{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0703857487275648, "eval_steps": 500, "global_step": 3500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.000669702652022502, "grad_norm": 1.0546875, "learning_rate": 6.693440428380188e-07, "loss": 2.3721, "step": 10 }, { "epoch": 0.001339405304045004, "grad_norm": 0.69921875, "learning_rate": 1.3386880856760376e-06, "loss": 2.3552, "step": 20 }, { "epoch": 0.002009107956067506, "grad_norm": 0.65234375, "learning_rate": 2.0080321285140564e-06, "loss": 2.3693, "step": 30 }, { "epoch": 0.002678810608090008, "grad_norm": 0.52734375, "learning_rate": 2.6773761713520752e-06, "loss": 2.345, "step": 40 }, { "epoch": 0.00334851326011251, "grad_norm": 1.4375, "learning_rate": 3.346720214190094e-06, "loss": 2.323, "step": 50 }, { "epoch": 0.004018215912135012, "grad_norm": 1.1953125, "learning_rate": 4.016064257028113e-06, "loss": 2.3123, "step": 60 }, { "epoch": 0.004687918564157514, "grad_norm": 1.109375, "learning_rate": 4.685408299866132e-06, "loss": 2.3204, "step": 70 }, { "epoch": 0.005357621216180016, "grad_norm": 0.9921875, "learning_rate": 5.3547523427041504e-06, "loss": 2.306, "step": 80 }, { "epoch": 0.006027323868202518, "grad_norm": 0.91796875, "learning_rate": 6.024096385542169e-06, "loss": 2.2981, "step": 90 }, { "epoch": 0.00669702652022502, "grad_norm": 0.275390625, "learning_rate": 6.693440428380188e-06, "loss": 2.3155, "step": 100 }, { "epoch": 0.0073667291722475225, "grad_norm": 0.267578125, "learning_rate": 7.362784471218207e-06, "loss": 2.3209, "step": 110 }, { "epoch": 0.008036431824270024, "grad_norm": 0.333984375, "learning_rate": 8.032128514056226e-06, "loss": 2.3155, "step": 120 }, { "epoch": 0.008706134476292525, "grad_norm": 0.29296875, "learning_rate": 8.701472556894244e-06, "loss": 2.303, "step": 130 }, { "epoch": 0.009375837128315028, "grad_norm": 0.322265625, "learning_rate": 9.370816599732263e-06, "loss": 2.3224, "step": 140 }, { "epoch": 0.01004553978033753, "grad_norm": 0.2890625, "learning_rate": 1.0040160642570281e-05, "loss": 2.2726, "step": 150 }, { "epoch": 0.010715242432360031, "grad_norm": 0.294921875, "learning_rate": 1.0709504685408301e-05, "loss": 2.2799, "step": 160 }, { "epoch": 0.011384945084382534, "grad_norm": 0.267578125, "learning_rate": 1.1378848728246319e-05, "loss": 2.2789, "step": 170 }, { "epoch": 0.012054647736405036, "grad_norm": 0.267578125, "learning_rate": 1.2048192771084338e-05, "loss": 2.2618, "step": 180 }, { "epoch": 0.012724350388427539, "grad_norm": 0.25, "learning_rate": 1.2717536813922356e-05, "loss": 2.2943, "step": 190 }, { "epoch": 0.01339405304045004, "grad_norm": 0.259765625, "learning_rate": 1.3386880856760376e-05, "loss": 2.2806, "step": 200 }, { "epoch": 0.014063755692472542, "grad_norm": 0.251953125, "learning_rate": 1.4056224899598394e-05, "loss": 2.2543, "step": 210 }, { "epoch": 0.014733458344495045, "grad_norm": 0.234375, "learning_rate": 1.4725568942436414e-05, "loss": 2.2559, "step": 220 }, { "epoch": 0.015403160996517546, "grad_norm": 0.2490234375, "learning_rate": 1.5394912985274433e-05, "loss": 2.2741, "step": 230 }, { "epoch": 0.016072863648540048, "grad_norm": 0.2333984375, "learning_rate": 1.606425702811245e-05, "loss": 2.2753, "step": 240 }, { "epoch": 0.01674256630056255, "grad_norm": 0.259765625, "learning_rate": 1.673360107095047e-05, "loss": 2.2856, "step": 250 }, { "epoch": 0.01741226895258505, "grad_norm": 0.2421875, "learning_rate": 1.7402945113788487e-05, "loss": 2.2625, "step": 260 }, { "epoch": 0.018081971604607554, "grad_norm": 0.234375, "learning_rate": 1.8072289156626505e-05, "loss": 2.2666, "step": 270 }, { "epoch": 0.018751674256630057, "grad_norm": 0.236328125, "learning_rate": 1.8741633199464527e-05, "loss": 2.2704, "step": 280 }, { "epoch": 0.019421376908652557, "grad_norm": 0.21484375, "learning_rate": 1.9410977242302544e-05, "loss": 2.2582, "step": 290 }, { "epoch": 0.02009107956067506, "grad_norm": 0.2421875, "learning_rate": 2.0080321285140562e-05, "loss": 2.2666, "step": 300 }, { "epoch": 0.020760782212697563, "grad_norm": 0.25390625, "learning_rate": 2.074966532797858e-05, "loss": 2.2732, "step": 310 }, { "epoch": 0.021430484864720063, "grad_norm": 0.2294921875, "learning_rate": 2.1419009370816602e-05, "loss": 2.2367, "step": 320 }, { "epoch": 0.022100187516742566, "grad_norm": 0.2236328125, "learning_rate": 2.208835341365462e-05, "loss": 2.2618, "step": 330 }, { "epoch": 0.02276989016876507, "grad_norm": 0.228515625, "learning_rate": 2.2757697456492638e-05, "loss": 2.2544, "step": 340 }, { "epoch": 0.023439592820787572, "grad_norm": 0.2578125, "learning_rate": 2.3427041499330656e-05, "loss": 2.2896, "step": 350 }, { "epoch": 0.02410929547281007, "grad_norm": 0.2353515625, "learning_rate": 2.4096385542168677e-05, "loss": 2.269, "step": 360 }, { "epoch": 0.024778998124832575, "grad_norm": 0.25, "learning_rate": 2.4765729585006695e-05, "loss": 2.2676, "step": 370 }, { "epoch": 0.025448700776855078, "grad_norm": 0.228515625, "learning_rate": 2.5435073627844713e-05, "loss": 2.2498, "step": 380 }, { "epoch": 0.026118403428877578, "grad_norm": 0.228515625, "learning_rate": 2.6104417670682734e-05, "loss": 2.2784, "step": 390 }, { "epoch": 0.02678810608090008, "grad_norm": 0.248046875, "learning_rate": 2.6773761713520752e-05, "loss": 2.2946, "step": 400 }, { "epoch": 0.027457808732922584, "grad_norm": 0.23828125, "learning_rate": 2.7443105756358774e-05, "loss": 2.2536, "step": 410 }, { "epoch": 0.028127511384945084, "grad_norm": 0.21875, "learning_rate": 2.8112449799196788e-05, "loss": 2.2978, "step": 420 }, { "epoch": 0.028797214036967587, "grad_norm": 0.22265625, "learning_rate": 2.878179384203481e-05, "loss": 2.2671, "step": 430 }, { "epoch": 0.02946691668899009, "grad_norm": 0.22265625, "learning_rate": 2.9451137884872827e-05, "loss": 2.2552, "step": 440 }, { "epoch": 0.03013661934101259, "grad_norm": 0.236328125, "learning_rate": 3.012048192771085e-05, "loss": 2.2439, "step": 450 }, { "epoch": 0.030806321993035093, "grad_norm": 0.2236328125, "learning_rate": 3.078982597054887e-05, "loss": 2.2697, "step": 460 }, { "epoch": 0.031476024645057596, "grad_norm": 0.2294921875, "learning_rate": 3.1459170013386885e-05, "loss": 2.2334, "step": 470 }, { "epoch": 0.032145727297080096, "grad_norm": 0.228515625, "learning_rate": 3.21285140562249e-05, "loss": 2.2282, "step": 480 }, { "epoch": 0.032815429949102595, "grad_norm": 0.228515625, "learning_rate": 3.279785809906292e-05, "loss": 2.2628, "step": 490 }, { "epoch": 0.0334851326011251, "grad_norm": 0.2177734375, "learning_rate": 3.346720214190094e-05, "loss": 2.2855, "step": 500 }, { "epoch": 0.0341548352531476, "grad_norm": 0.216796875, "learning_rate": 3.413654618473896e-05, "loss": 2.2603, "step": 510 }, { "epoch": 0.0348245379051701, "grad_norm": 0.2060546875, "learning_rate": 3.4805890227576974e-05, "loss": 2.2287, "step": 520 }, { "epoch": 0.03549424055719261, "grad_norm": 0.2001953125, "learning_rate": 3.5475234270415e-05, "loss": 2.2635, "step": 530 }, { "epoch": 0.03616394320921511, "grad_norm": 0.1943359375, "learning_rate": 3.614457831325301e-05, "loss": 2.2376, "step": 540 }, { "epoch": 0.03683364586123761, "grad_norm": 0.21875, "learning_rate": 3.6813922356091035e-05, "loss": 2.2782, "step": 550 }, { "epoch": 0.037503348513260114, "grad_norm": 0.220703125, "learning_rate": 3.748326639892905e-05, "loss": 2.2426, "step": 560 }, { "epoch": 0.038173051165282613, "grad_norm": 0.21875, "learning_rate": 3.815261044176707e-05, "loss": 2.2376, "step": 570 }, { "epoch": 0.03884275381730511, "grad_norm": 0.2109375, "learning_rate": 3.882195448460509e-05, "loss": 2.2515, "step": 580 }, { "epoch": 0.03951245646932762, "grad_norm": 0.205078125, "learning_rate": 3.949129852744311e-05, "loss": 2.223, "step": 590 }, { "epoch": 0.04018215912135012, "grad_norm": 0.2119140625, "learning_rate": 4.0160642570281125e-05, "loss": 2.2776, "step": 600 }, { "epoch": 0.04085186177337262, "grad_norm": 0.1953125, "learning_rate": 4.082998661311915e-05, "loss": 2.2262, "step": 610 }, { "epoch": 0.041521564425395126, "grad_norm": 0.1943359375, "learning_rate": 4.149933065595716e-05, "loss": 2.2124, "step": 620 }, { "epoch": 0.042191267077417625, "grad_norm": 0.1904296875, "learning_rate": 4.2168674698795186e-05, "loss": 2.2626, "step": 630 }, { "epoch": 0.042860969729440125, "grad_norm": 0.1884765625, "learning_rate": 4.2838018741633203e-05, "loss": 2.2565, "step": 640 }, { "epoch": 0.04353067238146263, "grad_norm": 0.2236328125, "learning_rate": 4.350736278447122e-05, "loss": 2.2399, "step": 650 }, { "epoch": 0.04420037503348513, "grad_norm": 0.2353515625, "learning_rate": 4.417670682730924e-05, "loss": 2.202, "step": 660 }, { "epoch": 0.04487007768550764, "grad_norm": 0.197265625, "learning_rate": 4.484605087014726e-05, "loss": 2.2397, "step": 670 }, { "epoch": 0.04553978033753014, "grad_norm": 0.2001953125, "learning_rate": 4.5515394912985275e-05, "loss": 2.2208, "step": 680 }, { "epoch": 0.04620948298955264, "grad_norm": 0.201171875, "learning_rate": 4.61847389558233e-05, "loss": 2.2599, "step": 690 }, { "epoch": 0.046879185641575144, "grad_norm": 0.2001953125, "learning_rate": 4.685408299866131e-05, "loss": 2.2363, "step": 700 }, { "epoch": 0.047548888293597644, "grad_norm": 0.2119140625, "learning_rate": 4.7523427041499336e-05, "loss": 2.2617, "step": 710 }, { "epoch": 0.04821859094562014, "grad_norm": 0.189453125, "learning_rate": 4.8192771084337354e-05, "loss": 2.2539, "step": 720 }, { "epoch": 0.04888829359764265, "grad_norm": 0.1962890625, "learning_rate": 4.886211512717537e-05, "loss": 2.242, "step": 730 }, { "epoch": 0.04955799624966515, "grad_norm": 0.1865234375, "learning_rate": 4.953145917001339e-05, "loss": 2.2315, "step": 740 }, { "epoch": 0.05022769890168765, "grad_norm": 0.1884765625, "learning_rate": 5.020080321285141e-05, "loss": 2.244, "step": 750 }, { "epoch": 0.050897401553710156, "grad_norm": 0.2041015625, "learning_rate": 5.0870147255689426e-05, "loss": 2.2261, "step": 760 }, { "epoch": 0.051567104205732656, "grad_norm": 0.18359375, "learning_rate": 5.1539491298527444e-05, "loss": 2.2324, "step": 770 }, { "epoch": 0.052236806857755155, "grad_norm": 0.1787109375, "learning_rate": 5.220883534136547e-05, "loss": 2.2235, "step": 780 }, { "epoch": 0.05290650950977766, "grad_norm": 0.1845703125, "learning_rate": 5.2878179384203486e-05, "loss": 2.2199, "step": 790 }, { "epoch": 0.05357621216180016, "grad_norm": 0.185546875, "learning_rate": 5.3547523427041504e-05, "loss": 2.2198, "step": 800 }, { "epoch": 0.05424591481382266, "grad_norm": 0.181640625, "learning_rate": 5.4216867469879516e-05, "loss": 2.2227, "step": 810 }, { "epoch": 0.05491561746584517, "grad_norm": 0.17578125, "learning_rate": 5.488621151271755e-05, "loss": 2.2328, "step": 820 }, { "epoch": 0.05558532011786767, "grad_norm": 0.16796875, "learning_rate": 5.555555555555556e-05, "loss": 2.2186, "step": 830 }, { "epoch": 0.05625502276989017, "grad_norm": 0.1650390625, "learning_rate": 5.6224899598393576e-05, "loss": 2.2523, "step": 840 }, { "epoch": 0.056924725421912674, "grad_norm": 0.1708984375, "learning_rate": 5.6894243641231594e-05, "loss": 2.2318, "step": 850 }, { "epoch": 0.057594428073935174, "grad_norm": 0.162109375, "learning_rate": 5.756358768406962e-05, "loss": 2.2264, "step": 860 }, { "epoch": 0.05826413072595767, "grad_norm": 0.1611328125, "learning_rate": 5.823293172690764e-05, "loss": 2.188, "step": 870 }, { "epoch": 0.05893383337798018, "grad_norm": 0.1552734375, "learning_rate": 5.8902275769745655e-05, "loss": 2.2095, "step": 880 }, { "epoch": 0.05960353603000268, "grad_norm": 0.15234375, "learning_rate": 5.9571619812583666e-05, "loss": 2.2364, "step": 890 }, { "epoch": 0.06027323868202518, "grad_norm": 0.1513671875, "learning_rate": 6.02409638554217e-05, "loss": 2.2416, "step": 900 }, { "epoch": 0.060942941334047686, "grad_norm": 0.14453125, "learning_rate": 6.091030789825971e-05, "loss": 2.231, "step": 910 }, { "epoch": 0.061612643986070185, "grad_norm": 0.1416015625, "learning_rate": 6.157965194109773e-05, "loss": 2.205, "step": 920 }, { "epoch": 0.062282346638092685, "grad_norm": 0.1416015625, "learning_rate": 6.224899598393574e-05, "loss": 2.225, "step": 930 }, { "epoch": 0.06295204929011519, "grad_norm": 0.1435546875, "learning_rate": 6.291834002677377e-05, "loss": 2.2272, "step": 940 }, { "epoch": 0.06362175194213769, "grad_norm": 0.134765625, "learning_rate": 6.358768406961179e-05, "loss": 2.1913, "step": 950 }, { "epoch": 0.06429145459416019, "grad_norm": 0.1376953125, "learning_rate": 6.42570281124498e-05, "loss": 2.2537, "step": 960 }, { "epoch": 0.06496115724618269, "grad_norm": 0.134765625, "learning_rate": 6.492637215528782e-05, "loss": 2.2265, "step": 970 }, { "epoch": 0.06563085989820519, "grad_norm": 0.1318359375, "learning_rate": 6.559571619812584e-05, "loss": 2.2061, "step": 980 }, { "epoch": 0.0663005625502277, "grad_norm": 0.1337890625, "learning_rate": 6.626506024096386e-05, "loss": 2.2064, "step": 990 }, { "epoch": 0.0669702652022502, "grad_norm": 0.1279296875, "learning_rate": 6.693440428380188e-05, "loss": 2.1914, "step": 1000 }, { "epoch": 0.0676399678542727, "grad_norm": 0.1240234375, "learning_rate": 6.76037483266399e-05, "loss": 2.2139, "step": 1010 }, { "epoch": 0.0683096705062952, "grad_norm": 0.1279296875, "learning_rate": 6.827309236947793e-05, "loss": 2.1974, "step": 1020 }, { "epoch": 0.0689793731583177, "grad_norm": 0.11865234375, "learning_rate": 6.894243641231593e-05, "loss": 2.1966, "step": 1030 }, { "epoch": 0.0696490758103402, "grad_norm": 0.123046875, "learning_rate": 6.961178045515395e-05, "loss": 2.2377, "step": 1040 }, { "epoch": 0.07031877846236272, "grad_norm": 0.1279296875, "learning_rate": 7.028112449799197e-05, "loss": 2.2083, "step": 1050 }, { "epoch": 0.07098848111438522, "grad_norm": 0.130859375, "learning_rate": 7.095046854083e-05, "loss": 2.219, "step": 1060 }, { "epoch": 0.07165818376640772, "grad_norm": 0.12060546875, "learning_rate": 7.161981258366802e-05, "loss": 2.2395, "step": 1070 }, { "epoch": 0.07232788641843022, "grad_norm": 0.1201171875, "learning_rate": 7.228915662650602e-05, "loss": 2.2282, "step": 1080 }, { "epoch": 0.07299758907045271, "grad_norm": 0.12109375, "learning_rate": 7.295850066934404e-05, "loss": 2.2116, "step": 1090 }, { "epoch": 0.07366729172247521, "grad_norm": 0.11572265625, "learning_rate": 7.362784471218207e-05, "loss": 2.2614, "step": 1100 }, { "epoch": 0.07433699437449773, "grad_norm": 0.10888671875, "learning_rate": 7.429718875502009e-05, "loss": 2.1997, "step": 1110 }, { "epoch": 0.07500669702652023, "grad_norm": 0.10693359375, "learning_rate": 7.49665327978581e-05, "loss": 2.209, "step": 1120 }, { "epoch": 0.07567639967854273, "grad_norm": 0.11083984375, "learning_rate": 7.563587684069612e-05, "loss": 2.1891, "step": 1130 }, { "epoch": 0.07634610233056523, "grad_norm": 0.11279296875, "learning_rate": 7.630522088353414e-05, "loss": 2.2325, "step": 1140 }, { "epoch": 0.07701580498258773, "grad_norm": 0.10205078125, "learning_rate": 7.697456492637216e-05, "loss": 2.2616, "step": 1150 }, { "epoch": 0.07768550763461023, "grad_norm": 0.1015625, "learning_rate": 7.764390896921018e-05, "loss": 2.2526, "step": 1160 }, { "epoch": 0.07835521028663274, "grad_norm": 0.10205078125, "learning_rate": 7.83132530120482e-05, "loss": 2.2339, "step": 1170 }, { "epoch": 0.07902491293865524, "grad_norm": 0.1025390625, "learning_rate": 7.898259705488621e-05, "loss": 2.2019, "step": 1180 }, { "epoch": 0.07969461559067774, "grad_norm": 0.0986328125, "learning_rate": 7.965194109772423e-05, "loss": 2.2269, "step": 1190 }, { "epoch": 0.08036431824270024, "grad_norm": 0.09765625, "learning_rate": 8.032128514056225e-05, "loss": 2.2104, "step": 1200 }, { "epoch": 0.08103402089472274, "grad_norm": 0.0966796875, "learning_rate": 8.099062918340027e-05, "loss": 2.2159, "step": 1210 }, { "epoch": 0.08170372354674524, "grad_norm": 0.0966796875, "learning_rate": 8.16599732262383e-05, "loss": 2.2059, "step": 1220 }, { "epoch": 1.0003683364586125, "grad_norm": 0.09814453125, "learning_rate": 8.232931726907632e-05, "loss": 2.2994, "step": 1230 }, { "epoch": 1.0010380391106348, "grad_norm": 0.1005859375, "learning_rate": 8.299866131191432e-05, "loss": 2.1363, "step": 1240 }, { "epoch": 1.0017077417626574, "grad_norm": 0.09912109375, "learning_rate": 8.366800535475234e-05, "loss": 2.1779, "step": 1250 }, { "epoch": 1.0023774444146798, "grad_norm": 0.10693359375, "learning_rate": 8.433734939759037e-05, "loss": 2.2136, "step": 1260 }, { "epoch": 1.0030471470667024, "grad_norm": 0.09521484375, "learning_rate": 8.500669344042839e-05, "loss": 2.1424, "step": 1270 }, { "epoch": 1.003716849718725, "grad_norm": 0.0947265625, "learning_rate": 8.567603748326641e-05, "loss": 2.158, "step": 1280 }, { "epoch": 1.0043865523707474, "grad_norm": 0.0947265625, "learning_rate": 8.634538152610442e-05, "loss": 2.2189, "step": 1290 }, { "epoch": 1.00505625502277, "grad_norm": 0.09619140625, "learning_rate": 8.701472556894244e-05, "loss": 2.1891, "step": 1300 }, { "epoch": 1.0057259576747923, "grad_norm": 0.095703125, "learning_rate": 8.768406961178046e-05, "loss": 2.2035, "step": 1310 }, { "epoch": 1.006395660326815, "grad_norm": 0.09619140625, "learning_rate": 8.835341365461848e-05, "loss": 2.1913, "step": 1320 }, { "epoch": 1.0070653629788373, "grad_norm": 0.0888671875, "learning_rate": 8.90227576974565e-05, "loss": 2.1379, "step": 1330 }, { "epoch": 1.00773506563086, "grad_norm": 0.083984375, "learning_rate": 8.969210174029451e-05, "loss": 2.1785, "step": 1340 }, { "epoch": 1.0084047682828825, "grad_norm": 0.10888671875, "learning_rate": 9.036144578313253e-05, "loss": 2.1621, "step": 1350 }, { "epoch": 1.0090744709349049, "grad_norm": 0.1005859375, "learning_rate": 9.103078982597055e-05, "loss": 2.2017, "step": 1360 }, { "epoch": 1.0097441735869275, "grad_norm": 0.08984375, "learning_rate": 9.170013386880857e-05, "loss": 2.1408, "step": 1370 }, { "epoch": 1.0104138762389498, "grad_norm": 0.09423828125, "learning_rate": 9.23694779116466e-05, "loss": 2.1555, "step": 1380 }, { "epoch": 1.0110835788909724, "grad_norm": 0.0927734375, "learning_rate": 9.303882195448462e-05, "loss": 2.1959, "step": 1390 }, { "epoch": 1.0117532815429948, "grad_norm": 0.10009765625, "learning_rate": 9.370816599732262e-05, "loss": 2.158, "step": 1400 }, { "epoch": 1.0124229841950174, "grad_norm": 0.1064453125, "learning_rate": 9.437751004016064e-05, "loss": 2.1815, "step": 1410 }, { "epoch": 1.01309268684704, "grad_norm": 0.09375, "learning_rate": 9.504685408299867e-05, "loss": 2.1719, "step": 1420 }, { "epoch": 1.0137623894990624, "grad_norm": 0.09326171875, "learning_rate": 9.571619812583669e-05, "loss": 2.1707, "step": 1430 }, { "epoch": 1.014432092151085, "grad_norm": 0.09619140625, "learning_rate": 9.638554216867471e-05, "loss": 2.179, "step": 1440 }, { "epoch": 1.0151017948031074, "grad_norm": 0.140625, "learning_rate": 9.705488621151271e-05, "loss": 2.1658, "step": 1450 }, { "epoch": 1.01577149745513, "grad_norm": 0.13671875, "learning_rate": 9.772423025435074e-05, "loss": 2.2005, "step": 1460 }, { "epoch": 1.0164412001071523, "grad_norm": 0.10888671875, "learning_rate": 9.839357429718876e-05, "loss": 2.1817, "step": 1470 }, { "epoch": 1.017110902759175, "grad_norm": 0.10791015625, "learning_rate": 9.906291834002678e-05, "loss": 2.163, "step": 1480 }, { "epoch": 1.0177806054111975, "grad_norm": 0.10400390625, "learning_rate": 9.97322623828648e-05, "loss": 2.1742, "step": 1490 }, { "epoch": 1.01845030806322, "grad_norm": 0.1591796875, "learning_rate": 9.999995081044314e-05, "loss": 2.19, "step": 1500 }, { "epoch": 1.0191200107152425, "grad_norm": 0.2138671875, "learning_rate": 9.999965020794615e-05, "loss": 2.1837, "step": 1510 }, { "epoch": 1.0197897133672649, "grad_norm": 0.265625, "learning_rate": 9.99990763321247e-05, "loss": 2.2064, "step": 1520 }, { "epoch": 1.0204594160192875, "grad_norm": 0.18359375, "learning_rate": 9.999822918611533e-05, "loss": 2.2009, "step": 1530 }, { "epoch": 1.0211291186713098, "grad_norm": 0.1513671875, "learning_rate": 9.999710877454811e-05, "loss": 2.2013, "step": 1540 }, { "epoch": 1.0217988213233324, "grad_norm": 0.47265625, "learning_rate": 9.999571510354664e-05, "loss": 2.1684, "step": 1550 }, { "epoch": 1.022468523975355, "grad_norm": 2.125, "learning_rate": 9.999404818072808e-05, "loss": 2.3948, "step": 1560 }, { "epoch": 1.0231382266273774, "grad_norm": 1.25, "learning_rate": 9.999210801520296e-05, "loss": 2.2336, "step": 1570 }, { "epoch": 1.0238079292794, "grad_norm": 1.046875, "learning_rate": 9.998989461757526e-05, "loss": 2.2113, "step": 1580 }, { "epoch": 1.0244776319314224, "grad_norm": 0.59765625, "learning_rate": 9.998740799994235e-05, "loss": 2.1963, "step": 1590 }, { "epoch": 1.025147334583445, "grad_norm": 0.435546875, "learning_rate": 9.998464817589484e-05, "loss": 2.1907, "step": 1600 }, { "epoch": 1.0258170372354674, "grad_norm": 0.51953125, "learning_rate": 9.998161516051656e-05, "loss": 2.1702, "step": 1610 }, { "epoch": 1.02648673988749, "grad_norm": 0.498046875, "learning_rate": 9.997830897038446e-05, "loss": 2.2083, "step": 1620 }, { "epoch": 1.0271564425395125, "grad_norm": 0.62890625, "learning_rate": 9.997472962356854e-05, "loss": 2.1807, "step": 1630 }, { "epoch": 1.027826145191535, "grad_norm": 0.8046875, "learning_rate": 9.997087713963174e-05, "loss": 2.2268, "step": 1640 }, { "epoch": 1.0284958478435575, "grad_norm": 1.3125, "learning_rate": 9.996675153962984e-05, "loss": 2.2356, "step": 1650 }, { "epoch": 1.02916555049558, "grad_norm": 1.0078125, "learning_rate": 9.996235284611131e-05, "loss": 2.2278, "step": 1660 }, { "epoch": 1.0298352531476025, "grad_norm": 0.9609375, "learning_rate": 9.995768108311722e-05, "loss": 2.2257, "step": 1670 }, { "epoch": 1.0305049557996249, "grad_norm": 0.88671875, "learning_rate": 9.995273627618116e-05, "loss": 2.2913, "step": 1680 }, { "epoch": 1.0311746584516475, "grad_norm": 2.421875, "learning_rate": 9.994751845232894e-05, "loss": 2.2634, "step": 1690 }, { "epoch": 1.03184436110367, "grad_norm": 0.294921875, "learning_rate": 9.994202764007865e-05, "loss": 2.1909, "step": 1700 }, { "epoch": 1.0325140637556924, "grad_norm": 0.3203125, "learning_rate": 9.993626386944031e-05, "loss": 2.2492, "step": 1710 }, { "epoch": 1.033183766407715, "grad_norm": 0.26953125, "learning_rate": 9.993022717191584e-05, "loss": 2.203, "step": 1720 }, { "epoch": 1.0338534690597374, "grad_norm": 0.291015625, "learning_rate": 9.992391758049889e-05, "loss": 2.2661, "step": 1730 }, { "epoch": 1.03452317171176, "grad_norm": 0.259765625, "learning_rate": 9.99173351296745e-05, "loss": 2.1845, "step": 1740 }, { "epoch": 1.0351928743637824, "grad_norm": 0.25390625, "learning_rate": 9.99104798554191e-05, "loss": 2.2189, "step": 1750 }, { "epoch": 1.035862577015805, "grad_norm": 0.263671875, "learning_rate": 9.990335179520023e-05, "loss": 2.4898, "step": 1760 }, { "epoch": 1.0365322796678276, "grad_norm": 0.248046875, "learning_rate": 9.989595098797635e-05, "loss": 2.2128, "step": 1770 }, { "epoch": 1.03720198231985, "grad_norm": 0.2138671875, "learning_rate": 9.988827747419659e-05, "loss": 2.2162, "step": 1780 }, { "epoch": 1.0378716849718725, "grad_norm": 0.265625, "learning_rate": 9.988033129580059e-05, "loss": 2.3657, "step": 1790 }, { "epoch": 1.038541387623895, "grad_norm": 0.26171875, "learning_rate": 9.98721124962182e-05, "loss": 2.2101, "step": 1800 }, { "epoch": 1.0392110902759175, "grad_norm": 0.2490234375, "learning_rate": 9.986362112036935e-05, "loss": 2.2097, "step": 1810 }, { "epoch": 1.03988079292794, "grad_norm": 0.205078125, "learning_rate": 9.985485721466366e-05, "loss": 2.2505, "step": 1820 }, { "epoch": 1.0405504955799625, "grad_norm": 0.2216796875, "learning_rate": 9.984582082700029e-05, "loss": 2.3071, "step": 1830 }, { "epoch": 1.041220198231985, "grad_norm": 0.2236328125, "learning_rate": 9.98365120067677e-05, "loss": 2.2063, "step": 1840 }, { "epoch": 1.0418899008840075, "grad_norm": 0.24609375, "learning_rate": 9.982693080484327e-05, "loss": 2.4559, "step": 1850 }, { "epoch": 1.04255960353603, "grad_norm": 0.2275390625, "learning_rate": 9.981707727359308e-05, "loss": 2.2928, "step": 1860 }, { "epoch": 1.0432293061880524, "grad_norm": 0.2216796875, "learning_rate": 9.98069514668717e-05, "loss": 2.2455, "step": 1870 }, { "epoch": 1.043899008840075, "grad_norm": 0.25, "learning_rate": 9.979655344002172e-05, "loss": 2.2244, "step": 1880 }, { "epoch": 1.0445687114920974, "grad_norm": 0.326171875, "learning_rate": 9.978588324987365e-05, "loss": 2.3511, "step": 1890 }, { "epoch": 1.04523841414412, "grad_norm": 0.259765625, "learning_rate": 9.977494095474546e-05, "loss": 2.1951, "step": 1900 }, { "epoch": 1.0459081167961426, "grad_norm": 0.232421875, "learning_rate": 9.97637266144423e-05, "loss": 2.2606, "step": 1910 }, { "epoch": 1.046577819448165, "grad_norm": 0.2080078125, "learning_rate": 9.975224029025619e-05, "loss": 2.2298, "step": 1920 }, { "epoch": 1.0472475221001876, "grad_norm": 0.19921875, "learning_rate": 9.974048204496572e-05, "loss": 2.2982, "step": 1930 }, { "epoch": 1.04791722475221, "grad_norm": 0.2080078125, "learning_rate": 9.97284519428356e-05, "loss": 2.2726, "step": 1940 }, { "epoch": 1.0485869274042325, "grad_norm": 0.2578125, "learning_rate": 9.971615004961645e-05, "loss": 2.2034, "step": 1950 }, { "epoch": 1.049256630056255, "grad_norm": 0.2138671875, "learning_rate": 9.970357643254429e-05, "loss": 2.2918, "step": 1960 }, { "epoch": 1.0499263327082775, "grad_norm": 0.208984375, "learning_rate": 9.96907311603403e-05, "loss": 2.2579, "step": 1970 }, { "epoch": 1.0505960353603, "grad_norm": 0.2294921875, "learning_rate": 9.967761430321037e-05, "loss": 2.2386, "step": 1980 }, { "epoch": 1.0512657380123225, "grad_norm": 0.2021484375, "learning_rate": 9.966422593284474e-05, "loss": 2.3197, "step": 1990 }, { "epoch": 1.051935440664345, "grad_norm": 0.23046875, "learning_rate": 9.965056612241764e-05, "loss": 2.2259, "step": 2000 }, { "epoch": 1.0526051433163675, "grad_norm": 0.2001953125, "learning_rate": 9.963663494658681e-05, "loss": 2.2615, "step": 2010 }, { "epoch": 1.05327484596839, "grad_norm": 0.201171875, "learning_rate": 9.962243248149314e-05, "loss": 2.2512, "step": 2020 }, { "epoch": 1.0539445486204126, "grad_norm": 0.2216796875, "learning_rate": 9.960795880476029e-05, "loss": 2.2659, "step": 2030 }, { "epoch": 1.054614251272435, "grad_norm": 0.1962890625, "learning_rate": 9.95932139954942e-05, "loss": 2.3473, "step": 2040 }, { "epoch": 1.0552839539244576, "grad_norm": 0.236328125, "learning_rate": 9.95781981342827e-05, "loss": 2.3062, "step": 2050 }, { "epoch": 1.05595365657648, "grad_norm": 0.2158203125, "learning_rate": 9.956291130319506e-05, "loss": 2.3375, "step": 2060 }, { "epoch": 1.0566233592285026, "grad_norm": 0.216796875, "learning_rate": 9.95473535857815e-05, "loss": 2.304, "step": 2070 }, { "epoch": 1.057293061880525, "grad_norm": 0.2080078125, "learning_rate": 9.953152506707283e-05, "loss": 2.2981, "step": 2080 }, { "epoch": 1.0579627645325476, "grad_norm": 0.2119140625, "learning_rate": 9.951542583357986e-05, "loss": 2.3371, "step": 2090 }, { "epoch": 1.0586324671845702, "grad_norm": 0.2265625, "learning_rate": 9.949905597329306e-05, "loss": 2.4041, "step": 2100 }, { "epoch": 1.0593021698365925, "grad_norm": 0.212890625, "learning_rate": 9.948241557568196e-05, "loss": 2.2564, "step": 2110 }, { "epoch": 1.0599718724886151, "grad_norm": 0.1865234375, "learning_rate": 9.946550473169474e-05, "loss": 2.2427, "step": 2120 }, { "epoch": 1.0606415751406375, "grad_norm": 0.208984375, "learning_rate": 9.944832353375769e-05, "loss": 2.3322, "step": 2130 }, { "epoch": 1.06131127779266, "grad_norm": 0.1845703125, "learning_rate": 9.943087207577473e-05, "loss": 2.3154, "step": 2140 }, { "epoch": 1.0619809804446825, "grad_norm": 0.2275390625, "learning_rate": 9.941315045312689e-05, "loss": 2.2429, "step": 2150 }, { "epoch": 1.062650683096705, "grad_norm": 0.2255859375, "learning_rate": 9.939515876267178e-05, "loss": 2.2805, "step": 2160 }, { "epoch": 1.0633203857487277, "grad_norm": 0.19921875, "learning_rate": 9.937689710274308e-05, "loss": 2.4389, "step": 2170 }, { "epoch": 1.06399008840075, "grad_norm": 0.189453125, "learning_rate": 9.935836557314995e-05, "loss": 2.2942, "step": 2180 }, { "epoch": 1.0646597910527726, "grad_norm": 0.19140625, "learning_rate": 9.933956427517657e-05, "loss": 2.4409, "step": 2190 }, { "epoch": 1.065329493704795, "grad_norm": 0.2265625, "learning_rate": 9.932049331158153e-05, "loss": 2.3839, "step": 2200 }, { "epoch": 1.0659991963568176, "grad_norm": 0.20703125, "learning_rate": 9.930115278659727e-05, "loss": 2.2474, "step": 2210 }, { "epoch": 1.06666889900884, "grad_norm": 0.1923828125, "learning_rate": 9.928154280592952e-05, "loss": 2.232, "step": 2220 }, { "epoch": 1.0673386016608626, "grad_norm": 0.1923828125, "learning_rate": 9.92616634767567e-05, "loss": 2.2308, "step": 2230 }, { "epoch": 1.0680083043128852, "grad_norm": 0.197265625, "learning_rate": 9.924151490772942e-05, "loss": 2.3316, "step": 2240 }, { "epoch": 1.0686780069649076, "grad_norm": 0.228515625, "learning_rate": 9.922109720896973e-05, "loss": 2.2364, "step": 2250 }, { "epoch": 1.0693477096169302, "grad_norm": 0.197265625, "learning_rate": 9.92004104920707e-05, "loss": 2.2502, "step": 2260 }, { "epoch": 1.0700174122689525, "grad_norm": 0.1923828125, "learning_rate": 9.917945487009566e-05, "loss": 2.4647, "step": 2270 }, { "epoch": 1.0706871149209751, "grad_norm": 0.1875, "learning_rate": 9.915823045757765e-05, "loss": 2.257, "step": 2280 }, { "epoch": 1.0713568175729975, "grad_norm": 0.1904296875, "learning_rate": 9.913673737051882e-05, "loss": 2.2682, "step": 2290 }, { "epoch": 1.07202652022502, "grad_norm": 0.208984375, "learning_rate": 9.911497572638973e-05, "loss": 2.3932, "step": 2300 }, { "epoch": 1.0726962228770427, "grad_norm": 0.2275390625, "learning_rate": 9.909294564412874e-05, "loss": 2.2642, "step": 2310 }, { "epoch": 1.073365925529065, "grad_norm": 0.1865234375, "learning_rate": 9.907064724414139e-05, "loss": 2.2652, "step": 2320 }, { "epoch": 1.0740356281810877, "grad_norm": 0.1845703125, "learning_rate": 9.904808064829967e-05, "loss": 2.2997, "step": 2330 }, { "epoch": 1.07470533083311, "grad_norm": 0.2177734375, "learning_rate": 9.902524597994143e-05, "loss": 2.5363, "step": 2340 }, { "epoch": 1.0753750334851326, "grad_norm": 0.201171875, "learning_rate": 9.900214336386964e-05, "loss": 2.4563, "step": 2350 }, { "epoch": 1.076044736137155, "grad_norm": 0.2314453125, "learning_rate": 9.897877292635179e-05, "loss": 2.3237, "step": 2360 }, { "epoch": 1.0767144387891776, "grad_norm": 0.1796875, "learning_rate": 9.895513479511907e-05, "loss": 2.4093, "step": 2370 }, { "epoch": 1.0773841414412002, "grad_norm": 0.1845703125, "learning_rate": 9.893122909936583e-05, "loss": 2.3122, "step": 2380 }, { "epoch": 1.0780538440932226, "grad_norm": 0.18359375, "learning_rate": 9.890705596974875e-05, "loss": 2.2948, "step": 2390 }, { "epoch": 1.0787235467452452, "grad_norm": 0.1904296875, "learning_rate": 9.88826155383862e-05, "loss": 2.3163, "step": 2400 }, { "epoch": 1.0793932493972676, "grad_norm": 0.18359375, "learning_rate": 9.885790793885746e-05, "loss": 2.2554, "step": 2410 }, { "epoch": 1.0800629520492901, "grad_norm": 0.1640625, "learning_rate": 9.883293330620205e-05, "loss": 2.3188, "step": 2420 }, { "epoch": 1.0807326547013125, "grad_norm": 0.169921875, "learning_rate": 9.880769177691892e-05, "loss": 2.4599, "step": 2430 }, { "epoch": 1.0814023573533351, "grad_norm": 0.17578125, "learning_rate": 9.878218348896577e-05, "loss": 2.269, "step": 2440 }, { "epoch": 2.0000669702652023, "grad_norm": 0.1728515625, "learning_rate": 9.875640858175827e-05, "loss": 2.3692, "step": 2450 }, { "epoch": 2.000736672917225, "grad_norm": 0.173828125, "learning_rate": 9.87303671961693e-05, "loss": 2.2924, "step": 2460 }, { "epoch": 2.001406375569247, "grad_norm": 0.166015625, "learning_rate": 9.870405947452819e-05, "loss": 2.215, "step": 2470 }, { "epoch": 2.0020760782212697, "grad_norm": 0.158203125, "learning_rate": 9.86774855606199e-05, "loss": 2.2131, "step": 2480 }, { "epoch": 2.0027457808732922, "grad_norm": 0.1591796875, "learning_rate": 9.86506455996843e-05, "loss": 2.2224, "step": 2490 }, { "epoch": 2.003415483525315, "grad_norm": 0.1650390625, "learning_rate": 9.862353973841526e-05, "loss": 2.2056, "step": 2500 }, { "epoch": 2.0040851861773374, "grad_norm": 0.1533203125, "learning_rate": 9.859616812496008e-05, "loss": 2.4625, "step": 2510 }, { "epoch": 2.0047548888293596, "grad_norm": 0.1533203125, "learning_rate": 9.856853090891843e-05, "loss": 2.3478, "step": 2520 }, { "epoch": 2.005424591481382, "grad_norm": 0.1484375, "learning_rate": 9.854062824134159e-05, "loss": 2.2959, "step": 2530 }, { "epoch": 2.006094294133405, "grad_norm": 0.1435546875, "learning_rate": 9.851246027473173e-05, "loss": 2.2018, "step": 2540 }, { "epoch": 2.0067639967854274, "grad_norm": 0.15625, "learning_rate": 9.848402716304106e-05, "loss": 2.1618, "step": 2550 }, { "epoch": 2.00743369943745, "grad_norm": 0.205078125, "learning_rate": 9.845532906167083e-05, "loss": 2.2995, "step": 2560 }, { "epoch": 2.008103402089472, "grad_norm": 0.1474609375, "learning_rate": 9.842636612747069e-05, "loss": 2.2655, "step": 2570 }, { "epoch": 2.0087731047414947, "grad_norm": 0.140625, "learning_rate": 9.839713851873766e-05, "loss": 2.2814, "step": 2580 }, { "epoch": 2.0094428073935173, "grad_norm": 0.1357421875, "learning_rate": 9.836764639521539e-05, "loss": 2.3161, "step": 2590 }, { "epoch": 2.01011251004554, "grad_norm": 0.1357421875, "learning_rate": 9.833788991809323e-05, "loss": 2.1999, "step": 2600 }, { "epoch": 2.010782212697562, "grad_norm": 0.1337890625, "learning_rate": 9.830786925000533e-05, "loss": 2.2726, "step": 2610 }, { "epoch": 2.0114519153495847, "grad_norm": 0.1494140625, "learning_rate": 9.827758455502978e-05, "loss": 2.3824, "step": 2620 }, { "epoch": 2.0121216180016073, "grad_norm": 0.1298828125, "learning_rate": 9.824703599868776e-05, "loss": 2.2622, "step": 2630 }, { "epoch": 2.01279132065363, "grad_norm": 0.1298828125, "learning_rate": 9.821622374794253e-05, "loss": 2.2553, "step": 2640 }, { "epoch": 2.0134610233056525, "grad_norm": 0.1328125, "learning_rate": 9.81851479711986e-05, "loss": 2.3075, "step": 2650 }, { "epoch": 2.0141307259576746, "grad_norm": 0.1708984375, "learning_rate": 9.815380883830076e-05, "loss": 2.24, "step": 2660 }, { "epoch": 2.014800428609697, "grad_norm": 0.1435546875, "learning_rate": 9.812220652053318e-05, "loss": 2.3989, "step": 2670 }, { "epoch": 2.01547013126172, "grad_norm": 0.169921875, "learning_rate": 9.80903411906185e-05, "loss": 2.3223, "step": 2680 }, { "epoch": 2.0161398339137424, "grad_norm": 0.1240234375, "learning_rate": 9.805821302271685e-05, "loss": 2.3143, "step": 2690 }, { "epoch": 2.016809536565765, "grad_norm": 0.11669921875, "learning_rate": 9.802582219242484e-05, "loss": 2.2776, "step": 2700 }, { "epoch": 2.017479239217787, "grad_norm": 0.11669921875, "learning_rate": 9.799316887677471e-05, "loss": 2.2133, "step": 2710 }, { "epoch": 2.0181489418698098, "grad_norm": 0.24609375, "learning_rate": 9.796025325423334e-05, "loss": 2.4009, "step": 2720 }, { "epoch": 2.0188186445218324, "grad_norm": 0.11328125, "learning_rate": 9.792707550470122e-05, "loss": 2.2619, "step": 2730 }, { "epoch": 2.019488347173855, "grad_norm": 0.11181640625, "learning_rate": 9.789363580951145e-05, "loss": 2.3545, "step": 2740 }, { "epoch": 2.020158049825877, "grad_norm": 0.1103515625, "learning_rate": 9.785993435142891e-05, "loss": 2.3078, "step": 2750 }, { "epoch": 2.0208277524778997, "grad_norm": 0.10791015625, "learning_rate": 9.782597131464901e-05, "loss": 2.2504, "step": 2760 }, { "epoch": 2.0214974551299223, "grad_norm": 0.1064453125, "learning_rate": 9.779174688479693e-05, "loss": 2.3193, "step": 2770 }, { "epoch": 2.022167157781945, "grad_norm": 0.10888671875, "learning_rate": 9.775726124892646e-05, "loss": 2.2184, "step": 2780 }, { "epoch": 2.0228368604339675, "grad_norm": 0.11474609375, "learning_rate": 9.772251459551897e-05, "loss": 2.3344, "step": 2790 }, { "epoch": 2.0235065630859896, "grad_norm": 0.10693359375, "learning_rate": 9.768750711448249e-05, "loss": 2.1897, "step": 2800 }, { "epoch": 2.0241762657380122, "grad_norm": 0.19921875, "learning_rate": 9.765223899715054e-05, "loss": 2.3175, "step": 2810 }, { "epoch": 2.024845968390035, "grad_norm": 0.10107421875, "learning_rate": 9.761671043628124e-05, "loss": 2.2005, "step": 2820 }, { "epoch": 2.0255156710420574, "grad_norm": 0.10009765625, "learning_rate": 9.758092162605604e-05, "loss": 2.1851, "step": 2830 }, { "epoch": 2.02618537369408, "grad_norm": 0.10400390625, "learning_rate": 9.754487276207889e-05, "loss": 2.2614, "step": 2840 }, { "epoch": 2.026855076346102, "grad_norm": 0.1025390625, "learning_rate": 9.750856404137502e-05, "loss": 2.2961, "step": 2850 }, { "epoch": 2.0275247789981248, "grad_norm": 0.11962890625, "learning_rate": 9.747199566238991e-05, "loss": 2.2607, "step": 2860 }, { "epoch": 2.0281944816501474, "grad_norm": 0.10498046875, "learning_rate": 9.743516782498822e-05, "loss": 2.2181, "step": 2870 }, { "epoch": 2.02886418430217, "grad_norm": 0.1220703125, "learning_rate": 9.739808073045264e-05, "loss": 2.5639, "step": 2880 }, { "epoch": 2.029533886954192, "grad_norm": 0.10302734375, "learning_rate": 9.73607345814829e-05, "loss": 2.1937, "step": 2890 }, { "epoch": 2.0302035896062147, "grad_norm": 0.1025390625, "learning_rate": 9.732312958219453e-05, "loss": 2.1907, "step": 2900 }, { "epoch": 2.0308732922582373, "grad_norm": 0.10205078125, "learning_rate": 9.728526593811783e-05, "loss": 2.4198, "step": 2910 }, { "epoch": 2.03154299491026, "grad_norm": 0.11083984375, "learning_rate": 9.724714385619673e-05, "loss": 2.195, "step": 2920 }, { "epoch": 2.0322126975622825, "grad_norm": 0.109375, "learning_rate": 9.720876354478765e-05, "loss": 2.2016, "step": 2930 }, { "epoch": 2.0328824002143047, "grad_norm": 0.09912109375, "learning_rate": 9.717012521365836e-05, "loss": 2.3368, "step": 2940 }, { "epoch": 2.0335521028663273, "grad_norm": 0.10693359375, "learning_rate": 9.713122907398686e-05, "loss": 2.2595, "step": 2950 }, { "epoch": 2.03422180551835, "grad_norm": 0.23046875, "learning_rate": 9.709207533836016e-05, "loss": 2.2384, "step": 2960 }, { "epoch": 2.0348915081703725, "grad_norm": 0.10400390625, "learning_rate": 9.70526642207732e-05, "loss": 2.2645, "step": 2970 }, { "epoch": 2.035561210822395, "grad_norm": 0.10205078125, "learning_rate": 9.701299593662763e-05, "loss": 2.2913, "step": 2980 }, { "epoch": 2.036230913474417, "grad_norm": 0.103515625, "learning_rate": 9.697307070273062e-05, "loss": 2.2553, "step": 2990 }, { "epoch": 2.03690061612644, "grad_norm": 0.1015625, "learning_rate": 9.693288873729376e-05, "loss": 2.3177, "step": 3000 }, { "epoch": 2.0375703187784624, "grad_norm": 0.10546875, "learning_rate": 9.689245025993175e-05, "loss": 2.1987, "step": 3010 }, { "epoch": 2.038240021430485, "grad_norm": 0.10546875, "learning_rate": 9.68517554916613e-05, "loss": 2.2584, "step": 3020 }, { "epoch": 2.038909724082507, "grad_norm": 0.09912109375, "learning_rate": 9.681080465489983e-05, "loss": 2.2419, "step": 3030 }, { "epoch": 2.0395794267345297, "grad_norm": 0.09912109375, "learning_rate": 9.676959797346435e-05, "loss": 2.3091, "step": 3040 }, { "epoch": 2.0402491293865523, "grad_norm": 0.11669921875, "learning_rate": 9.672813567257017e-05, "loss": 2.2239, "step": 3050 }, { "epoch": 2.040918832038575, "grad_norm": 0.1376953125, "learning_rate": 9.668641797882969e-05, "loss": 2.2053, "step": 3060 }, { "epoch": 2.0415885346905975, "grad_norm": 0.11474609375, "learning_rate": 9.664444512025116e-05, "loss": 2.1763, "step": 3070 }, { "epoch": 2.0422582373426197, "grad_norm": 0.10595703125, "learning_rate": 9.660221732623744e-05, "loss": 2.3468, "step": 3080 }, { "epoch": 2.0429279399946423, "grad_norm": 0.1044921875, "learning_rate": 9.655973482758473e-05, "loss": 2.2388, "step": 3090 }, { "epoch": 2.043597642646665, "grad_norm": 0.158203125, "learning_rate": 9.651699785648135e-05, "loss": 2.1826, "step": 3100 }, { "epoch": 2.0442673452986875, "grad_norm": 0.185546875, "learning_rate": 9.647400664650638e-05, "loss": 2.1837, "step": 3110 }, { "epoch": 2.04493704795071, "grad_norm": 0.1728515625, "learning_rate": 9.643076143262851e-05, "loss": 2.285, "step": 3120 }, { "epoch": 2.0456067506027322, "grad_norm": 0.16015625, "learning_rate": 9.638726245120466e-05, "loss": 2.2987, "step": 3130 }, { "epoch": 2.046276453254755, "grad_norm": 0.181640625, "learning_rate": 9.634350993997871e-05, "loss": 2.3399, "step": 3140 }, { "epoch": 2.0469461559067774, "grad_norm": 0.71484375, "learning_rate": 9.629950413808022e-05, "loss": 2.2561, "step": 3150 }, { "epoch": 2.0476158585588, "grad_norm": 3.71875, "learning_rate": 9.62552452860231e-05, "loss": 2.8187, "step": 3160 }, { "epoch": 2.0482855612108226, "grad_norm": 1.8828125, "learning_rate": 9.621073362570432e-05, "loss": 2.3827, "step": 3170 }, { "epoch": 2.0489552638628448, "grad_norm": 0.5625, "learning_rate": 9.616596940040257e-05, "loss": 2.3021, "step": 3180 }, { "epoch": 2.0496249665148674, "grad_norm": 0.51171875, "learning_rate": 9.612095285477694e-05, "loss": 2.2065, "step": 3190 }, { "epoch": 2.05029466916689, "grad_norm": 0.62109375, "learning_rate": 9.607568423486558e-05, "loss": 2.1997, "step": 3200 }, { "epoch": 2.0509643718189126, "grad_norm": 2.0, "learning_rate": 9.603016378808432e-05, "loss": 2.8713, "step": 3210 }, { "epoch": 2.0516340744709347, "grad_norm": 1.1328125, "learning_rate": 9.59843917632254e-05, "loss": 2.2858, "step": 3220 }, { "epoch": 2.0523037771229573, "grad_norm": 0.44921875, "learning_rate": 9.593836841045602e-05, "loss": 2.2142, "step": 3230 }, { "epoch": 2.05297347977498, "grad_norm": 0.73046875, "learning_rate": 9.589209398131706e-05, "loss": 2.2057, "step": 3240 }, { "epoch": 2.0536431824270025, "grad_norm": 2.09375, "learning_rate": 9.584556872872159e-05, "loss": 2.2197, "step": 3250 }, { "epoch": 2.054312885079025, "grad_norm": 156.0, "learning_rate": 9.579879290695364e-05, "loss": 2.265, "step": 3260 }, { "epoch": 2.0549825877310472, "grad_norm": 1.1640625, "learning_rate": 9.575176677166667e-05, "loss": 2.2266, "step": 3270 }, { "epoch": 2.05565229038307, "grad_norm": 2.0625, "learning_rate": 9.570449057988222e-05, "loss": 2.2296, "step": 3280 }, { "epoch": 2.0563219930350924, "grad_norm": 1.6796875, "learning_rate": 9.565696458998858e-05, "loss": 2.2136, "step": 3290 }, { "epoch": 2.056991695687115, "grad_norm": 0.515625, "learning_rate": 9.560918906173923e-05, "loss": 2.2098, "step": 3300 }, { "epoch": 2.0576613983391376, "grad_norm": 0.421875, "learning_rate": 9.556116425625159e-05, "loss": 2.4203, "step": 3310 }, { "epoch": 2.05833110099116, "grad_norm": 0.55078125, "learning_rate": 9.551289043600542e-05, "loss": 2.1719, "step": 3320 }, { "epoch": 2.0590008036431824, "grad_norm": 0.2373046875, "learning_rate": 9.546436786484155e-05, "loss": 2.2131, "step": 3330 }, { "epoch": 2.059670506295205, "grad_norm": 0.240234375, "learning_rate": 9.541559680796029e-05, "loss": 2.2326, "step": 3340 }, { "epoch": 2.0603402089472276, "grad_norm": 0.2275390625, "learning_rate": 9.536657753192011e-05, "loss": 2.2091, "step": 3350 }, { "epoch": 2.0610099115992497, "grad_norm": 0.228515625, "learning_rate": 9.531731030463607e-05, "loss": 2.2013, "step": 3360 }, { "epoch": 2.0616796142512723, "grad_norm": 0.25, "learning_rate": 9.526779539537845e-05, "loss": 2.1891, "step": 3370 }, { "epoch": 2.062349316903295, "grad_norm": 0.259765625, "learning_rate": 9.52180330747712e-05, "loss": 2.2255, "step": 3380 }, { "epoch": 2.0630190195553175, "grad_norm": 0.322265625, "learning_rate": 9.516802361479056e-05, "loss": 2.1719, "step": 3390 }, { "epoch": 2.06368872220734, "grad_norm": 0.2412109375, "learning_rate": 9.511776728876341e-05, "loss": 2.1987, "step": 3400 }, { "epoch": 2.0643584248593623, "grad_norm": 0.2412109375, "learning_rate": 9.506726437136599e-05, "loss": 2.2044, "step": 3410 }, { "epoch": 2.065028127511385, "grad_norm": 0.224609375, "learning_rate": 9.501651513862222e-05, "loss": 2.223, "step": 3420 }, { "epoch": 2.0656978301634075, "grad_norm": 0.2041015625, "learning_rate": 9.496551986790225e-05, "loss": 2.2268, "step": 3430 }, { "epoch": 2.06636753281543, "grad_norm": 0.197265625, "learning_rate": 9.4914278837921e-05, "loss": 2.2123, "step": 3440 }, { "epoch": 2.0670372354674527, "grad_norm": 0.2177734375, "learning_rate": 9.486279232873654e-05, "loss": 2.1541, "step": 3450 }, { "epoch": 2.067706938119475, "grad_norm": 0.220703125, "learning_rate": 9.481106062174863e-05, "loss": 2.1894, "step": 3460 }, { "epoch": 2.0683766407714974, "grad_norm": 0.2421875, "learning_rate": 9.475908399969718e-05, "loss": 2.2281, "step": 3470 }, { "epoch": 2.06904634342352, "grad_norm": 0.212890625, "learning_rate": 9.470686274666065e-05, "loss": 2.1997, "step": 3480 }, { "epoch": 2.0697160460755426, "grad_norm": 0.21875, "learning_rate": 9.465439714805455e-05, "loss": 2.214, "step": 3490 }, { "epoch": 2.0703857487275648, "grad_norm": 0.24609375, "learning_rate": 9.460168749062985e-05, "loss": 2.2174, "step": 3500 } ], "logging_steps": 10, "max_steps": 14932, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "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": 1.038291318521856e+19, "train_batch_size": 1, "trial_name": null, "trial_params": null }