| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.4618937644341801, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004618937644341801, | |
| "grad_norm": 10.975909233093262, | |
| "learning_rate": 3.4642032332563515e-07, | |
| "loss": 2.4942, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009237875288683603, | |
| "grad_norm": 9.186948776245117, | |
| "learning_rate": 9.237875288683603e-07, | |
| "loss": 2.7582, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013856812933025405, | |
| "grad_norm": 10.907584190368652, | |
| "learning_rate": 1.443418013856813e-06, | |
| "loss": 2.6814, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.018475750577367205, | |
| "grad_norm": 9.531168937683105, | |
| "learning_rate": 2.0207852193995383e-06, | |
| "loss": 2.6982, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.023094688221709007, | |
| "grad_norm": 14.727725982666016, | |
| "learning_rate": 2.5981524249422633e-06, | |
| "loss": 2.0218, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02771362586605081, | |
| "grad_norm": 8.314309120178223, | |
| "learning_rate": 3.117782909930716e-06, | |
| "loss": 1.5595, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03233256351039261, | |
| "grad_norm": 4.944284915924072, | |
| "learning_rate": 3.6951501154734412e-06, | |
| "loss": 1.02, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03695150115473441, | |
| "grad_norm": 10.882843971252441, | |
| "learning_rate": 4.272517321016166e-06, | |
| "loss": 1.0419, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04157043879907621, | |
| "grad_norm": 12.310320854187012, | |
| "learning_rate": 4.849884526558892e-06, | |
| "loss": 1.0801, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.046189376443418015, | |
| "grad_norm": 6.49992036819458, | |
| "learning_rate": 5.427251732101616e-06, | |
| "loss": 0.8444, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.050808314087759814, | |
| "grad_norm": 2.89493465423584, | |
| "learning_rate": 6.004618937644342e-06, | |
| "loss": 0.8884, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05542725173210162, | |
| "grad_norm": 3.966763734817505, | |
| "learning_rate": 6.581986143187067e-06, | |
| "loss": 0.8672, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06004618937644342, | |
| "grad_norm": 4.442293167114258, | |
| "learning_rate": 7.159353348729793e-06, | |
| "loss": 0.8037, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06466512702078522, | |
| "grad_norm": 2.506918430328369, | |
| "learning_rate": 7.736720554272519e-06, | |
| "loss": 0.724, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06928406466512702, | |
| "grad_norm": 5.733686447143555, | |
| "learning_rate": 8.314087759815242e-06, | |
| "loss": 0.7692, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07390300230946882, | |
| "grad_norm": 4.161188125610352, | |
| "learning_rate": 8.891454965357968e-06, | |
| "loss": 0.7413, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07852193995381063, | |
| "grad_norm": 3.9434962272644043, | |
| "learning_rate": 9.468822170900693e-06, | |
| "loss": 0.7386, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08314087759815242, | |
| "grad_norm": 2.9100701808929443, | |
| "learning_rate": 1.0046189376443418e-05, | |
| "loss": 0.6942, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.08775981524249422, | |
| "grad_norm": 5.367318153381348, | |
| "learning_rate": 1.0623556581986144e-05, | |
| "loss": 0.8011, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09237875288683603, | |
| "grad_norm": 3.1690614223480225, | |
| "learning_rate": 1.1200923787528869e-05, | |
| "loss": 0.6816, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09699769053117784, | |
| "grad_norm": 4.35976505279541, | |
| "learning_rate": 1.1778290993071595e-05, | |
| "loss": 0.7408, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10161662817551963, | |
| "grad_norm": 3.330937623977661, | |
| "learning_rate": 1.235565819861432e-05, | |
| "loss": 0.7159, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.10623556581986143, | |
| "grad_norm": 5.761129379272461, | |
| "learning_rate": 1.2933025404157046e-05, | |
| "loss": 0.6838, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11085450346420324, | |
| "grad_norm": 7.05668830871582, | |
| "learning_rate": 1.351039260969977e-05, | |
| "loss": 0.7135, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11547344110854503, | |
| "grad_norm": 3.7135939598083496, | |
| "learning_rate": 1.4087759815242497e-05, | |
