| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.99695843190267, |
| "eval_steps": 500, |
| "global_step": 1107, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.027036160865157147, |
| "grad_norm": 3.6503875872895457, |
| "learning_rate": 1e-06, |
| "loss": 0.7953, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.054072321730314295, |
| "grad_norm": 1.7569721856122498, |
| "learning_rate": 1e-06, |
| "loss": 0.7005, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.08110848259547145, |
| "grad_norm": 1.6809598137883086, |
| "learning_rate": 1e-06, |
| "loss": 0.6698, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.10814464346062859, |
| "grad_norm": 1.9898234262086616, |
| "learning_rate": 1e-06, |
| "loss": 0.6586, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.13518080432578575, |
| "grad_norm": 1.370909250299239, |
| "learning_rate": 1e-06, |
| "loss": 0.6425, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.1622169651909429, |
| "grad_norm": 1.4635655217944983, |
| "learning_rate": 1e-06, |
| "loss": 0.6363, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.18925312605610004, |
| "grad_norm": 1.630786915276097, |
| "learning_rate": 1e-06, |
| "loss": 0.6315, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.21628928692125718, |
| "grad_norm": 1.4646243743445129, |
| "learning_rate": 1e-06, |
| "loss": 0.623, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.24332544778641432, |
| "grad_norm": 1.8839329924596884, |
| "learning_rate": 1e-06, |
| "loss": 0.6197, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.2703616086515715, |
| "grad_norm": 1.654871462446455, |
| "learning_rate": 1e-06, |
| "loss": 0.6221, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.29739776951672864, |
| "grad_norm": 1.3904449493880144, |
| "learning_rate": 1e-06, |
| "loss": 0.6192, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.3244339303818858, |
| "grad_norm": 1.5013994250001494, |
| "learning_rate": 1e-06, |
| "loss": 0.6142, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.3514700912470429, |
| "grad_norm": 1.6179719292710375, |
| "learning_rate": 1e-06, |
| "loss": 0.6159, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.37850625211220007, |
| "grad_norm": 1.539789253137618, |
| "learning_rate": 1e-06, |
| "loss": 0.6155, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.4055424129773572, |
| "grad_norm": 1.9347475550306212, |
| "learning_rate": 1e-06, |
| "loss": 0.6064, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.43257857384251436, |
| "grad_norm": 1.456515522014398, |
| "learning_rate": 1e-06, |
| "loss": 0.6078, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.4596147347076715, |
| "grad_norm": 1.6552492368307141, |
| "learning_rate": 1e-06, |
| "loss": 0.6077, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.48665089557282865, |
| "grad_norm": 1.3788390074752856, |
| "learning_rate": 1e-06, |
| "loss": 0.6071, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.5136870564379858, |
| "grad_norm": 1.364845844005561, |
| "learning_rate": 1e-06, |
| "loss": 0.6033, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.540723217303143, |
| "grad_norm": 1.80889758488186, |
| "learning_rate": 1e-06, |
| "loss": 0.601, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.5677593781683001, |
| "grad_norm": 1.6680485988791665, |
| "learning_rate": 1e-06, |
| "loss": 0.6026, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.5947955390334573, |
| "grad_norm": 1.5916624635090106, |
| "learning_rate": 1e-06, |
| "loss": 0.598, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.6218316998986144, |
| "grad_norm": 1.4187846256482866, |
| "learning_rate": 1e-06, |
| "loss": 0.5995, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.6488678607637716, |
| "grad_norm": 1.6789789212917976, |
| "learning_rate": 1e-06, |
| "loss": 0.5962, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.6759040216289287, |
| "grad_norm": 1.940377894908514, |
| "learning_rate": 1e-06, |
| "loss": 0.5882, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.7029401824940859, |
| "grad_norm": 1.4614866124657333, |
| "learning_rate": 1e-06, |
| "loss": 0.5916, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.729976343359243, |
| "grad_norm": 1.385546402101677, |
| "learning_rate": 1e-06, |
| "loss": 0.5953, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.7570125042244001, |
| "grad_norm": 1.594020164827515, |
| "learning_rate": 1e-06, |
| "loss": 0.5905, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.7840486650895573, |
| "grad_norm": 1.4916422621093772, |
| "learning_rate": 1e-06, |
| "loss": 0.5903, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.8110848259547144, |
| "grad_norm": 1.4814008971564634, |
| "learning_rate": 1e-06, |
| "loss": 0.5896, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.8381209868198716, |
| "grad_norm": 1.342321531719104, |
| "learning_rate": 1e-06, |
| "loss": 0.5866, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.8651571476850287, |
| "grad_norm": 1.3717754933291688, |
| "learning_rate": 1e-06, |
| "loss": 0.5856, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.8921933085501859, |
| "grad_norm": 1.524506169690246, |
| "learning_rate": 1e-06, |
| "loss": 0.5865, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.919229469415343, |
| "grad_norm": 1.3828833626570336, |
| "learning_rate": 1e-06, |
| "loss": 0.5934, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.9462656302805001, |
| "grad_norm": 1.789960289867802, |
| "learning_rate": 1e-06, |
| "loss": 0.5848, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.9733017911456573, |
| "grad_norm": 1.8633160327770453, |
| "learning_rate": 1e-06, |
| "loss": 0.5846, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.9976343359242987, |
| "eval_loss": 0.07291737198829651, |
| "eval_runtime": 382.2719, |
| "eval_samples_per_second": 26.068, |
| "eval_steps_per_second": 0.408, |
| "step": 369 |
| }, |
| { |
| "epoch": 1.0023656640757013, |
| "grad_norm": 2.056752608460438, |
| "learning_rate": 1e-06, |
| "loss": 0.5848, |
| "step": 370 |
| }, |
| { |
| "epoch": 1.0294018249408583, |
| "grad_norm": 2.0382350808065524, |
| "learning_rate": 1e-06, |
| "loss": 0.555, |
| "step": 380 |
| }, |
| { |
| "epoch": 1.0564379858060156, |
| "grad_norm": 1.4275159131594684, |
| "learning_rate": 1e-06, |
| "loss": 0.5544, |
| "step": 390 |
| }, |
| { |
| "epoch": 1.0834741466711728, |
| "grad_norm": 1.534877702230249, |
| "learning_rate": 1e-06, |
| "loss": 0.5568, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.1105103075363298, |
| "grad_norm": 1.5744230812862943, |
| "learning_rate": 1e-06, |
| "loss": 0.554, |
| "step": 410 |
| }, |
| { |
| "epoch": 1.1375464684014869, |
| "grad_norm": 1.7333588090630478, |
| "learning_rate": 1e-06, |
| "loss": 0.5558, |
| "step": 420 |
| }, |
| { |
| "epoch": 1.1645826292666441, |
| "grad_norm": 1.5892673969061217, |
| "learning_rate": 1e-06, |
| "loss": 0.5621, |
| "step": 430 |
| }, |
| { |
| "epoch": 1.1916187901318014, |
| "grad_norm": 1.507039269649461, |
| "learning_rate": 1e-06, |
| "loss": 0.5522, |
| "step": 440 |
| }, |
| { |
| "epoch": 1.2186549509969584, |
| "grad_norm": 1.7746833778464899, |
| "learning_rate": 1e-06, |
| "loss": 0.5569, |
| "step": 450 |
| }, |
| { |
| "epoch": 1.2456911118621157, |
| "grad_norm": 2.3175308464303446, |
| "learning_rate": 1e-06, |
| "loss": 0.5582, |
| "step": 460 |
| }, |
| { |
| "epoch": 1.2727272727272727, |
| "grad_norm": 1.5903292287916326, |
| "learning_rate": 1e-06, |
| "loss": 0.558, |
| "step": 470 |
| }, |
| { |
| "epoch": 1.29976343359243, |
| "grad_norm": 1.889265359568593, |
| "learning_rate": 1e-06, |
| "loss": 0.5534, |
| "step": 480 |
| }, |
| { |
| "epoch": 1.326799594457587, |
| "grad_norm": 1.859329532891828, |
| "learning_rate": 1e-06, |
| "loss": 0.553, |
| "step": 490 |
| }, |
| { |
| "epoch": 1.3538357553227442, |
| "grad_norm": 1.5185821864198052, |
| "learning_rate": 1e-06, |
| "loss": 0.5557, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.3808719161879013, |
| "grad_norm": 1.5352745363271778, |
| "learning_rate": 1e-06, |
| "loss": 0.5551, |
| "step": 510 |
| }, |
| { |
| "epoch": 1.4079080770530585, |
| "grad_norm": 1.4967747324416663, |
| "learning_rate": 1e-06, |
| "loss": 0.557, |
| "step": 520 |
| }, |
| { |
| "epoch": 1.4349442379182156, |
| "grad_norm": 1.5838303148730983, |
| "learning_rate": 1e-06, |
| "loss": 0.5494, |
| "step": 530 |
| }, |
| { |
| "epoch": 1.4619803987833728, |
| "grad_norm": 1.589952312711665, |
| "learning_rate": 1e-06, |
| "loss": 0.554, |
| "step": 540 |
| }, |
| { |
| "epoch": 1.4890165596485299, |
| "grad_norm": 1.4884571279344156, |
| "learning_rate": 1e-06, |
| "loss": 0.5511, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.5160527205136871, |
| "grad_norm": 1.6807740089852297, |
| "learning_rate": 1e-06, |
| "loss": 0.553, |
| "step": 560 |
| }, |
| { |
| "epoch": 1.5430888813788441, |
| "grad_norm": 1.5738922116096086, |
| "learning_rate": 1e-06, |
| "loss": 0.5474, |
| "step": 570 |
| }, |
| { |
| "epoch": 1.5701250422440014, |
| "grad_norm": 1.2816702873452306, |
| "learning_rate": 1e-06, |
| "loss": 0.5496, |
| "step": 580 |
| }, |
| { |
| "epoch": 1.5971612031091587, |
| "grad_norm": 1.387328243437945, |
| "learning_rate": 1e-06, |
| "loss": 0.553, |
| "step": 590 |
| }, |
| { |
| "epoch": 1.6241973639743157, |
| "grad_norm": 1.4440752863559743, |
| "learning_rate": 1e-06, |
| "loss": 0.5478, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.6512335248394727, |
| "grad_norm": 1.5659569703518923, |
| "learning_rate": 1e-06, |
| "loss": 0.5463, |
| "step": 610 |
| }, |
| { |
| "epoch": 1.67826968570463, |
| "grad_norm": 1.5693078814309174, |
| "learning_rate": 1e-06, |
| "loss": 0.5504, |
| "step": 620 |
| }, |
| { |
| "epoch": 1.7053058465697872, |
| "grad_norm": 1.4566281165614279, |
| "learning_rate": 1e-06, |
| "loss": 0.5553, |
| "step": 630 |
| }, |
| { |
| "epoch": 1.7323420074349443, |
| "grad_norm": 1.513761932224734, |
| "learning_rate": 1e-06, |
| "loss": 0.5537, |
| "step": 640 |
| }, |
| { |
| "epoch": 1.7593781683001013, |
| "grad_norm": 1.748694998470595, |
| "learning_rate": 1e-06, |
| "loss": 0.551, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.7864143291652586, |
| "grad_norm": 1.4391703049392393, |
| "learning_rate": 1e-06, |
| "loss": 0.5471, |
| "step": 660 |
| }, |
| { |
| "epoch": 1.8134504900304158, |
| "grad_norm": 1.5593553566151679, |
| "learning_rate": 1e-06, |
| "loss": 0.5486, |
| "step": 670 |
| }, |
| { |
| "epoch": 1.8404866508955728, |
| "grad_norm": 1.4276061302519432, |
| "learning_rate": 1e-06, |
| "loss": 0.5515, |
| "step": 680 |
| }, |
| { |
| "epoch": 1.8675228117607299, |
| "grad_norm": 1.6599988060375446, |
| "learning_rate": 1e-06, |
| "loss": 0.5485, |
| "step": 690 |
| }, |
| { |
| "epoch": 1.8945589726258871, |
| "grad_norm": 1.5056855484458003, |
| "learning_rate": 1e-06, |
| "loss": 0.5509, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.9215951334910444, |
| "grad_norm": 1.5501347776721288, |
| "learning_rate": 1e-06, |
| "loss": 0.5554, |
| "step": 710 |
| }, |
| { |
| "epoch": 1.9486312943562014, |
| "grad_norm": 1.4591555519225876, |
| "learning_rate": 1e-06, |
| "loss": 0.5472, |
| "step": 720 |
| }, |
| { |
| "epoch": 1.9756674552213584, |
| "grad_norm": 1.5262930847918643, |
| "learning_rate": 1e-06, |
| "loss": 0.5426, |
| "step": 730 |
| }, |
| { |
| "epoch": 1.9972963839134843, |
| "eval_loss": 0.07129764556884766, |
| "eval_runtime": 379.8155, |
| "eval_samples_per_second": 26.236, |
| "eval_steps_per_second": 0.411, |
| "step": 738 |
| }, |
| { |
| "epoch": 2.0047313281514025, |
| "grad_norm": 1.8995776642671698, |
| "learning_rate": 1e-06, |
| "loss": 0.5424, |
| "step": 740 |
| }, |
| { |
| "epoch": 2.0317674890165596, |
| "grad_norm": 1.4844116810041432, |
| "learning_rate": 1e-06, |
| "loss": 0.5105, |
| "step": 750 |
| }, |
| { |
| "epoch": 2.0588036498817166, |
| "grad_norm": 1.585715405294997, |
| "learning_rate": 1e-06, |
| "loss": 0.5195, |
| "step": 760 |
| }, |
| { |
| "epoch": 2.085839810746874, |
| "grad_norm": 1.5428298988588043, |
| "learning_rate": 1e-06, |
| "loss": 0.5162, |
| "step": 770 |
| }, |
| { |
| "epoch": 2.112875971612031, |
| "grad_norm": 1.6719458135129328, |
| "learning_rate": 1e-06, |
| "loss": 0.5156, |
| "step": 780 |
| }, |
| { |
| "epoch": 2.139912132477188, |
| "grad_norm": 1.7452805074588629, |
| "learning_rate": 1e-06, |
| "loss": 0.514, |
| "step": 790 |
| }, |
| { |
| "epoch": 2.1669482933423456, |
| "grad_norm": 1.6858208907796273, |
| "learning_rate": 1e-06, |
| "loss": 0.5196, |
| "step": 800 |
| }, |
| { |
| "epoch": 2.1939844542075027, |
| "grad_norm": 1.7264763784102355, |
| "learning_rate": 1e-06, |
| "loss": 0.5188, |
| "step": 810 |
| }, |
| { |
| "epoch": 2.2210206150726597, |
| "grad_norm": 1.6080058488464875, |
| "learning_rate": 1e-06, |
| "loss": 0.518, |
| "step": 820 |
| }, |
| { |
| "epoch": 2.2480567759378167, |
| "grad_norm": 1.5056848901498003, |
| "learning_rate": 1e-06, |
| "loss": 0.5217, |
| "step": 830 |
| }, |
| { |
| "epoch": 2.2750929368029738, |
| "grad_norm": 1.453297884431512, |
| "learning_rate": 1e-06, |
| "loss": 0.5206, |
| "step": 840 |
| }, |
| { |
| "epoch": 2.3021290976681312, |
| "grad_norm": 1.3502061562893424, |
| "learning_rate": 1e-06, |
| "loss": 0.5142, |
| "step": 850 |
| }, |
| { |
| "epoch": 2.3291652585332883, |
| "grad_norm": 1.9113035803831369, |
| "learning_rate": 1e-06, |
| "loss": 0.5187, |
| "step": 860 |
| }, |
| { |
| "epoch": 2.3562014193984453, |
| "grad_norm": 1.6114663336658883, |
| "learning_rate": 1e-06, |
| "loss": 0.5239, |
| "step": 870 |
| }, |
| { |
| "epoch": 2.3832375802636028, |
| "grad_norm": 1.5823476098768134, |
| "learning_rate": 1e-06, |
| "loss": 0.5208, |
| "step": 880 |
| }, |
| { |
| "epoch": 2.41027374112876, |
| "grad_norm": 1.6865550821087416, |
| "learning_rate": 1e-06, |
| "loss": 0.518, |
| "step": 890 |
| }, |
| { |
| "epoch": 2.437309901993917, |
| "grad_norm": 1.4118433994151605, |
| "learning_rate": 1e-06, |
| "loss": 0.5189, |
| "step": 900 |
| }, |
| { |
| "epoch": 2.464346062859074, |
| "grad_norm": 1.7616393267246282, |
| "learning_rate": 1e-06, |
| "loss": 0.5159, |
| "step": 910 |
| }, |
| { |
| "epoch": 2.4913822237242313, |
| "grad_norm": 1.6619254134657746, |
| "learning_rate": 1e-06, |
| "loss": 0.5155, |
| "step": 920 |
| }, |
| { |
| "epoch": 2.5184183845893884, |
| "grad_norm": 2.005467952797637, |
| "learning_rate": 1e-06, |
| "loss": 0.5132, |
| "step": 930 |
| }, |
| { |
| "epoch": 2.5454545454545454, |
| "grad_norm": 1.6751622324469126, |
| "learning_rate": 1e-06, |
| "loss": 0.5222, |
| "step": 940 |
| }, |
| { |
| "epoch": 2.5724907063197024, |
| "grad_norm": 1.5513069018510968, |
| "learning_rate": 1e-06, |
| "loss": 0.5146, |
| "step": 950 |
| }, |
| { |
| "epoch": 2.59952686718486, |
| "grad_norm": 1.439704052556968, |
| "learning_rate": 1e-06, |
| "loss": 0.5223, |
| "step": 960 |
| }, |
| { |
| "epoch": 2.626563028050017, |
| "grad_norm": 1.5280893876048505, |
| "learning_rate": 1e-06, |
| "loss": 0.5186, |
| "step": 970 |
| }, |
| { |
| "epoch": 2.653599188915174, |
| "grad_norm": 1.5581163726315426, |
| "learning_rate": 1e-06, |
| "loss": 0.5233, |
| "step": 980 |
| }, |
| { |
| "epoch": 2.6806353497803315, |
| "grad_norm": 1.455068563258345, |
| "learning_rate": 1e-06, |
| "loss": 0.5194, |
| "step": 990 |
| }, |
| { |
| "epoch": 2.7076715106454885, |
| "grad_norm": 1.524690245630394, |
| "learning_rate": 1e-06, |
| "loss": 0.5116, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.7347076715106455, |
| "grad_norm": 1.8005325743206846, |
| "learning_rate": 1e-06, |
| "loss": 0.5231, |
| "step": 1010 |
| }, |
| { |
| "epoch": 2.7617438323758026, |
| "grad_norm": 1.6922089883418125, |
| "learning_rate": 1e-06, |
| "loss": 0.5226, |
| "step": 1020 |
| }, |
| { |
| "epoch": 2.7887799932409596, |
| "grad_norm": 1.5115079161070315, |
| "learning_rate": 1e-06, |
| "loss": 0.5194, |
| "step": 1030 |
| }, |
| { |
| "epoch": 2.815816154106117, |
| "grad_norm": 1.606885984413175, |
| "learning_rate": 1e-06, |
| "loss": 0.5181, |
| "step": 1040 |
| }, |
| { |
| "epoch": 2.842852314971274, |
| "grad_norm": 1.6825309137829672, |
| "learning_rate": 1e-06, |
| "loss": 0.5187, |
| "step": 1050 |
| }, |
| { |
| "epoch": 2.869888475836431, |
| "grad_norm": 1.3724424151336712, |
| "learning_rate": 1e-06, |
| "loss": 0.5134, |
| "step": 1060 |
| }, |
| { |
| "epoch": 2.8969246367015886, |
| "grad_norm": 1.4408482381758962, |
| "learning_rate": 1e-06, |
| "loss": 0.523, |
| "step": 1070 |
| }, |
| { |
| "epoch": 2.9239607975667457, |
| "grad_norm": 1.6112080385104113, |
| "learning_rate": 1e-06, |
| "loss": 0.5169, |
| "step": 1080 |
| }, |
| { |
| "epoch": 2.9509969584319027, |
| "grad_norm": 1.5431444104863004, |
| "learning_rate": 1e-06, |
| "loss": 0.52, |
| "step": 1090 |
| }, |
| { |
| "epoch": 2.9780331192970597, |
| "grad_norm": 1.4847130780309508, |
| "learning_rate": 1e-06, |
| "loss": 0.5185, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.99695843190267, |
| "eval_loss": 0.07088814675807953, |
| "eval_runtime": 382.8031, |
| "eval_samples_per_second": 26.032, |
| "eval_steps_per_second": 0.408, |
| "step": 1107 |
| }, |
| { |
| "epoch": 2.99695843190267, |
| "step": 1107, |
| "total_flos": 1854056851046400.0, |
| "train_loss": 0.561927645079381, |
| "train_runtime": 63217.8679, |
| "train_samples_per_second": 8.985, |
| "train_steps_per_second": 0.018 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1107, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1854056851046400.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
