| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.6768970424214453, |
| "eval_steps": 500, |
| "global_step": 1100, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.006153609476558594, |
| "grad_norm": 25468.8046875, |
| "learning_rate": 1.9999999999999996e-07, |
| "loss": 8.7082, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.012307218953117188, |
| "grad_norm": 1228.4501953125, |
| "learning_rate": 1.2e-06, |
| "loss": 8.0776, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.018460828429675783, |
| "grad_norm": 355.1081237792969, |
| "learning_rate": 2.1999999999999997e-06, |
| "loss": 7.2372, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.024614437906234375, |
| "grad_norm": 255.58433532714844, |
| "learning_rate": 3.1999999999999994e-06, |
| "loss": 6.9725, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.03076804738279297, |
| "grad_norm": 253.23995971679688, |
| "learning_rate": 4.2e-06, |
| "loss": 6.8585, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.036921656859351566, |
| "grad_norm": 241.82064819335938, |
| "learning_rate": 5.199999999999999e-06, |
| "loss": 6.7367, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.043075266335910155, |
| "grad_norm": 229.0261993408203, |
| "learning_rate": 6.199999999999999e-06, |
| "loss": 6.6534, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.04922887581246875, |
| "grad_norm": 321.6871337890625, |
| "learning_rate": 7.2e-06, |
| "loss": 6.6305, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.055382485289027346, |
| "grad_norm": 465.00543212890625, |
| "learning_rate": 8.2e-06, |
| "loss": 6.6253, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.06153609476558594, |
| "grad_norm": 566.6929321289062, |
| "learning_rate": 9.199999999999998e-06, |
| "loss": 6.5724, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.06768970424214453, |
| "grad_norm": 432.230712890625, |
| "learning_rate": 1.02e-05, |
| "loss": 6.5328, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.07384331371870313, |
| "grad_norm": 289.2374267578125, |
| "learning_rate": 1.12e-05, |
| "loss": 6.5547, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.07999692319526172, |
| "grad_norm": 595.7386474609375, |
| "learning_rate": 1.2199999999999998e-05, |
| "loss": 6.5651, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.08615053267182031, |
| "grad_norm": 861.4376220703125, |
| "learning_rate": 1.3199999999999997e-05, |
| "loss": 6.5337, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.09230414214837891, |
| "grad_norm": 689.42919921875, |
| "learning_rate": 1.4199999999999998e-05, |
| "loss": 6.4927, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.0984577516249375, |
| "grad_norm": 380.9496765136719, |
| "learning_rate": 1.5199999999999998e-05, |
| "loss": 6.4812, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.1046113611014961, |
| "grad_norm": 221.21005249023438, |
| "learning_rate": 1.6199999999999997e-05, |
| "loss": 6.4508, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.11076497057805469, |
| "grad_norm": 199.30874633789062, |
| "learning_rate": 1.7199999999999998e-05, |
| "loss": 6.4642, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.11691858005461328, |
| "grad_norm": 252.44459533691406, |
| "learning_rate": 1.82e-05, |
| "loss": 6.45, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.12307218953117188, |
| "grad_norm": 121.32320404052734, |
| "learning_rate": 1.92e-05, |
| "loss": 6.4115, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.12922579900773049, |
| "grad_norm": 207.04344177246094, |
| "learning_rate": 2.0199999999999996e-05, |
| "loss": 6.4532, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.13537940848428906, |
| "grad_norm": 296.6274719238281, |
| "learning_rate": 2.12e-05, |
| "loss": 6.3985, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.14153301796084766, |
| "grad_norm": 140.3836212158203, |
| "learning_rate": 2.2199999999999998e-05, |
| "loss": 6.4152, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.14768662743740626, |
| "grad_norm": 168.14443969726562, |
| "learning_rate": 2.3199999999999998e-05, |
| "loss": 6.3686, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.15384023691396484, |
| "grad_norm": 141.07810974121094, |
| "learning_rate": 2.42e-05, |
| "loss": 6.3809, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.15999384639052344, |
| "grad_norm": 117.08052062988281, |
| "learning_rate": 2.52e-05, |
| "loss": 6.3423, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.16614745586708204, |
| "grad_norm": 122.97789764404297, |
| "learning_rate": 2.6199999999999996e-05, |
| "loss": 6.3486, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.17230106534364062, |
| "grad_norm": 73.85701751708984, |
| "learning_rate": 2.72e-05, |
| "loss": 6.329, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.17845467482019922, |
| "grad_norm": 78.00408935546875, |
| "learning_rate": 2.8199999999999998e-05, |
| "loss": 6.3096, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.18460828429675782, |
| "grad_norm": 150.3341827392578, |
| "learning_rate": 2.9199999999999995e-05, |
| "loss": 6.3063, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.1907618937733164, |
| "grad_norm": 61.75702667236328, |
| "learning_rate": 3.02e-05, |
| "loss": 6.3004, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.196915503249875, |
| "grad_norm": 83.836181640625, |
| "learning_rate": 3.119999999999999e-05, |
| "loss": 6.2534, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.2030691127264336, |
| "grad_norm": 190.2839813232422, |
| "learning_rate": 3.22e-05, |
| "loss": 6.2603, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.2092227222029922, |
| "grad_norm": 52.196563720703125, |
| "learning_rate": 3.3199999999999994e-05, |
| "loss": 6.2439, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.21537633167955078, |
| "grad_norm": 65.71736907958984, |
| "learning_rate": 3.42e-05, |
| "loss": 6.227, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.22152994115610938, |
| "grad_norm": 66.71468353271484, |
| "learning_rate": 3.5199999999999995e-05, |
| "loss": 6.2164, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.22768355063266799, |
| "grad_norm": 37.26243209838867, |
| "learning_rate": 3.62e-05, |
| "loss": 6.212, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.23383716010922656, |
| "grad_norm": 190.98501586914062, |
| "learning_rate": 3.7199999999999996e-05, |
| "loss": 6.1833, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.23999076958578516, |
| "grad_norm": 51.51064682006836, |
| "learning_rate": 3.8199999999999993e-05, |
| "loss": 6.1796, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.24614437906234377, |
| "grad_norm": 71.14685821533203, |
| "learning_rate": 3.919999999999999e-05, |
| "loss": 6.1614, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.25229798853890234, |
| "grad_norm": 47.48910140991211, |
| "learning_rate": 4.02e-05, |
| "loss": 6.1277, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.25845159801546097, |
| "grad_norm": 49.60618209838867, |
| "learning_rate": 4.12e-05, |
| "loss": 6.107, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.26460520749201955, |
| "grad_norm": 38.03694534301758, |
| "learning_rate": 4.2199999999999996e-05, |
| "loss": 6.0939, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.2707588169685781, |
| "grad_norm": 38.81101989746094, |
| "learning_rate": 4.319999999999999e-05, |
| "loss": 6.118, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.27691242644513675, |
| "grad_norm": 45.36625671386719, |
| "learning_rate": 4.42e-05, |
| "loss": 6.0641, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.2830660359216953, |
| "grad_norm": 44.09037780761719, |
| "learning_rate": 4.5199999999999994e-05, |
| "loss": 6.0417, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.2892196453982539, |
| "grad_norm": 39.48929214477539, |
| "learning_rate": 4.62e-05, |
| "loss": 6.0371, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.29537325487481253, |
| "grad_norm": 44.747039794921875, |
| "learning_rate": 4.7199999999999995e-05, |
| "loss": 6.0157, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.3015268643513711, |
| "grad_norm": 51.372249603271484, |
| "learning_rate": 4.82e-05, |
| "loss": 6.0378, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.3076804738279297, |
| "grad_norm": 46.31863021850586, |
| "learning_rate": 4.9199999999999997e-05, |
| "loss": 5.9815, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.3138340833044883, |
| "grad_norm": 74.35610961914062, |
| "learning_rate": 5.0199999999999994e-05, |
| "loss": 6.0108, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.3199876927810469, |
| "grad_norm": 31.986116409301758, |
| "learning_rate": 5.119999999999999e-05, |
| "loss": 5.9452, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.32614130225760546, |
| "grad_norm": 62.82748794555664, |
| "learning_rate": 5.2199999999999995e-05, |
| "loss": 5.9396, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.3322949117341641, |
| "grad_norm": 36.623085021972656, |
| "learning_rate": 5.32e-05, |
| "loss": 5.9078, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.33844852121072266, |
| "grad_norm": 66.77005004882812, |
| "learning_rate": 5.4199999999999996e-05, |
| "loss": 5.9127, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.34460213068728124, |
| "grad_norm": 42.2696533203125, |
| "learning_rate": 5.519999999999999e-05, |
| "loss": 5.8751, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.35075574016383987, |
| "grad_norm": 36.03275680541992, |
| "learning_rate": 5.619999999999999e-05, |
| "loss": 5.8704, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.35690934964039844, |
| "grad_norm": 34.99599838256836, |
| "learning_rate": 5.72e-05, |
| "loss": 5.8357, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.363062959116957, |
| "grad_norm": 37.03520584106445, |
| "learning_rate": 5.82e-05, |
| "loss": 5.8248, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.36921656859351565, |
| "grad_norm": 39.70176696777344, |
| "learning_rate": 5.9199999999999996e-05, |
| "loss": 5.782, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.3753701780700742, |
| "grad_norm": 39.41380310058594, |
| "learning_rate": 6.019999999999999e-05, |
| "loss": 5.7875, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.3815237875466328, |
| "grad_norm": 36.758148193359375, |
| "learning_rate": 6.12e-05, |
| "loss": 5.7705, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.38767739702319143, |
| "grad_norm": 47.42232894897461, |
| "learning_rate": 6.22e-05, |
| "loss": 5.7891, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.39383100649975, |
| "grad_norm": 41.998783111572266, |
| "learning_rate": 6.319999999999999e-05, |
| "loss": 5.735, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.3999846159763086, |
| "grad_norm": 35.05141830444336, |
| "learning_rate": 6.419999999999999e-05, |
| "loss": 5.7239, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.4061382254528672, |
| "grad_norm": 38.990840911865234, |
| "learning_rate": 6.519999999999999e-05, |
| "loss": 5.7106, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.4122918349294258, |
| "grad_norm": 38.231101989746094, |
| "learning_rate": 6.62e-05, |
| "loss": 5.6812, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.4184454444059844, |
| "grad_norm": 35.416473388671875, |
| "learning_rate": 6.72e-05, |
| "loss": 5.6728, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.424599053882543, |
| "grad_norm": 45.61351776123047, |
| "learning_rate": 6.819999999999999e-05, |
| "loss": 5.6484, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.43075266335910156, |
| "grad_norm": 41.97138595581055, |
| "learning_rate": 6.919999999999999e-05, |
| "loss": 5.6279, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.4369062728356602, |
| "grad_norm": 35.231204986572266, |
| "learning_rate": 7.02e-05, |
| "loss": 5.6131, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.44305988231221877, |
| "grad_norm": 43.76620864868164, |
| "learning_rate": 7.12e-05, |
| "loss": 5.576, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.44921349178877734, |
| "grad_norm": 34.20133972167969, |
| "learning_rate": 7.219999999999999e-05, |
| "loss": 5.5689, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.45536710126533597, |
| "grad_norm": 37.153480529785156, |
| "learning_rate": 7.319999999999999e-05, |
| "loss": 5.5499, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.46152071074189455, |
| "grad_norm": 55.655784606933594, |
| "learning_rate": 7.419999999999999e-05, |
| "loss": 5.5353, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.4676743202184531, |
| "grad_norm": 33.66733169555664, |
| "learning_rate": 7.519999999999998e-05, |
| "loss": 5.5018, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.47382792969501175, |
| "grad_norm": 33.522857666015625, |
| "learning_rate": 7.62e-05, |
| "loss": 5.483, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.4799815391715703, |
| "grad_norm": 35.31947326660156, |
| "learning_rate": 7.72e-05, |
| "loss": 5.4719, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.4861351486481289, |
| "grad_norm": 34.84073257446289, |
| "learning_rate": 7.819999999999999e-05, |
| "loss": 5.4676, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.49228875812468753, |
| "grad_norm": 31.39910316467285, |
| "learning_rate": 7.92e-05, |
| "loss": 5.4421, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.4984423676012461, |
| "grad_norm": 33.14794158935547, |
| "learning_rate": 8.019999999999998e-05, |
| "loss": 5.3994, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.5045959770778047, |
| "grad_norm": 36.15346145629883, |
| "learning_rate": 8.12e-05, |
| "loss": 5.3823, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.5107495865543633, |
| "grad_norm": 32.183895111083984, |
| "learning_rate": 8.22e-05, |
| "loss": 5.3428, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.5169031960309219, |
| "grad_norm": 32.388916015625, |
| "learning_rate": 8.319999999999999e-05, |
| "loss": 5.3514, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.5230568055074805, |
| "grad_norm": 31.176891326904297, |
| "learning_rate": 8.42e-05, |
| "loss": 5.3256, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.5292104149840391, |
| "grad_norm": 33.613014221191406, |
| "learning_rate": 8.519999999999998e-05, |
| "loss": 5.3117, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.5353640244605977, |
| "grad_norm": 30.59576416015625, |
| "learning_rate": 8.62e-05, |
| "loss": 5.2776, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.5415176339371562, |
| "grad_norm": 34.864505767822266, |
| "learning_rate": 8.719999999999999e-05, |
| "loss": 5.2651, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.5476712434137149, |
| "grad_norm": 33.03851318359375, |
| "learning_rate": 8.819999999999999e-05, |
| "loss": 5.2467, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.5538248528902735, |
| "grad_norm": 36.94324493408203, |
| "learning_rate": 8.92e-05, |
| "loss": 5.2287, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.559978462366832, |
| "grad_norm": 32.03374481201172, |
| "learning_rate": 9.02e-05, |
| "loss": 5.197, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.5661320718433906, |
| "grad_norm": 31.21794891357422, |
| "learning_rate": 9.12e-05, |
| "loss": 5.1565, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.5722856813199493, |
| "grad_norm": 29.664560317993164, |
| "learning_rate": 9.219999999999999e-05, |
| "loss": 5.1325, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.5784392907965078, |
| "grad_norm": 29.547014236450195, |
| "learning_rate": 9.319999999999999e-05, |
| "loss": 5.1322, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.5845929002730664, |
| "grad_norm": 30.9229679107666, |
| "learning_rate": 9.419999999999999e-05, |
| "loss": 5.1176, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.5907465097496251, |
| "grad_norm": 32.148712158203125, |
| "learning_rate": 9.52e-05, |
| "loss": 5.0631, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.5969001192261836, |
| "grad_norm": 30.888355255126953, |
| "learning_rate": 9.62e-05, |
| "loss": 5.0937, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.6030537287027422, |
| "grad_norm": 37.31905746459961, |
| "learning_rate": 9.719999999999999e-05, |
| "loss": 5.0572, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.6092073381793008, |
| "grad_norm": 27.809349060058594, |
| "learning_rate": 9.819999999999999e-05, |
| "loss": 5.0144, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.6153609476558594, |
| "grad_norm": 34.41085433959961, |
| "learning_rate": 9.919999999999999e-05, |
| "loss": 5.0256, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.621514557132418, |
| "grad_norm": 29.413028717041016, |
| "learning_rate": 0.0001002, |
| "loss": 4.9648, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.6276681666089766, |
| "grad_norm": 30.73712158203125, |
| "learning_rate": 0.0001012, |
| "loss": 4.9744, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.6338217760855351, |
| "grad_norm": 30.328628540039062, |
| "learning_rate": 0.00010219999999999999, |
| "loss": 4.9165, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.6399753855620938, |
| "grad_norm": 29.132736206054688, |
| "learning_rate": 0.00010319999999999999, |
| "loss": 4.9574, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.6461289950386524, |
| "grad_norm": 28.395519256591797, |
| "learning_rate": 0.00010419999999999998, |
| "loss": 4.8759, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.6522826045152109, |
| "grad_norm": 26.395824432373047, |
| "learning_rate": 0.0001052, |
| "loss": 4.9002, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.6584362139917695, |
| "grad_norm": 27.33197593688965, |
| "learning_rate": 0.00010619999999999998, |
| "loss": 4.8743, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.6645898234683282, |
| "grad_norm": 25.773096084594727, |
| "learning_rate": 0.00010719999999999999, |
| "loss": 4.8093, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.6707434329448867, |
| "grad_norm": 27.487308502197266, |
| "learning_rate": 0.0001082, |
| "loss": 4.8178, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.6768970424214453, |
| "grad_norm": 26.95354652404785, |
| "learning_rate": 0.00010919999999999998, |
| "loss": 4.7463, |
| "step": 1100 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 100000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 62, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 8.33224387325952e+16, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