| "loss": 0.6385, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12009237875288684, | |
| "grad_norm": 5.477907657623291, | |
| "learning_rate": 1.4665127020785218e-05, | |
| "loss": 0.6292, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.12471131639722864, | |
| "grad_norm": 6.577059268951416, | |
| "learning_rate": 1.5242494226327944e-05, | |
| "loss": 0.6921, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.12933025404157045, | |
| "grad_norm": 3.6328892707824707, | |
| "learning_rate": 1.581986143187067e-05, | |
| "loss": 0.6621, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.13394919168591224, | |
| "grad_norm": 4.084783554077148, | |
| "learning_rate": 1.6397228637413393e-05, | |
| "loss": 0.6667, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.13856812933025403, | |
| "grad_norm": 3.8719701766967773, | |
| "learning_rate": 1.697459584295612e-05, | |
| "loss": 0.6692, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14318706697459585, | |
| "grad_norm": 7.860931873321533, | |
| "learning_rate": 1.7551963048498846e-05, | |
| "loss": 0.6251, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.14780600461893764, | |
| "grad_norm": 4.381837368011475, | |
| "learning_rate": 1.812933025404157e-05, | |
| "loss": 0.6297, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.15242494226327943, | |
| "grad_norm": 3.7145886421203613, | |
| "learning_rate": 1.8706697459584295e-05, | |
| "loss": 0.6483, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.15704387990762125, | |
| "grad_norm": 2.609006643295288, | |
| "learning_rate": 1.9284064665127023e-05, | |
| "loss": 0.6149, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.16166281755196305, | |
| "grad_norm": 4.774081230163574, | |
| "learning_rate": 1.9861431870669748e-05, | |
| "loss": 0.6034, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.16628175519630484, | |
| "grad_norm": 7.305100440979004, | |
| "learning_rate": 2.0438799076212473e-05, | |
| "loss": 0.6496, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.17090069284064666, | |
| "grad_norm": 5.507181644439697, | |
| "learning_rate": 2.1016166281755197e-05, | |
| "loss": 0.643, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.17551963048498845, | |
| "grad_norm": 4.033135890960693, | |
| "learning_rate": 2.1593533487297922e-05, | |
| "loss": 0.6186, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.18013856812933027, | |
| "grad_norm": 3.903007745742798, | |
| "learning_rate": 2.217090069284065e-05, | |
| "loss": 0.6041, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.18475750577367206, | |
| "grad_norm": 4.785562992095947, | |
| "learning_rate": 2.2748267898383374e-05, | |
| "loss": 0.5527, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.18937644341801385, | |
| "grad_norm": 3.4289231300354004, | |
| "learning_rate": 2.3325635103926096e-05, | |
| "loss": 0.5936, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19399538106235567, | |
| "grad_norm": 2.384840965270996, | |
| "learning_rate": 2.3903002309468824e-05, | |
| "loss": 0.5421, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.19861431870669746, | |
| "grad_norm": 4.025755882263184, | |
| "learning_rate": 2.448036951501155e-05, | |
| "loss": 0.5839, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.20323325635103925, | |
| "grad_norm": 4.832013130187988, | |
| "learning_rate": 2.5057736720554276e-05, | |
| "loss": 0.5938, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.20785219399538107, | |
| "grad_norm": 3.66886305809021, | |
| "learning_rate": 2.5635103926096998e-05, | |
| "loss": 0.5607, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.21247113163972287, | |
| "grad_norm": 3.7285852432250977, | |
| "learning_rate": 2.6212471131639726e-05, | |
| "loss": 0.5457, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.21709006928406466, | |
| "grad_norm": 3.755711555480957, | |
| "learning_rate": 2.678983833718245e-05, | |
| "loss": 0.5721, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.22170900692840648, | |
| "grad_norm": 4.016116619110107, | |
| "learning_rate": 2.7367205542725178e-05, | |
| "loss": 0.59, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.22632794457274827, | |
| "grad_norm": 6.123377799987793, | |
| "learning_rate": 2.79445727482679e-05, | |
| "loss": 0.6236, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.23094688221709006, | |
| "grad_norm": 3.77093505859375, | |
| "learning_rate": 2.8521939953810624e-05, | |
| "loss": 0.6306, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23556581986143188, | |
| "grad_norm": 5.101199626922607, | |
| "learning_rate": 2.9099307159353352e-05, | |
| "loss": 0.5806, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.24018475750577367, | |
| "grad_norm": 4.425400257110596, | |
| "learning_rate": 2.9676674364896073e-05, | |
| "loss": 0.5644, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.24480369515011546, | |
| "grad_norm": 3.661381244659424, | |
| "learning_rate": 3.02540415704388e-05, | |
| "loss": 0.538, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.24942263279445728, | |
| "grad_norm": 3.271655559539795, | |
| "learning_rate": 3.0831408775981526e-05, | |
| "loss": 0.5927, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2540415704387991, | |
| "grad_norm": 4.603795051574707, | |
| "learning_rate": 3.140877598152425e-05, | |
| "loss": 0.5772, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2586605080831409, | |
| "grad_norm": 3.379786968231201, | |
| "learning_rate": 3.1986143187066975e-05, | |
| "loss": 0.5577, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2632794457274827, | |
| "grad_norm": 3.3702409267425537, | |
| "learning_rate": 3.25635103926097e-05, | |
| "loss": 0.6258, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2678983833718245, | |
| "grad_norm": 3.1498706340789795, | |
| "learning_rate": 3.3140877598152424e-05, | |
| "loss": 0.5631, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.27251732101616627, | |
| "grad_norm": 4.846761703491211, | |
| "learning_rate": 3.3718244803695156e-05, | |
| "loss": 0.5944, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.27713625866050806, | |
| "grad_norm": 4.397818088531494, | |
| "learning_rate": 3.4295612009237874e-05, | |
| "loss": 0.5698, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2817551963048499, | |
| "grad_norm": 2.2724201679229736, | |
| "learning_rate": 3.4872979214780605e-05, | |
| "loss": 0.5468, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2863741339491917, | |
| "grad_norm": 4.73584508895874, | |
| "learning_rate": 3.545034642032333e-05, | |
| "loss": 0.5697, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2909930715935335, | |
| "grad_norm": 3.557711124420166, | |
| "learning_rate": 3.6027713625866054e-05, | |
| "loss": 0.5195, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2956120092378753, | |
| "grad_norm": 4.273311614990234, | |
| "learning_rate": 3.660508083140878e-05, | |
| "loss": 0.5561, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3002309468822171, | |
| "grad_norm": 3.6489686965942383, | |
| "learning_rate": 3.7182448036951504e-05, | |
| "loss": 0.566, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.30484988452655887, | |
| "grad_norm": 3.4011149406433105, | |
| "learning_rate": 3.775981524249423e-05, | |
| "loss": 0.5468, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3094688221709007, | |
| "grad_norm": 3.517822742462158, | |
| "learning_rate": 3.833718244803695e-05, | |
| "loss": 0.5834, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3140877598152425, | |
| "grad_norm": 2.7577602863311768, | |
| "learning_rate": 3.891454965357968e-05, | |
| "loss": 0.5489, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3187066974595843, | |
| "grad_norm": 2.856598138809204, | |
| "learning_rate": 3.94919168591224e-05, | |
| "loss": 0.4906, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3233256351039261, | |
| "grad_norm": 4.136464595794678, | |
| "learning_rate": 4.0069284064665133e-05, | |
| "loss": 0.5705, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3279445727482679, | |
| "grad_norm": 4.121008396148682, | |
| "learning_rate": 4.064665127020785e-05, | |
| "loss": 0.6241, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3325635103926097, | |
| "grad_norm": 3.611814498901367, | |
| "learning_rate": 4.122401847575058e-05, | |
| "loss": 0.5034, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3371824480369515, | |
| "grad_norm": 3.4624834060668945, | |
| "learning_rate": 4.18013856812933e-05, | |
| "loss": 0.5643, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3418013856812933, | |
| "grad_norm": 3.3586976528167725, | |
| "learning_rate": 4.237875288683603e-05, | |
| "loss": 0.5439, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3464203233256351, | |
| "grad_norm": 2.5129313468933105, | |
| "learning_rate": 4.2956120092378757e-05, | |
| "loss": 0.5062, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3510392609699769, | |
| "grad_norm": 3.0052690505981445, | |
| "learning_rate": 4.353348729792148e-05, | |
| "loss": 0.5219, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.3556581986143187, | |
| "grad_norm": 3.7070388793945312, | |
| "learning_rate": 4.4110854503464206e-05, | |
| "loss": 0.5425, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.36027713625866054, | |
| "grad_norm": 2.8378756046295166, | |
| "learning_rate": 4.468822170900693e-05, | |
| "loss": 0.5208, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3648960739030023, | |
| "grad_norm": 2.3988196849823, | |
| "learning_rate": 4.5265588914549655e-05, | |
| "loss": 0.5149, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3695150115473441, | |
| "grad_norm": 2.2305569648742676, | |
| "learning_rate": 4.584295612009238e-05, | |
| "loss": 0.5028, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3741339491916859, | |
| "grad_norm": 3.0817391872406006, | |
| "learning_rate": 4.6420323325635104e-05, | |
| "loss": 0.5419, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.3787528868360277, | |
| "grad_norm": 2.767381429672241, | |
| "learning_rate": 4.699769053117783e-05, | |
| "loss": 0.5036, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.3833718244803695, | |
| "grad_norm": 2.8563129901885986, | |
| "learning_rate": 4.757505773672056e-05, | |
| "loss": 0.5079, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.38799076212471134, | |
| "grad_norm": 4.459218978881836, | |
| "learning_rate": 4.815242494226328e-05, | |
| "loss": 0.4891, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.39260969976905313, | |
| "grad_norm": 2.825631856918335, | |
| "learning_rate": 4.872979214780601e-05, | |
| "loss": 0.4921, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.3972286374133949, | |
| "grad_norm": 2.835643768310547, | |
| "learning_rate": 4.9307159353348734e-05, | |
| "loss": 0.4736, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4018475750577367, | |
| "grad_norm": 3.237922430038452, | |
| "learning_rate": 4.988452655889146e-05, | |
| "loss": 0.5254, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4064665127020785, | |
| "grad_norm": 3.913771390914917, | |
| "learning_rate": 4.9999870022388165e-05, | |
| "loss": 0.5212, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4110854503464203, | |
| "grad_norm": 4.357654094696045, | |
| "learning_rate": 4.999934199065641e-05, | |
| "loss": 0.6004, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.41570438799076215, | |
| "grad_norm": 2.1519246101379395, | |
| "learning_rate": 4.999840778977644e-05, | |
| "loss": 0.5106, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.42032332563510394, | |
| "grad_norm": 3.196580410003662, | |
| "learning_rate": 4.9997067434926386e-05, | |
| "loss": 0.4708, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.42494226327944573, | |
| "grad_norm": 3.0933518409729004, | |
| "learning_rate": 4.9995320947883265e-05, | |
| "loss": 0.5228, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.4295612009237875, | |
| "grad_norm": 2.5883092880249023, | |
| "learning_rate": 4.999316835702259e-05, | |
| "loss": 0.54, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.4341801385681293, | |
| "grad_norm": 2.738126277923584, | |
| "learning_rate": 4.9990609697317916e-05, | |
| "loss": 0.4965, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.4387990762124711, | |
| "grad_norm": 2.903364419937134, | |
| "learning_rate": 4.998764501034028e-05, | |
| "loss": 0.4732, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.44341801385681295, | |
| "grad_norm": 1.895765781402588, | |
| "learning_rate": 4.998427434425753e-05, | |
| "loss": 0.4857, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.44803695150115475, | |
| "grad_norm": 5.765718936920166, | |
| "learning_rate": 4.998049775383353e-05, | |
| "loss": 0.4901, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.45265588914549654, | |
| "grad_norm": 3.9257349967956543, | |
| "learning_rate": 4.997631530042727e-05, | |
| "loss": 0.493, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.45727482678983833, | |
| "grad_norm": 2.3068912029266357, | |
| "learning_rate": 4.997172705199189e-05, | |
| "loss": 0.5061, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.4618937644341801, | |
| "grad_norm": 2.966820001602173, | |
| "learning_rate": 4.996673308307355e-05, | |
| "loss": 0.4557, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 8660, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.714316759199908e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |