103, | |
| "learning_rate": 4.3502027660844606e-06, | |
| "loss": 1.6771, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.2704918032786887, | |
| "grad_norm": 2.258947100849125, | |
| "learning_rate": 4.28225404173254e-06, | |
| "loss": 1.5915, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.2868852459016393, | |
| "grad_norm": 2.1050457746344313, | |
| "learning_rate": 4.214440397900983e-06, | |
| "loss": 1.6608, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.30327868852459, | |
| "grad_norm": 1.948114104146267, | |
| "learning_rate": 4.146774597186622e-06, | |
| "loss": 1.5369, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.319672131147541, | |
| "grad_norm": 1.864917840283612, | |
| "learning_rate": 4.0792693743620695e-06, | |
| "loss": 1.5717, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.3360655737704916, | |
| "grad_norm": 1.7461920012860392, | |
| "learning_rate": 4.011937433979014e-06, | |
| "loss": 1.5871, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.3524590163934427, | |
| "grad_norm": 1.798186392967839, | |
| "learning_rate": 3.944791447977213e-06, | |
| "loss": 1.5592, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.3688524590163933, | |
| "grad_norm": 1.840314539800089, | |
| "learning_rate": 3.87784405329962e-06, | |
| "loss": 1.5807, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.3852459016393444, | |
| "grad_norm": 1.9329074085202937, | |
| "learning_rate": 3.811107849514098e-06, | |
| "loss": 1.5505, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.401639344262295, | |
| "grad_norm": 1.652283737337734, | |
| "learning_rate": 3.744595396442169e-06, | |
| "loss": 1.5601, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.418032786885246, | |
| "grad_norm": 1.6942011471416905, | |
| "learning_rate": 3.6783192117952427e-06, | |
| "loss": 1.5762, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.4344262295081966, | |
| "grad_norm": 1.6999827546312305, | |
| "learning_rate": 3.612291768818772e-06, | |
| "loss": 1.5286, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4508196721311477, | |
| "grad_norm": 1.717480901806302, | |
| "learning_rate": 3.5465254939447737e-06, | |
| "loss": 1.6165, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.4672131147540983, | |
| "grad_norm": 1.7354378842697238, | |
| "learning_rate": 3.4810327644531606e-06, | |
| "loss": 1.5636, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.4836065573770494, | |
| "grad_norm": 2.076605836354899, | |
| "learning_rate": 3.415825906142326e-06, | |
| "loss": 1.5426, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.807787441701596, | |
| "learning_rate": 3.3509171910094162e-06, | |
| "loss": 1.4229, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.5163934426229506, | |
| "grad_norm": 2.1162841013271, | |
| "learning_rate": 3.2863188349407293e-06, | |
| "loss": 1.5828, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.5327868852459017, | |
| "grad_norm": 2.06404980427501, | |
| "learning_rate": 3.222042995412669e-06, | |
| "loss": 1.6511, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.5491803278688527, | |
| "grad_norm": 1.954408720302879, | |
| "learning_rate": 3.1581017692036986e-06, | |
| "loss": 1.5611, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.5655737704918034, | |
| "grad_norm": 1.8718671260321513, | |
| "learning_rate": 3.094507190117715e-06, | |
| "loss": 1.5528, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.581967213114754, | |
| "grad_norm": 1.8323058517821103, | |
| "learning_rate": 3.0312712267192713e-06, | |
| "loss": 1.5525, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.598360655737705, | |
| "grad_norm": 4.8860562051247385, | |
| "learning_rate": 2.9684057800810844e-06, | |
| "loss": 1.5571, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6147540983606556, | |
| "grad_norm": 2.1738936906241237, | |
| "learning_rate": 2.9059226815442386e-06, | |
| "loss": 1.4133, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.6311475409836067, | |
| "grad_norm": 2.1707521374782246, | |
| "learning_rate": 2.8438336904915186e-06, | |
| "loss": 1.7308, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.6475409836065573, | |
| "grad_norm": 1.848892940417498, | |
| "learning_rate": 2.782150492134278e-06, | |
| "loss": 1.6001, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.663934426229508, | |
| "grad_norm": 1.7230291366018273, | |
| "learning_rate": 2.7208846953132685e-06, | |
| "loss": 1.6081, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.680327868852459, | |
| "grad_norm": 1.6901177596723493, | |
| "learning_rate": 2.6600478303138503e-06, | |
| "loss": 1.4536, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.69672131147541, | |
| "grad_norm": 1.770734235373982, | |
| "learning_rate": 2.599651346695979e-06, | |
| "loss": 1.5516, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.7131147540983607, | |
| "grad_norm": 1.9713299716871673, | |
| "learning_rate": 2.539706611139385e-06, | |
| "loss": 1.6036, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.7295081967213113, | |
| "grad_norm": 1.944020272636568, | |
| "learning_rate": 2.4802249053043525e-06, | |
| "loss": 1.6063, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.7459016393442623, | |
| "grad_norm": 1.917025051456852, | |
| "learning_rate": 2.4212174237085007e-06, | |
| "loss": 1.484, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.762295081967213, | |
| "grad_norm": 1.7078308287195758, | |
| "learning_rate": 2.3626952716199647e-06, | |
| "loss": 1.6206, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.778688524590164, | |
| "grad_norm": 2.105784906420741, | |
| "learning_rate": 2.3046694629673715e-06, | |
| "loss": 1.5728, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.7950819672131146, | |
| "grad_norm": 1.8395129946614983, | |
| "learning_rate": 2.247150918267008e-06, | |
| "loss": 1.5713, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.8114754098360657, | |
| "grad_norm": 1.6207009651182827, | |
| "learning_rate": 2.190150462567569e-06, | |
| "loss": 1.5185, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.8278688524590163, | |
| "grad_norm": 1.5443883689967923, | |
| "learning_rate": 2.133678823412873e-06, | |
| "loss": 1.5119, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.8442622950819674, | |
| "grad_norm": 1.6761100058671115, | |
| "learning_rate": 2.077746628822921e-06, | |
| "loss": 1.5463, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.860655737704918, | |
| "grad_norm": 1.982740476512917, | |
| "learning_rate": 2.022364405293703e-06, | |
| "loss": 1.5463, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.8770491803278686, | |
| "grad_norm": 1.792200717708596, | |
| "learning_rate": 1.9675425758160927e-06, | |
| "loss": 1.6641, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.8934426229508197, | |
| "grad_norm": 2.129611048939019, | |
| "learning_rate": 1.913291457914234e-06, | |
| "loss": 1.6427, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.9098360655737707, | |
| "grad_norm": 1.703026911107873, | |
| "learning_rate": 1.8596212617037695e-06, | |
| "loss": 1.5071, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.9262295081967213, | |
| "grad_norm": 1.684347363697613, | |
| "learning_rate": 1.8065420879702888e-06, | |
| "loss": 1.613, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.942622950819672, | |
| "grad_norm": 4.094268965277064, | |
| "learning_rate": 1.754063926268349e-06, | |
| "loss": 1.562, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.959016393442623, | |
| "grad_norm": 1.8320884272338482, | |
| "learning_rate": 1.7021966530414303e-06, | |
| "loss": 1.5323, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.9754098360655736, | |
| "grad_norm": 2.1968448353373833, | |
| "learning_rate": 1.6509500297631786e-06, | |
| "loss": 1.6683, | |
| "step": 183 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 244, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 61, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.784065850068173e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |