|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.99968, |
|
"eval_steps": 300, |
|
"global_step": 781, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 51.5, |
|
"learning_rate": 6.329113924050633e-09, |
|
"logits/generated": 0.1862616091966629, |
|
"logits/real": -0.24196800589561462, |
|
"logps/generated": -538.626953125, |
|
"logps/real": -411.1020202636719, |
|
"loss": 0.6931, |
|
"rewards/accuracies": 0.0, |
|
"rewards/generated": 0.0, |
|
"rewards/margins": 0.0, |
|
"rewards/real": 0.0, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 56.0, |
|
"learning_rate": 3.1645569620253166e-08, |
|
"logits/generated": 0.348865807056427, |
|
"logits/real": -0.1247345358133316, |
|
"logps/generated": -534.8592529296875, |
|
"logps/real": -401.95263671875, |
|
"loss": 0.7116, |
|
"rewards/accuracies": 0.34375, |
|
"rewards/generated": 0.018089786171913147, |
|
"rewards/margins": -0.02859283983707428, |
|
"rewards/real": -0.010503048077225685, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 52.25, |
|
"learning_rate": 6.329113924050633e-08, |
|
"logits/generated": 0.37161511182785034, |
|
"logits/real": -0.12119238078594208, |
|
"logps/generated": -555.2481689453125, |
|
"logps/real": -418.7288513183594, |
|
"loss": 0.707, |
|
"rewards/accuracies": 0.5249999761581421, |
|
"rewards/generated": -0.016772981733083725, |
|
"rewards/margins": -0.006616026163101196, |
|
"rewards/real": -0.02338900789618492, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 50.75, |
|
"learning_rate": 9.49367088607595e-08, |
|
"logits/generated": 0.3857704699039459, |
|
"logits/real": -0.10262864828109741, |
|
"logps/generated": -564.0440673828125, |
|
"logps/real": -434.94561767578125, |
|
"loss": 0.6843, |
|
"rewards/accuracies": 0.550000011920929, |
|
"rewards/generated": -0.02160583809018135, |
|
"rewards/margins": 0.07228925079107285, |
|
"rewards/real": 0.05068342760205269, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 52.5, |
|
"learning_rate": 1.2658227848101266e-07, |
|
"logits/generated": 0.3662206530570984, |
|
"logits/real": -0.06378354132175446, |
|
"logps/generated": -558.4364013671875, |
|
"logps/real": -439.30816650390625, |
|
"loss": 0.7284, |
|
"rewards/accuracies": 0.4375, |
|
"rewards/generated": 0.002506026765331626, |
|
"rewards/margins": -0.03106103278696537, |
|
"rewards/real": -0.028555002063512802, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 53.25, |
|
"learning_rate": 1.5822784810126582e-07, |
|
"logits/generated": 0.39896726608276367, |
|
"logits/real": -0.12075452506542206, |
|
"logps/generated": -565.0833740234375, |
|
"logps/real": -439.34344482421875, |
|
"loss": 0.708, |
|
"rewards/accuracies": 0.48750001192092896, |
|
"rewards/generated": 0.013953039422631264, |
|
"rewards/margins": -0.00905968714505434, |
|
"rewards/real": 0.004893349949270487, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 54.75, |
|
"learning_rate": 1.89873417721519e-07, |
|
"logits/generated": 0.36376339197158813, |
|
"logits/real": -0.15110069513320923, |
|
"logps/generated": -563.3927612304688, |
|
"logps/real": -444.236083984375, |
|
"loss": 0.6823, |
|
"rewards/accuracies": 0.5874999761581421, |
|
"rewards/generated": -0.03245529532432556, |
|
"rewards/margins": 0.041769616305828094, |
|
"rewards/real": 0.009314320981502533, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 53.0, |
|
"learning_rate": 2.2151898734177212e-07, |
|
"logits/generated": 0.3546258807182312, |
|
"logits/real": -0.08531169593334198, |
|
"logps/generated": -567.1099853515625, |
|
"logps/real": -426.37530517578125, |
|
"loss": 0.6792, |
|
"rewards/accuracies": 0.5249999761581421, |
|
"rewards/generated": -0.04005850851535797, |
|
"rewards/margins": 0.062456123530864716, |
|
"rewards/real": 0.022397616878151894, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 54.0, |
|
"learning_rate": 2.5316455696202533e-07, |
|
"logits/generated": 0.3955411911010742, |
|
"logits/real": -0.09925423562526703, |
|
"logps/generated": -584.2412719726562, |
|
"logps/real": -481.74151611328125, |
|
"loss": 0.7088, |
|
"rewards/accuracies": 0.518750011920929, |
|
"rewards/generated": 0.002044306369498372, |
|
"rewards/margins": 0.014470143243670464, |
|
"rewards/real": 0.016514450311660767, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 49.75, |
|
"learning_rate": 2.848101265822785e-07, |
|
"logits/generated": 0.3811967372894287, |
|
"logits/real": -0.10687746107578278, |
|
"logps/generated": -589.9839477539062, |
|
"logps/real": -475.3309020996094, |
|
"loss": 0.6826, |
|
"rewards/accuracies": 0.518750011920929, |
|
"rewards/generated": -0.01990634948015213, |
|
"rewards/margins": 0.01696391962468624, |
|
"rewards/real": -0.0029424328822642565, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 48.0, |
|
"learning_rate": 3.1645569620253163e-07, |
|
"logits/generated": 0.36268311738967896, |
|
"logits/real": -0.13508598506450653, |
|
"logps/generated": -559.43505859375, |
|
"logps/real": -434.68145751953125, |
|
"loss": 0.6677, |
|
"rewards/accuracies": 0.581250011920929, |
|
"rewards/generated": -0.050643421709537506, |
|
"rewards/margins": 0.07995937764644623, |
|
"rewards/real": 0.029315968975424767, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 54.25, |
|
"learning_rate": 3.481012658227848e-07, |
|
"logits/generated": 0.3685925602912903, |
|
"logits/real": -0.11418122053146362, |
|
"logps/generated": -591.776611328125, |
|
"logps/real": -450.0094299316406, |
|
"loss": 0.6876, |
|
"rewards/accuracies": 0.5687500238418579, |
|
"rewards/generated": -0.06491654366254807, |
|
"rewards/margins": 0.06822647154331207, |
|
"rewards/real": 0.0033099246211349964, |
|
"step": 55 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 52.0, |
|
"learning_rate": 3.79746835443038e-07, |
|
"logits/generated": 0.39532652497291565, |
|
"logits/real": -0.08878536522388458, |
|
"logps/generated": -570.0611572265625, |
|
"logps/real": -456.876708984375, |
|
"loss": 0.6667, |
|
"rewards/accuracies": 0.606249988079071, |
|
"rewards/generated": -0.08224920183420181, |
|
"rewards/margins": 0.05731937289237976, |
|
"rewards/real": -0.02492983639240265, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 48.0, |
|
"learning_rate": 4.1139240506329117e-07, |
|
"logits/generated": 0.3644171953201294, |
|
"logits/real": -0.11488844454288483, |
|
"logps/generated": -542.10498046875, |
|
"logps/real": -441.2933044433594, |
|
"loss": 0.6619, |
|
"rewards/accuracies": 0.581250011920929, |
|
"rewards/generated": -0.07368786633014679, |
|
"rewards/margins": 0.10115363448858261, |
|
"rewards/real": 0.02746577560901642, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 48.25, |
|
"learning_rate": 4.4303797468354424e-07, |
|
"logits/generated": 0.3911452889442444, |
|
"logits/real": -0.15676851570606232, |
|
"logps/generated": -570.861083984375, |
|
"logps/real": -487.00848388671875, |
|
"loss": 0.6486, |
|
"rewards/accuracies": 0.637499988079071, |
|
"rewards/generated": -0.1312028169631958, |
|
"rewards/margins": 0.1361098289489746, |
|
"rewards/real": 0.004907005000859499, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 51.25, |
|
"learning_rate": 4.746835443037974e-07, |
|
"logits/generated": 0.38353046774864197, |
|
"logits/real": -0.09031276404857635, |
|
"logps/generated": -571.59765625, |
|
"logps/real": -419.74285888671875, |
|
"loss": 0.6689, |
|
"rewards/accuracies": 0.5562499761581421, |
|
"rewards/generated": -0.09446991980075836, |
|
"rewards/margins": 0.06767363846302032, |
|
"rewards/real": -0.026796285063028336, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 51.25, |
|
"learning_rate": 4.992877492877492e-07, |
|
"logits/generated": 0.3525523245334625, |
|
"logits/real": -0.14858296513557434, |
|
"logps/generated": -547.0704345703125, |
|
"logps/real": -442.40228271484375, |
|
"loss": 0.638, |
|
"rewards/accuracies": 0.643750011920929, |
|
"rewards/generated": -0.11326056718826294, |
|
"rewards/margins": 0.12944021821022034, |
|
"rewards/real": 0.016179624944925308, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 48.0, |
|
"learning_rate": 4.957264957264958e-07, |
|
"logits/generated": 0.360097736120224, |
|
"logits/real": -0.18511034548282623, |
|
"logps/generated": -535.1893310546875, |
|
"logps/real": -456.3688049316406, |
|
"loss": 0.611, |
|
"rewards/accuracies": 0.737500011920929, |
|
"rewards/generated": -0.18863049149513245, |
|
"rewards/margins": 0.2453247308731079, |
|
"rewards/real": 0.05669426918029785, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 51.0, |
|
"learning_rate": 4.921652421652421e-07, |
|
"logits/generated": 0.3094131052494049, |
|
"logits/real": -0.13698112964630127, |
|
"logps/generated": -578.9557495117188, |
|
"logps/real": -435.10626220703125, |
|
"loss": 0.6183, |
|
"rewards/accuracies": 0.706250011920929, |
|
"rewards/generated": -0.16466441750526428, |
|
"rewards/margins": 0.2127438485622406, |
|
"rewards/real": 0.04807942733168602, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 49.5, |
|
"learning_rate": 4.886039886039886e-07, |
|
"logits/generated": 0.3409915864467621, |
|
"logits/real": -0.15773525834083557, |
|
"logps/generated": -562.3658447265625, |
|
"logps/real": -472.12261962890625, |
|
"loss": 0.5971, |
|
"rewards/accuracies": 0.7124999761581421, |
|
"rewards/generated": -0.2226313352584839, |
|
"rewards/margins": 0.24577657878398895, |
|
"rewards/real": 0.023145252838730812, |
|
"step": 95 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 46.5, |
|
"learning_rate": 4.850427350427351e-07, |
|
"logits/generated": 0.3600820302963257, |
|
"logits/real": -0.22068698704242706, |
|
"logps/generated": -518.3560791015625, |
|
"logps/real": -505.414794921875, |
|
"loss": 0.5968, |
|
"rewards/accuracies": 0.762499988079071, |
|
"rewards/generated": -0.22140724956989288, |
|
"rewards/margins": 0.2667246460914612, |
|
"rewards/real": 0.045317377895116806, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 47.0, |
|
"learning_rate": 4.814814814814814e-07, |
|
"logits/generated": 0.3369310796260834, |
|
"logits/real": -0.16511984169483185, |
|
"logps/generated": -597.7069091796875, |
|
"logps/real": -469.93951416015625, |
|
"loss": 0.5653, |
|
"rewards/accuracies": 0.7875000238418579, |
|
"rewards/generated": -0.24688784778118134, |
|
"rewards/margins": 0.31749141216278076, |
|
"rewards/real": 0.07060358673334122, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 49.75, |
|
"learning_rate": 4.779202279202279e-07, |
|
"logits/generated": 0.31372249126434326, |
|
"logits/real": -0.13822853565216064, |
|
"logps/generated": -559.5579223632812, |
|
"logps/real": -436.1619567871094, |
|
"loss": 0.5843, |
|
"rewards/accuracies": 0.768750011920929, |
|
"rewards/generated": -0.24767298996448517, |
|
"rewards/margins": 0.2614055275917053, |
|
"rewards/real": 0.013732570223510265, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 44.75, |
|
"learning_rate": 4.743589743589743e-07, |
|
"logits/generated": 0.3527405858039856, |
|
"logits/real": -0.16191214323043823, |
|
"logps/generated": -540.7269287109375, |
|
"logps/real": -481.5582580566406, |
|
"loss": 0.5474, |
|
"rewards/accuracies": 0.824999988079071, |
|
"rewards/generated": -0.3221510648727417, |
|
"rewards/margins": 0.33944302797317505, |
|
"rewards/real": 0.017291929572820663, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 44.0, |
|
"learning_rate": 4.707977207977208e-07, |
|
"logits/generated": 0.31554654240608215, |
|
"logits/real": -0.14937585592269897, |
|
"logps/generated": -533.3251953125, |
|
"logps/real": -425.991455078125, |
|
"loss": 0.5488, |
|
"rewards/accuracies": 0.831250011920929, |
|
"rewards/generated": -0.350756973028183, |
|
"rewards/margins": 0.37362727522850037, |
|
"rewards/real": 0.022870313376188278, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 43.25, |
|
"learning_rate": 4.672364672364672e-07, |
|
"logits/generated": 0.31335240602493286, |
|
"logits/real": -0.16108378767967224, |
|
"logps/generated": -556.67919921875, |
|
"logps/real": -451.10296630859375, |
|
"loss": 0.532, |
|
"rewards/accuracies": 0.856249988079071, |
|
"rewards/generated": -0.3505321741104126, |
|
"rewards/margins": 0.4209725260734558, |
|
"rewards/real": 0.07044035941362381, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 45.75, |
|
"learning_rate": 4.6367521367521367e-07, |
|
"logits/generated": 0.32702523469924927, |
|
"logits/real": -0.16390064358711243, |
|
"logps/generated": -558.4036254882812, |
|
"logps/real": -435.8067932128906, |
|
"loss": 0.5325, |
|
"rewards/accuracies": 0.8687499761581421, |
|
"rewards/generated": -0.3638128936290741, |
|
"rewards/margins": 0.40913742780685425, |
|
"rewards/real": 0.045324526727199554, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 44.0, |
|
"learning_rate": 4.601139601139601e-07, |
|
"logits/generated": 0.28192949295043945, |
|
"logits/real": -0.23023930191993713, |
|
"logps/generated": -578.7821655273438, |
|
"logps/real": -447.467529296875, |
|
"loss": 0.532, |
|
"rewards/accuracies": 0.856249988079071, |
|
"rewards/generated": -0.37364333868026733, |
|
"rewards/margins": 0.4050399363040924, |
|
"rewards/real": 0.03139660134911537, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 45.25, |
|
"learning_rate": 4.5655270655270654e-07, |
|
"logits/generated": 0.3297092914581299, |
|
"logits/real": -0.24259880185127258, |
|
"logps/generated": -557.1586303710938, |
|
"logps/real": -533.3529663085938, |
|
"loss": 0.5193, |
|
"rewards/accuracies": 0.84375, |
|
"rewards/generated": -0.3680626153945923, |
|
"rewards/margins": 0.42688456177711487, |
|
"rewards/real": 0.058821968734264374, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 41.25, |
|
"learning_rate": 4.5299145299145297e-07, |
|
"logits/generated": 0.2855086922645569, |
|
"logits/real": -0.16449978947639465, |
|
"logps/generated": -531.9002685546875, |
|
"logps/real": -427.7098693847656, |
|
"loss": 0.5324, |
|
"rewards/accuracies": 0.856249988079071, |
|
"rewards/generated": -0.4032559394836426, |
|
"rewards/margins": 0.4310643672943115, |
|
"rewards/real": 0.027808407321572304, |
|
"step": 145 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 47.25, |
|
"learning_rate": 4.494301994301994e-07, |
|
"logits/generated": 0.26499122381210327, |
|
"logits/real": -0.17654789984226227, |
|
"logps/generated": -544.5510864257812, |
|
"logps/real": -443.2236328125, |
|
"loss": 0.5146, |
|
"rewards/accuracies": 0.8374999761581421, |
|
"rewards/generated": -0.46017971634864807, |
|
"rewards/margins": 0.47648102045059204, |
|
"rewards/real": 0.016301285475492477, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 43.75, |
|
"learning_rate": 4.4586894586894584e-07, |
|
"logits/generated": 0.32447490096092224, |
|
"logits/real": -0.19060441851615906, |
|
"logps/generated": -565.9913330078125, |
|
"logps/real": -485.25762939453125, |
|
"loss": 0.5035, |
|
"rewards/accuracies": 0.8374999761581421, |
|
"rewards/generated": -0.44145339727401733, |
|
"rewards/margins": 0.4792710244655609, |
|
"rewards/real": 0.03781764581799507, |
|
"step": 155 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 43.75, |
|
"learning_rate": 4.423076923076923e-07, |
|
"logits/generated": 0.3035694360733032, |
|
"logits/real": -0.22318044304847717, |
|
"logps/generated": -566.83251953125, |
|
"logps/real": -444.7403259277344, |
|
"loss": 0.5005, |
|
"rewards/accuracies": 0.862500011920929, |
|
"rewards/generated": -0.4472742974758148, |
|
"rewards/margins": 0.5138620138168335, |
|
"rewards/real": 0.06658776849508286, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 41.25, |
|
"learning_rate": 4.3874643874643876e-07, |
|
"logits/generated": 0.3123244643211365, |
|
"logits/real": -0.1916072815656662, |
|
"logps/generated": -558.3717041015625, |
|
"logps/real": -460.4476623535156, |
|
"loss": 0.4883, |
|
"rewards/accuracies": 0.800000011920929, |
|
"rewards/generated": -0.4848392605781555, |
|
"rewards/margins": 0.4970267713069916, |
|
"rewards/real": 0.012187558226287365, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 41.25, |
|
"learning_rate": 4.3518518518518514e-07, |
|
"logits/generated": 0.250892698764801, |
|
"logits/real": -0.18243929743766785, |
|
"logps/generated": -536.6226196289062, |
|
"logps/real": -416.77984619140625, |
|
"loss": 0.496, |
|
"rewards/accuracies": 0.8687499761581421, |
|
"rewards/generated": -0.4702886939048767, |
|
"rewards/margins": 0.5208121538162231, |
|
"rewards/real": 0.05052344128489494, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 41.0, |
|
"learning_rate": 4.3162393162393163e-07, |
|
"logits/generated": 0.29632076621055603, |
|
"logits/real": -0.2431728094816208, |
|
"logps/generated": -538.0128173828125, |
|
"logps/real": -478.095703125, |
|
"loss": 0.4654, |
|
"rewards/accuracies": 0.8999999761581421, |
|
"rewards/generated": -0.5014994144439697, |
|
"rewards/margins": 0.5601893663406372, |
|
"rewards/real": 0.0586898997426033, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 44.5, |
|
"learning_rate": 4.2806267806267807e-07, |
|
"logits/generated": 0.277912437915802, |
|
"logits/real": -0.2574039101600647, |
|
"logps/generated": -561.8548583984375, |
|
"logps/real": -444.5439453125, |
|
"loss": 0.4828, |
|
"rewards/accuracies": 0.875, |
|
"rewards/generated": -0.5042586922645569, |
|
"rewards/margins": 0.5849383473396301, |
|
"rewards/real": 0.08067961782217026, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 43.25, |
|
"learning_rate": 4.245014245014245e-07, |
|
"logits/generated": 0.33910220861434937, |
|
"logits/real": -0.11496976763010025, |
|
"logps/generated": -599.3956298828125, |
|
"logps/real": -421.74334716796875, |
|
"loss": 0.4619, |
|
"rewards/accuracies": 0.8999999761581421, |
|
"rewards/generated": -0.572574257850647, |
|
"rewards/margins": 0.6046732664108276, |
|
"rewards/real": 0.032099105417728424, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 43.5, |
|
"learning_rate": 4.2094017094017093e-07, |
|
"logits/generated": 0.32159385085105896, |
|
"logits/real": -0.17387528717517853, |
|
"logps/generated": -586.5929565429688, |
|
"logps/real": -426.36444091796875, |
|
"loss": 0.483, |
|
"rewards/accuracies": 0.875, |
|
"rewards/generated": -0.5183109641075134, |
|
"rewards/margins": 0.5664752721786499, |
|
"rewards/real": 0.0481642410159111, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 39.5, |
|
"learning_rate": 4.173789173789173e-07, |
|
"logits/generated": 0.26057395339012146, |
|
"logits/real": -0.21449565887451172, |
|
"logps/generated": -527.934326171875, |
|
"logps/real": -452.62567138671875, |
|
"loss": 0.465, |
|
"rewards/accuracies": 0.893750011920929, |
|
"rewards/generated": -0.53789222240448, |
|
"rewards/margins": 0.5843125581741333, |
|
"rewards/real": 0.04642019420862198, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 39.5, |
|
"learning_rate": 4.138176638176638e-07, |
|
"logits/generated": 0.3313707411289215, |
|
"logits/real": -0.19944889843463898, |
|
"logps/generated": -580.3385620117188, |
|
"logps/real": -472.051513671875, |
|
"loss": 0.4567, |
|
"rewards/accuracies": 0.862500011920929, |
|
"rewards/generated": -0.5696550607681274, |
|
"rewards/margins": 0.5835729837417603, |
|
"rewards/real": 0.013917927630245686, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 41.0, |
|
"learning_rate": 4.1025641025641024e-07, |
|
"logits/generated": 0.3330136835575104, |
|
"logits/real": -0.1847701370716095, |
|
"logps/generated": -576.3521728515625, |
|
"logps/real": -470.685302734375, |
|
"loss": 0.4635, |
|
"rewards/accuracies": 0.8687499761581421, |
|
"rewards/generated": -0.5276187062263489, |
|
"rewards/margins": 0.5738595724105835, |
|
"rewards/real": 0.04624096304178238, |
|
"step": 205 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 42.25, |
|
"learning_rate": 4.0669515669515667e-07, |
|
"logits/generated": 0.25724920630455017, |
|
"logits/real": -0.2856508791446686, |
|
"logps/generated": -558.7901611328125, |
|
"logps/real": -467.52008056640625, |
|
"loss": 0.4516, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -0.5960500836372375, |
|
"rewards/margins": 0.6805692911148071, |
|
"rewards/real": 0.08451925963163376, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 40.25, |
|
"learning_rate": 4.031339031339031e-07, |
|
"logits/generated": 0.28837472200393677, |
|
"logits/real": -0.19708415865898132, |
|
"logps/generated": -549.651123046875, |
|
"logps/real": -457.85919189453125, |
|
"loss": 0.4424, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -0.6616542935371399, |
|
"rewards/margins": 0.6647108793258667, |
|
"rewards/real": 0.0030566223431378603, |
|
"step": 215 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 39.5, |
|
"learning_rate": 3.995726495726496e-07, |
|
"logits/generated": 0.30279049277305603, |
|
"logits/real": -0.18387384712696075, |
|
"logps/generated": -541.563720703125, |
|
"logps/real": -444.663818359375, |
|
"loss": 0.4583, |
|
"rewards/accuracies": 0.887499988079071, |
|
"rewards/generated": -0.5965785384178162, |
|
"rewards/margins": 0.6331891417503357, |
|
"rewards/real": 0.03661050274968147, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 40.5, |
|
"learning_rate": 3.9601139601139597e-07, |
|
"logits/generated": 0.25506392121315, |
|
"logits/real": -0.23414012789726257, |
|
"logps/generated": -574.35595703125, |
|
"logps/real": -450.7875061035156, |
|
"loss": 0.4459, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -0.6340950131416321, |
|
"rewards/margins": 0.6694762110710144, |
|
"rewards/real": 0.0353812575340271, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 37.75, |
|
"learning_rate": 3.924501424501424e-07, |
|
"logits/generated": 0.25567466020584106, |
|
"logits/real": -0.20427341759204865, |
|
"logps/generated": -546.7042236328125, |
|
"logps/real": -421.8016662597656, |
|
"loss": 0.4303, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -0.5935453176498413, |
|
"rewards/margins": 0.685859203338623, |
|
"rewards/real": 0.09231381118297577, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 41.75, |
|
"learning_rate": 3.888888888888889e-07, |
|
"logits/generated": 0.3079773783683777, |
|
"logits/real": -0.21253804862499237, |
|
"logps/generated": -598.437744140625, |
|
"logps/real": -443.98492431640625, |
|
"loss": 0.4276, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.722990870475769, |
|
"rewards/margins": 0.7524241209030151, |
|
"rewards/real": 0.029433395713567734, |
|
"step": 235 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 37.5, |
|
"learning_rate": 3.853276353276353e-07, |
|
"logits/generated": 0.30521565675735474, |
|
"logits/real": -0.1906270682811737, |
|
"logps/generated": -541.72607421875, |
|
"logps/real": -424.0105895996094, |
|
"loss": 0.4269, |
|
"rewards/accuracies": 0.8812500238418579, |
|
"rewards/generated": -0.667791485786438, |
|
"rewards/margins": 0.7342725992202759, |
|
"rewards/real": 0.06648121774196625, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 36.75, |
|
"learning_rate": 3.8176638176638176e-07, |
|
"logits/generated": 0.3006458282470703, |
|
"logits/real": -0.2351672649383545, |
|
"logps/generated": -559.7643432617188, |
|
"logps/real": -504.17987060546875, |
|
"loss": 0.4507, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -0.6496528387069702, |
|
"rewards/margins": 0.67330002784729, |
|
"rewards/real": 0.023647133260965347, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 37.0, |
|
"learning_rate": 3.782051282051282e-07, |
|
"logits/generated": 0.2616468071937561, |
|
"logits/real": -0.2550817131996155, |
|
"logps/generated": -556.2052001953125, |
|
"logps/real": -438.43743896484375, |
|
"loss": 0.4118, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -0.7023671865463257, |
|
"rewards/margins": 0.7656707763671875, |
|
"rewards/real": 0.06330356001853943, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 37.0, |
|
"learning_rate": 3.7464387464387463e-07, |
|
"logits/generated": 0.3042835593223572, |
|
"logits/real": -0.22142143547534943, |
|
"logps/generated": -538.9241333007812, |
|
"logps/real": -484.5042419433594, |
|
"loss": 0.4347, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.7196381092071533, |
|
"rewards/margins": 0.7357234358787537, |
|
"rewards/real": 0.016085291281342506, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 38.75, |
|
"learning_rate": 3.7108262108262107e-07, |
|
"logits/generated": 0.28268715739250183, |
|
"logits/real": -0.2414659708738327, |
|
"logps/generated": -552.683837890625, |
|
"logps/real": -439.206298828125, |
|
"loss": 0.4091, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -0.738116443157196, |
|
"rewards/margins": 0.7844719886779785, |
|
"rewards/real": 0.04635554179549217, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"grad_norm": 36.5, |
|
"learning_rate": 3.6752136752136755e-07, |
|
"logits/generated": 0.2755090296268463, |
|
"logits/real": -0.2653561234474182, |
|
"logps/generated": -575.4991455078125, |
|
"logps/real": -458.960693359375, |
|
"loss": 0.4082, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.7462833523750305, |
|
"rewards/margins": 0.7978280782699585, |
|
"rewards/real": 0.05154471844434738, |
|
"step": 265 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 40.25, |
|
"learning_rate": 3.6396011396011393e-07, |
|
"logits/generated": 0.3112264573574066, |
|
"logits/real": -0.2405599057674408, |
|
"logps/generated": -537.7279052734375, |
|
"logps/real": -463.67718505859375, |
|
"loss": 0.4195, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.7078372240066528, |
|
"rewards/margins": 0.7539125084877014, |
|
"rewards/real": 0.04607529193162918, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 36.75, |
|
"learning_rate": 3.6039886039886037e-07, |
|
"logits/generated": 0.2953110635280609, |
|
"logits/real": -0.21746914088726044, |
|
"logps/generated": -562.6829833984375, |
|
"logps/real": -432.52105712890625, |
|
"loss": 0.3822, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -0.7816409468650818, |
|
"rewards/margins": 0.8443856239318848, |
|
"rewards/real": 0.06274472177028656, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 36.75, |
|
"learning_rate": 3.5683760683760686e-07, |
|
"logits/generated": 0.29402461647987366, |
|
"logits/real": -0.16522033512592316, |
|
"logps/generated": -568.8770751953125, |
|
"logps/real": -452.81866455078125, |
|
"loss": 0.3823, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -0.8258803486824036, |
|
"rewards/margins": 0.9027735590934753, |
|
"rewards/real": 0.07689327746629715, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 37.5, |
|
"learning_rate": 3.5327635327635324e-07, |
|
"logits/generated": 0.2609436810016632, |
|
"logits/real": -0.23396773636341095, |
|
"logps/generated": -557.7428588867188, |
|
"logps/real": -437.74493408203125, |
|
"loss": 0.4063, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -0.762947678565979, |
|
"rewards/margins": 0.7624271512031555, |
|
"rewards/real": -0.0005204956978559494, |
|
"step": 285 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 39.0, |
|
"learning_rate": 3.497150997150997e-07, |
|
"logits/generated": 0.29128846526145935, |
|
"logits/real": -0.29488933086395264, |
|
"logps/generated": -569.3885498046875, |
|
"logps/real": -493.6559143066406, |
|
"loss": 0.3907, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -0.8542326092720032, |
|
"rewards/margins": 0.8633776903152466, |
|
"rewards/real": 0.009145094081759453, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 40.0, |
|
"learning_rate": 3.461538461538461e-07, |
|
"logits/generated": 0.2510134279727936, |
|
"logits/real": -0.23930807411670685, |
|
"logps/generated": -571.7144775390625, |
|
"logps/real": -420.68890380859375, |
|
"loss": 0.3903, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -0.8466684222221375, |
|
"rewards/margins": 0.8632675409317017, |
|
"rewards/real": 0.016599006950855255, |
|
"step": 295 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 34.5, |
|
"learning_rate": 3.425925925925926e-07, |
|
"logits/generated": 0.24713429808616638, |
|
"logits/real": -0.2522990107536316, |
|
"logps/generated": -579.431396484375, |
|
"logps/real": -446.91668701171875, |
|
"loss": 0.4016, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -0.852130115032196, |
|
"rewards/margins": 0.845871090888977, |
|
"rewards/real": -0.006259086541831493, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"eval_logits/generated": 0.25726401805877686, |
|
"eval_logits/real": -0.2516762316226959, |
|
"eval_logps/generated": -568.6871948242188, |
|
"eval_logps/real": -468.50299072265625, |
|
"eval_loss": 0.3913811147212982, |
|
"eval_rewards/accuracies": 0.9319999814033508, |
|
"eval_rewards/generated": -0.8142948746681213, |
|
"eval_rewards/margins": 0.8413105010986328, |
|
"eval_rewards/real": 0.027015656232833862, |
|
"eval_runtime": 262.9656, |
|
"eval_samples_per_second": 7.606, |
|
"eval_steps_per_second": 0.951, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 34.0, |
|
"learning_rate": 3.3903133903133903e-07, |
|
"logits/generated": 0.2464553862810135, |
|
"logits/real": -0.24754126369953156, |
|
"logps/generated": -557.3573608398438, |
|
"logps/real": -447.5314025878906, |
|
"loss": 0.3659, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -0.8741347193717957, |
|
"rewards/margins": 0.9578303098678589, |
|
"rewards/real": 0.08369561284780502, |
|
"step": 305 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 36.0, |
|
"learning_rate": 3.354700854700854e-07, |
|
"logits/generated": 0.2558444142341614, |
|
"logits/real": -0.2749106287956238, |
|
"logps/generated": -581.9246826171875, |
|
"logps/real": -470.64117431640625, |
|
"loss": 0.3883, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.870457649230957, |
|
"rewards/margins": 0.9120520353317261, |
|
"rewards/real": 0.04159443825483322, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 40.0, |
|
"learning_rate": 3.319088319088319e-07, |
|
"logits/generated": 0.2201302945613861, |
|
"logits/real": -0.2533838152885437, |
|
"logps/generated": -551.3731689453125, |
|
"logps/real": -479.00115966796875, |
|
"loss": 0.3958, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -0.7957965135574341, |
|
"rewards/margins": 0.8228376507759094, |
|
"rewards/real": 0.027041062712669373, |
|
"step": 315 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 38.75, |
|
"learning_rate": 3.2834757834757833e-07, |
|
"logits/generated": 0.2794730067253113, |
|
"logits/real": -0.1945553719997406, |
|
"logps/generated": -570.4600830078125, |
|
"logps/real": -423.16229248046875, |
|
"loss": 0.3755, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -0.8808556795120239, |
|
"rewards/margins": 0.8986592292785645, |
|
"rewards/real": 0.01780361495912075, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 36.25, |
|
"learning_rate": 3.2478632478632476e-07, |
|
"logits/generated": 0.28926241397857666, |
|
"logits/real": -0.2784452438354492, |
|
"logps/generated": -584.1411743164062, |
|
"logps/real": -480.142578125, |
|
"loss": 0.3778, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -0.8571932911872864, |
|
"rewards/margins": 0.9160116910934448, |
|
"rewards/real": 0.058818407356739044, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 34.75, |
|
"learning_rate": 3.212250712250712e-07, |
|
"logits/generated": 0.25457730889320374, |
|
"logits/real": -0.2175012081861496, |
|
"logps/generated": -571.3527221679688, |
|
"logps/real": -437.7066345214844, |
|
"loss": 0.3903, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -0.8724948167800903, |
|
"rewards/margins": 0.9130793809890747, |
|
"rewards/real": 0.04058459773659706, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 35.5, |
|
"learning_rate": 3.176638176638177e-07, |
|
"logits/generated": 0.2662949562072754, |
|
"logits/real": -0.2381666898727417, |
|
"logps/generated": -601.4517211914062, |
|
"logps/real": -483.5282287597656, |
|
"loss": 0.3865, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -0.8890978693962097, |
|
"rewards/margins": 0.8633550405502319, |
|
"rewards/real": -0.02574281021952629, |
|
"step": 335 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 35.25, |
|
"learning_rate": 3.1410256410256407e-07, |
|
"logits/generated": 0.25792551040649414, |
|
"logits/real": -0.28233546018600464, |
|
"logps/generated": -575.9357299804688, |
|
"logps/real": -461.888916015625, |
|
"loss": 0.3606, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -0.9487969279289246, |
|
"rewards/margins": 1.0083829164505005, |
|
"rewards/real": 0.05958594009280205, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 35.5, |
|
"learning_rate": 3.1054131054131055e-07, |
|
"logits/generated": 0.24235877394676208, |
|
"logits/real": -0.19490960240364075, |
|
"logps/generated": -574.0829467773438, |
|
"logps/real": -401.38031005859375, |
|
"loss": 0.3805, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -0.8425151705741882, |
|
"rewards/margins": 0.8939135670661926, |
|
"rewards/real": 0.0513983853161335, |
|
"step": 345 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 33.25, |
|
"learning_rate": 3.06980056980057e-07, |
|
"logits/generated": 0.2665616571903229, |
|
"logits/real": -0.24914798140525818, |
|
"logps/generated": -547.2540283203125, |
|
"logps/real": -449.8600158691406, |
|
"loss": 0.3741, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.8794205784797668, |
|
"rewards/margins": 0.8983514904975891, |
|
"rewards/real": 0.01893080770969391, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 38.0, |
|
"learning_rate": 3.0341880341880337e-07, |
|
"logits/generated": 0.2690719962120056, |
|
"logits/real": -0.26707571744918823, |
|
"logps/generated": -573.3060913085938, |
|
"logps/real": -484.48492431640625, |
|
"loss": 0.3759, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -0.9073358774185181, |
|
"rewards/margins": 0.927741527557373, |
|
"rewards/real": 0.020405706018209457, |
|
"step": 355 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 35.0, |
|
"learning_rate": 2.9985754985754986e-07, |
|
"logits/generated": 0.23270320892333984, |
|
"logits/real": -0.2126883715391159, |
|
"logps/generated": -566.2892456054688, |
|
"logps/real": -420.14471435546875, |
|
"loss": 0.373, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -0.9121645092964172, |
|
"rewards/margins": 0.9226058721542358, |
|
"rewards/real": 0.01044147927314043, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 33.0, |
|
"learning_rate": 2.962962962962963e-07, |
|
"logits/generated": 0.24159935116767883, |
|
"logits/real": -0.26122623682022095, |
|
"logps/generated": -570.7769775390625, |
|
"logps/real": -483.9549865722656, |
|
"loss": 0.3691, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.9642108082771301, |
|
"rewards/margins": 0.9902675747871399, |
|
"rewards/real": 0.0260569266974926, |
|
"step": 365 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 36.0, |
|
"learning_rate": 2.927350427350427e-07, |
|
"logits/generated": 0.23782770335674286, |
|
"logits/real": -0.2670837640762329, |
|
"logps/generated": -573.4016723632812, |
|
"logps/real": -469.85491943359375, |
|
"loss": 0.3758, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.9776171445846558, |
|
"rewards/margins": 0.9925469160079956, |
|
"rewards/real": 0.01492965966463089, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 34.0, |
|
"learning_rate": 2.8917378917378916e-07, |
|
"logits/generated": 0.2320900857448578, |
|
"logits/real": -0.18077997863292694, |
|
"logps/generated": -559.6778564453125, |
|
"logps/real": -410.2744140625, |
|
"loss": 0.3682, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -0.9255533218383789, |
|
"rewards/margins": 0.9920336008071899, |
|
"rewards/real": 0.06648023426532745, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 37.25, |
|
"learning_rate": 2.8561253561253565e-07, |
|
"logits/generated": 0.21389861404895782, |
|
"logits/real": -0.26197221875190735, |
|
"logps/generated": -571.7706909179688, |
|
"logps/real": -451.16253662109375, |
|
"loss": 0.3799, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -0.9532166719436646, |
|
"rewards/margins": 0.9206304550170898, |
|
"rewards/real": -0.03258631005883217, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 36.0, |
|
"learning_rate": 2.8205128205128203e-07, |
|
"logits/generated": 0.22054708003997803, |
|
"logits/real": -0.25541600584983826, |
|
"logps/generated": -549.7349853515625, |
|
"logps/real": -407.8997497558594, |
|
"loss": 0.3712, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.875189483165741, |
|
"rewards/margins": 0.9158505201339722, |
|
"rewards/real": 0.04066096618771553, |
|
"step": 385 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 35.25, |
|
"learning_rate": 2.7849002849002846e-07, |
|
"logits/generated": 0.24272923171520233, |
|
"logits/real": -0.2981414198875427, |
|
"logps/generated": -567.8425903320312, |
|
"logps/real": -499.34588623046875, |
|
"loss": 0.3782, |
|
"rewards/accuracies": 0.918749988079071, |
|
"rewards/generated": -0.8826544880867004, |
|
"rewards/margins": 0.8991689682006836, |
|
"rewards/real": 0.01651454158127308, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 36.25, |
|
"learning_rate": 2.749287749287749e-07, |
|
"logits/generated": 0.25423914194107056, |
|
"logits/real": -0.27515119314193726, |
|
"logps/generated": -561.5721435546875, |
|
"logps/real": -458.6827087402344, |
|
"loss": 0.3479, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -0.976852536201477, |
|
"rewards/margins": 1.0118058919906616, |
|
"rewards/real": 0.034953318536281586, |
|
"step": 395 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 31.125, |
|
"learning_rate": 2.7136752136752133e-07, |
|
"logits/generated": 0.23657219111919403, |
|
"logits/real": -0.28398722410202026, |
|
"logps/generated": -593.6157836914062, |
|
"logps/real": -420.7015686035156, |
|
"loss": 0.3569, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -0.9960700869560242, |
|
"rewards/margins": 1.0824501514434814, |
|
"rewards/real": 0.08638016134500504, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 34.75, |
|
"learning_rate": 2.678062678062678e-07, |
|
"logits/generated": 0.24148711562156677, |
|
"logits/real": -0.21437129378318787, |
|
"logps/generated": -577.3068237304688, |
|
"logps/real": -450.6416015625, |
|
"loss": 0.3691, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.9539799690246582, |
|
"rewards/margins": 1.0042264461517334, |
|
"rewards/real": 0.05024648830294609, |
|
"step": 405 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 34.0, |
|
"learning_rate": 2.642450142450142e-07, |
|
"logits/generated": 0.27527815103530884, |
|
"logits/real": -0.2802174687385559, |
|
"logps/generated": -584.6751708984375, |
|
"logps/real": -474.64678955078125, |
|
"loss": 0.3432, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0197203159332275, |
|
"rewards/margins": 1.0799630880355835, |
|
"rewards/real": 0.06024279445409775, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 34.25, |
|
"learning_rate": 2.606837606837607e-07, |
|
"logits/generated": 0.24102333188056946, |
|
"logits/real": -0.2790641188621521, |
|
"logps/generated": -580.9298706054688, |
|
"logps/real": -423.9723205566406, |
|
"loss": 0.3673, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -0.9787546396255493, |
|
"rewards/margins": 1.0065668821334839, |
|
"rewards/real": 0.027812320739030838, |
|
"step": 415 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 34.75, |
|
"learning_rate": 2.571225071225071e-07, |
|
"logits/generated": 0.2540944218635559, |
|
"logits/real": -0.19810980558395386, |
|
"logps/generated": -562.8670654296875, |
|
"logps/real": -426.3075256347656, |
|
"loss": 0.35, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.0469673871994019, |
|
"rewards/margins": 1.0949853658676147, |
|
"rewards/real": 0.04801792651414871, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 33.0, |
|
"learning_rate": 2.5356125356125355e-07, |
|
"logits/generated": 0.20098896324634552, |
|
"logits/real": -0.24582453072071075, |
|
"logps/generated": -573.8992309570312, |
|
"logps/real": -420.2386779785156, |
|
"loss": 0.3456, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0094668865203857, |
|
"rewards/margins": 1.0302053689956665, |
|
"rewards/real": 0.020738471299409866, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 36.25, |
|
"learning_rate": 2.5e-07, |
|
"logits/generated": 0.2519596219062805, |
|
"logits/real": -0.27877140045166016, |
|
"logps/generated": -565.7482299804688, |
|
"logps/real": -534.8162841796875, |
|
"loss": 0.3519, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -0.9428899884223938, |
|
"rewards/margins": 0.976216197013855, |
|
"rewards/real": 0.033326275646686554, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 36.75, |
|
"learning_rate": 2.464387464387464e-07, |
|
"logits/generated": 0.23958304524421692, |
|
"logits/real": -0.3024999499320984, |
|
"logps/generated": -558.3704223632812, |
|
"logps/real": -470.7825622558594, |
|
"loss": 0.3492, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0124276876449585, |
|
"rewards/margins": 1.0718011856079102, |
|
"rewards/real": 0.059373389929533005, |
|
"step": 435 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 30.75, |
|
"learning_rate": 2.4287749287749286e-07, |
|
"logits/generated": 0.2602802515029907, |
|
"logits/real": -0.28142985701560974, |
|
"logps/generated": -619.5274047851562, |
|
"logps/real": -502.71832275390625, |
|
"loss": 0.3466, |
|
"rewards/accuracies": 0.9750000238418579, |
|
"rewards/generated": -1.0051062107086182, |
|
"rewards/margins": 1.0426944494247437, |
|
"rewards/real": 0.037588391453027725, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 33.25, |
|
"learning_rate": 2.393162393162393e-07, |
|
"logits/generated": 0.21535196900367737, |
|
"logits/real": -0.2805056869983673, |
|
"logps/generated": -587.7852172851562, |
|
"logps/real": -429.0276794433594, |
|
"loss": 0.3489, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -0.9763972163200378, |
|
"rewards/margins": 0.9719440340995789, |
|
"rewards/real": -0.004453244619071484, |
|
"step": 445 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 33.75, |
|
"learning_rate": 2.3575498575498575e-07, |
|
"logits/generated": 0.22773954272270203, |
|
"logits/real": -0.2382933795452118, |
|
"logps/generated": -591.6124877929688, |
|
"logps/real": -456.0254821777344, |
|
"loss": 0.3504, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.0408380031585693, |
|
"rewards/margins": 1.030290961265564, |
|
"rewards/real": -0.010547079145908356, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 35.25, |
|
"learning_rate": 2.3219373219373216e-07, |
|
"logits/generated": 0.22827064990997314, |
|
"logits/real": -0.3192758560180664, |
|
"logps/generated": -563.3525390625, |
|
"logps/real": -482.61541748046875, |
|
"loss": 0.3454, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.9959697723388672, |
|
"rewards/margins": 1.0202291011810303, |
|
"rewards/real": 0.024259375408291817, |
|
"step": 455 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 36.25, |
|
"learning_rate": 2.2863247863247862e-07, |
|
"logits/generated": 0.2440454512834549, |
|
"logits/real": -0.22086623311042786, |
|
"logps/generated": -557.6519165039062, |
|
"logps/real": -446.55230712890625, |
|
"loss": 0.3376, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.0488998889923096, |
|
"rewards/margins": 1.0662249326705933, |
|
"rewards/real": 0.017325039952993393, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 34.5, |
|
"learning_rate": 2.2507122507122505e-07, |
|
"logits/generated": 0.21471574902534485, |
|
"logits/real": -0.2694615423679352, |
|
"logps/generated": -556.4073486328125, |
|
"logps/real": -466.55120849609375, |
|
"loss": 0.3528, |
|
"rewards/accuracies": 0.90625, |
|
"rewards/generated": -1.063711404800415, |
|
"rewards/margins": 1.067757487297058, |
|
"rewards/real": 0.004046155605465174, |
|
"step": 465 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 35.75, |
|
"learning_rate": 2.215099715099715e-07, |
|
"logits/generated": 0.223112553358078, |
|
"logits/real": -0.2675575315952301, |
|
"logps/generated": -554.9754028320312, |
|
"logps/real": -409.7915954589844, |
|
"loss": 0.3554, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.9735461473464966, |
|
"rewards/margins": 0.9401170015335083, |
|
"rewards/real": -0.03342916816473007, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 32.75, |
|
"learning_rate": 2.1794871794871795e-07, |
|
"logits/generated": 0.24918599426746368, |
|
"logits/real": -0.28682878613471985, |
|
"logps/generated": -571.9830322265625, |
|
"logps/real": -467.03125, |
|
"loss": 0.3409, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.0399566888809204, |
|
"rewards/margins": 1.0961642265319824, |
|
"rewards/real": 0.056207604706287384, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 30.875, |
|
"learning_rate": 2.1438746438746438e-07, |
|
"logits/generated": 0.2370508462190628, |
|
"logits/real": -0.285112202167511, |
|
"logps/generated": -598.0405883789062, |
|
"logps/real": -468.107421875, |
|
"loss": 0.3299, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1196238994598389, |
|
"rewards/margins": 1.1285395622253418, |
|
"rewards/real": 0.008915687911212444, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 33.25, |
|
"learning_rate": 2.1082621082621082e-07, |
|
"logits/generated": 0.21018759906291962, |
|
"logits/real": -0.2482105940580368, |
|
"logps/generated": -559.3743896484375, |
|
"logps/real": -418.2901916503906, |
|
"loss": 0.354, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.0670310258865356, |
|
"rewards/margins": 1.0917372703552246, |
|
"rewards/real": 0.02470635250210762, |
|
"step": 485 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 32.5, |
|
"learning_rate": 2.0726495726495728e-07, |
|
"logits/generated": 0.2351360023021698, |
|
"logits/real": -0.3030211925506592, |
|
"logps/generated": -561.6700439453125, |
|
"logps/real": -508.479248046875, |
|
"loss": 0.3462, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -1.0717344284057617, |
|
"rewards/margins": 1.0173569917678833, |
|
"rewards/real": -0.05437753349542618, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 33.25, |
|
"learning_rate": 2.0370370370370369e-07, |
|
"logits/generated": 0.21695998311042786, |
|
"logits/real": -0.262247234582901, |
|
"logps/generated": -560.5780029296875, |
|
"logps/real": -447.3353576660156, |
|
"loss": 0.346, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.0143311023712158, |
|
"rewards/margins": 1.0173934698104858, |
|
"rewards/real": 0.0030622382182627916, |
|
"step": 495 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 33.5, |
|
"learning_rate": 2.0014245014245012e-07, |
|
"logits/generated": 0.24932141602039337, |
|
"logits/real": -0.28666001558303833, |
|
"logps/generated": -548.8845825195312, |
|
"logps/real": -473.7200622558594, |
|
"loss": 0.359, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -0.9833143353462219, |
|
"rewards/margins": 0.9953653216362, |
|
"rewards/real": 0.012050976976752281, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 30.5, |
|
"learning_rate": 1.9658119658119656e-07, |
|
"logits/generated": 0.2177625447511673, |
|
"logits/real": -0.27629366517066956, |
|
"logps/generated": -541.3976440429688, |
|
"logps/real": -459.02227783203125, |
|
"loss": 0.3403, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -0.9844304919242859, |
|
"rewards/margins": 1.0443334579467773, |
|
"rewards/real": 0.0599028654396534, |
|
"step": 505 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 33.25, |
|
"learning_rate": 1.9301994301994302e-07, |
|
"logits/generated": 0.22767607867717743, |
|
"logits/real": -0.25904038548469543, |
|
"logps/generated": -568.2036743164062, |
|
"logps/real": -443.6368713378906, |
|
"loss": 0.3417, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0040171146392822, |
|
"rewards/margins": 1.078296184539795, |
|
"rewards/real": 0.07427913695573807, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 34.75, |
|
"learning_rate": 1.8945868945868945e-07, |
|
"logits/generated": 0.2245851755142212, |
|
"logits/real": -0.3053443431854248, |
|
"logps/generated": -569.513671875, |
|
"logps/real": -470.54547119140625, |
|
"loss": 0.3494, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -1.0781540870666504, |
|
"rewards/margins": 1.0763808488845825, |
|
"rewards/real": -0.0017731055850163102, |
|
"step": 515 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 33.25, |
|
"learning_rate": 1.8589743589743588e-07, |
|
"logits/generated": 0.228353351354599, |
|
"logits/real": -0.27608901262283325, |
|
"logps/generated": -549.1878662109375, |
|
"logps/real": -457.79681396484375, |
|
"loss": 0.3254, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0618976354599, |
|
"rewards/margins": 1.0646899938583374, |
|
"rewards/real": 0.0027922815643250942, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 32.5, |
|
"learning_rate": 1.8233618233618234e-07, |
|
"logits/generated": 0.24573054909706116, |
|
"logits/real": -0.24587111175060272, |
|
"logps/generated": -527.3595581054688, |
|
"logps/real": -461.79351806640625, |
|
"loss": 0.3365, |
|
"rewards/accuracies": 0.9312499761581421, |
|
"rewards/generated": -0.9843952059745789, |
|
"rewards/margins": 1.0428409576416016, |
|
"rewards/real": 0.05844569206237793, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 30.5, |
|
"learning_rate": 1.7877492877492878e-07, |
|
"logits/generated": 0.2456216812133789, |
|
"logits/real": -0.3523867428302765, |
|
"logps/generated": -584.2765502929688, |
|
"logps/real": -485.2999572753906, |
|
"loss": 0.3442, |
|
"rewards/accuracies": 0.96875, |
|
"rewards/generated": -1.1297935247421265, |
|
"rewards/margins": 1.1257268190383911, |
|
"rewards/real": -0.0040667857974767685, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 32.75, |
|
"learning_rate": 1.752136752136752e-07, |
|
"logits/generated": 0.24020275473594666, |
|
"logits/real": -0.2194860428571701, |
|
"logps/generated": -575.4654541015625, |
|
"logps/real": -414.576416015625, |
|
"loss": 0.3328, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1157715320587158, |
|
"rewards/margins": 1.0961633920669556, |
|
"rewards/real": -0.01960797980427742, |
|
"step": 535 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 31.75, |
|
"learning_rate": 1.7165242165242165e-07, |
|
"logits/generated": 0.20955972373485565, |
|
"logits/real": -0.25930672883987427, |
|
"logps/generated": -591.9100341796875, |
|
"logps/real": -403.3531799316406, |
|
"loss": 0.3286, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -1.076982855796814, |
|
"rewards/margins": 1.1210120916366577, |
|
"rewards/real": 0.044029366225004196, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 31.375, |
|
"learning_rate": 1.6809116809116808e-07, |
|
"logits/generated": 0.18421868979930878, |
|
"logits/real": -0.29251617193222046, |
|
"logps/generated": -552.1446533203125, |
|
"logps/real": -425.1080627441406, |
|
"loss": 0.3362, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -0.986895740032196, |
|
"rewards/margins": 1.0399030447006226, |
|
"rewards/real": 0.05300729721784592, |
|
"step": 545 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 32.5, |
|
"learning_rate": 1.6452991452991452e-07, |
|
"logits/generated": 0.22791898250579834, |
|
"logits/real": -0.28665608167648315, |
|
"logps/generated": -598.5328369140625, |
|
"logps/real": -407.1868591308594, |
|
"loss": 0.3265, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.205278754234314, |
|
"rewards/margins": 1.2098186016082764, |
|
"rewards/real": 0.004539764020591974, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"grad_norm": 31.625, |
|
"learning_rate": 1.6096866096866095e-07, |
|
"logits/generated": 0.22963444888591766, |
|
"logits/real": -0.24107995629310608, |
|
"logps/generated": -566.4512329101562, |
|
"logps/real": -450.96978759765625, |
|
"loss": 0.3351, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.081211805343628, |
|
"rewards/margins": 1.0905853509902954, |
|
"rewards/real": 0.009373527020215988, |
|
"step": 555 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 34.0, |
|
"learning_rate": 1.574074074074074e-07, |
|
"logits/generated": 0.24042251706123352, |
|
"logits/real": -0.21857628226280212, |
|
"logps/generated": -569.6424560546875, |
|
"logps/real": -448.65948486328125, |
|
"loss": 0.331, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.0992162227630615, |
|
"rewards/margins": 1.157705545425415, |
|
"rewards/real": 0.058489274233579636, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 33.25, |
|
"learning_rate": 1.5384615384615385e-07, |
|
"logits/generated": 0.1823839247226715, |
|
"logits/real": -0.2623021900653839, |
|
"logps/generated": -577.726318359375, |
|
"logps/real": -445.66534423828125, |
|
"loss": 0.3352, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.0678300857543945, |
|
"rewards/margins": 1.0674610137939453, |
|
"rewards/real": -0.000368914392311126, |
|
"step": 565 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 30.75, |
|
"learning_rate": 1.5028490028490028e-07, |
|
"logits/generated": 0.2117844521999359, |
|
"logits/real": -0.2711123526096344, |
|
"logps/generated": -577.0066528320312, |
|
"logps/real": -406.86175537109375, |
|
"loss": 0.3154, |
|
"rewards/accuracies": 0.981249988079071, |
|
"rewards/generated": -1.0694202184677124, |
|
"rewards/margins": 1.1183080673217773, |
|
"rewards/real": 0.04888775572180748, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 32.75, |
|
"learning_rate": 1.4672364672364671e-07, |
|
"logits/generated": 0.2499610185623169, |
|
"logits/real": -0.2550189197063446, |
|
"logps/generated": -583.8297119140625, |
|
"logps/real": -480.6053771972656, |
|
"loss": 0.3408, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.01523756980896, |
|
"rewards/margins": 1.0397374629974365, |
|
"rewards/real": 0.024500016123056412, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 32.5, |
|
"learning_rate": 1.4316239316239315e-07, |
|
"logits/generated": 0.1771518439054489, |
|
"logits/real": -0.3018808662891388, |
|
"logps/generated": -560.4537963867188, |
|
"logps/real": -397.22137451171875, |
|
"loss": 0.3314, |
|
"rewards/accuracies": 0.981249988079071, |
|
"rewards/generated": -1.1572097539901733, |
|
"rewards/margins": 1.1902085542678833, |
|
"rewards/real": 0.03299868851900101, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 31.375, |
|
"learning_rate": 1.3960113960113958e-07, |
|
"logits/generated": 0.192919060587883, |
|
"logits/real": -0.31041043996810913, |
|
"logps/generated": -532.7413330078125, |
|
"logps/real": -462.16693115234375, |
|
"loss": 0.3311, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.1133582592010498, |
|
"rewards/margins": 1.1442750692367554, |
|
"rewards/real": 0.030916890129446983, |
|
"step": 585 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 32.75, |
|
"learning_rate": 1.3603988603988604e-07, |
|
"logits/generated": 0.20842699706554413, |
|
"logits/real": -0.30351361632347107, |
|
"logps/generated": -582.7373657226562, |
|
"logps/real": -480.19073486328125, |
|
"loss": 0.3336, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.1188085079193115, |
|
"rewards/margins": 1.1176929473876953, |
|
"rewards/real": -0.0011154465610161424, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 31.125, |
|
"learning_rate": 1.3247863247863248e-07, |
|
"logits/generated": 0.1861368715763092, |
|
"logits/real": -0.2930435240268707, |
|
"logps/generated": -560.4434814453125, |
|
"logps/real": -442.20574951171875, |
|
"loss": 0.3037, |
|
"rewards/accuracies": 0.9750000238418579, |
|
"rewards/generated": -1.1219825744628906, |
|
"rewards/margins": 1.1991544961929321, |
|
"rewards/real": 0.0771719291806221, |
|
"step": 595 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 32.5, |
|
"learning_rate": 1.289173789173789e-07, |
|
"logits/generated": 0.2440781146287918, |
|
"logits/real": -0.2519104480743408, |
|
"logps/generated": -577.5955810546875, |
|
"logps/real": -426.29083251953125, |
|
"loss": 0.3451, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.0716168880462646, |
|
"rewards/margins": 1.0959482192993164, |
|
"rewards/real": 0.024331340566277504, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"eval_logits/generated": 0.22350193560123444, |
|
"eval_logits/real": -0.28475213050842285, |
|
"eval_logps/generated": -571.62451171875, |
|
"eval_logps/real": -468.817138671875, |
|
"eval_loss": 0.3355465829372406, |
|
"eval_rewards/accuracies": 0.9599999785423279, |
|
"eval_rewards/generated": -1.108030080795288, |
|
"eval_rewards/margins": 1.1036338806152344, |
|
"eval_rewards/real": -0.004396026488393545, |
|
"eval_runtime": 262.6885, |
|
"eval_samples_per_second": 7.614, |
|
"eval_steps_per_second": 0.952, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 31.25, |
|
"learning_rate": 1.2535612535612535e-07, |
|
"logits/generated": 0.26258888840675354, |
|
"logits/real": -0.27824801206588745, |
|
"logps/generated": -587.973876953125, |
|
"logps/real": -470.12603759765625, |
|
"loss": 0.3066, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.2446457147598267, |
|
"rewards/margins": 1.2537733316421509, |
|
"rewards/real": 0.009127420373260975, |
|
"step": 605 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 34.0, |
|
"learning_rate": 1.2179487179487178e-07, |
|
"logits/generated": 0.22428826987743378, |
|
"logits/real": -0.24908749759197235, |
|
"logps/generated": -574.6671752929688, |
|
"logps/real": -423.6082458496094, |
|
"loss": 0.3503, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.131711721420288, |
|
"rewards/margins": 1.1664810180664062, |
|
"rewards/real": 0.03476935625076294, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 36.0, |
|
"learning_rate": 1.1823361823361823e-07, |
|
"logits/generated": 0.2149450033903122, |
|
"logits/real": -0.3522172272205353, |
|
"logps/generated": -569.8680419921875, |
|
"logps/real": -522.5927734375, |
|
"loss": 0.3041, |
|
"rewards/accuracies": 0.987500011920929, |
|
"rewards/generated": -1.102595567703247, |
|
"rewards/margins": 1.1690142154693604, |
|
"rewards/real": 0.06641869246959686, |
|
"step": 615 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 34.0, |
|
"learning_rate": 1.1467236467236467e-07, |
|
"logits/generated": 0.21769335865974426, |
|
"logits/real": -0.3188803493976593, |
|
"logps/generated": -525.9594116210938, |
|
"logps/real": -459.16900634765625, |
|
"loss": 0.3244, |
|
"rewards/accuracies": 0.96875, |
|
"rewards/generated": -1.1512959003448486, |
|
"rewards/margins": 1.1399847269058228, |
|
"rewards/real": -0.011311108246445656, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 37.25, |
|
"learning_rate": 1.111111111111111e-07, |
|
"logits/generated": 0.1936269998550415, |
|
"logits/real": -0.2803877294063568, |
|
"logps/generated": -540.1295166015625, |
|
"logps/real": -444.78375244140625, |
|
"loss": 0.3322, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -1.0809428691864014, |
|
"rewards/margins": 1.1129690408706665, |
|
"rewards/real": 0.03202588111162186, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 29.625, |
|
"learning_rate": 1.0754985754985754e-07, |
|
"logits/generated": 0.19856727123260498, |
|
"logits/real": -0.27393585443496704, |
|
"logps/generated": -560.6895751953125, |
|
"logps/real": -426.6712951660156, |
|
"loss": 0.3194, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.1213538646697998, |
|
"rewards/margins": 1.142456293106079, |
|
"rewards/real": 0.021102434024214745, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 31.75, |
|
"learning_rate": 1.0398860398860399e-07, |
|
"logits/generated": 0.21278016269207, |
|
"logits/real": -0.24991516768932343, |
|
"logps/generated": -587.5706176757812, |
|
"logps/real": -404.28033447265625, |
|
"loss": 0.3189, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.1712287664413452, |
|
"rewards/margins": 1.2088580131530762, |
|
"rewards/real": 0.037629302591085434, |
|
"step": 635 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 31.75, |
|
"learning_rate": 1.0042735042735042e-07, |
|
"logits/generated": 0.22561617195606232, |
|
"logits/real": -0.20534536242485046, |
|
"logps/generated": -575.3756713867188, |
|
"logps/real": -442.997802734375, |
|
"loss": 0.3369, |
|
"rewards/accuracies": 0.96875, |
|
"rewards/generated": -1.1977014541625977, |
|
"rewards/margins": 1.1940991878509521, |
|
"rewards/real": -0.003602252807468176, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 31.875, |
|
"learning_rate": 9.686609686609686e-08, |
|
"logits/generated": 0.2502003610134125, |
|
"logits/real": -0.2464480847120285, |
|
"logps/generated": -574.8699340820312, |
|
"logps/real": -462.56134033203125, |
|
"loss": 0.3468, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -1.0637601613998413, |
|
"rewards/margins": 1.0312858819961548, |
|
"rewards/real": -0.032474346458911896, |
|
"step": 645 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 29.875, |
|
"learning_rate": 9.33048433048433e-08, |
|
"logits/generated": 0.2645814120769501, |
|
"logits/real": -0.2936360836029053, |
|
"logps/generated": -551.9849243164062, |
|
"logps/real": -463.1893005371094, |
|
"loss": 0.3202, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -1.102562665939331, |
|
"rewards/margins": 1.100036382675171, |
|
"rewards/real": -0.002526202006265521, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 33.25, |
|
"learning_rate": 8.974358974358974e-08, |
|
"logits/generated": 0.24026402831077576, |
|
"logits/real": -0.24332275986671448, |
|
"logps/generated": -586.1027221679688, |
|
"logps/real": -457.7566833496094, |
|
"loss": 0.3439, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -1.0856201648712158, |
|
"rewards/margins": 1.0498621463775635, |
|
"rewards/real": -0.035757891833782196, |
|
"step": 655 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 32.75, |
|
"learning_rate": 8.618233618233619e-08, |
|
"logits/generated": 0.22486528754234314, |
|
"logits/real": -0.33010414242744446, |
|
"logps/generated": -575.5369262695312, |
|
"logps/real": -444.61712646484375, |
|
"loss": 0.3306, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1358319520950317, |
|
"rewards/margins": 1.1036947965621948, |
|
"rewards/real": -0.03213699907064438, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 35.0, |
|
"learning_rate": 8.262108262108261e-08, |
|
"logits/generated": 0.22603563964366913, |
|
"logits/real": -0.2633499503135681, |
|
"logps/generated": -588.1068725585938, |
|
"logps/real": -409.90362548828125, |
|
"loss": 0.3222, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.1506733894348145, |
|
"rewards/margins": 1.1825590133666992, |
|
"rewards/real": 0.03188558295369148, |
|
"step": 665 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 32.0, |
|
"learning_rate": 7.905982905982906e-08, |
|
"logits/generated": 0.2575734257698059, |
|
"logits/real": -0.26269882917404175, |
|
"logps/generated": -613.2880859375, |
|
"logps/real": -436.75689697265625, |
|
"loss": 0.3172, |
|
"rewards/accuracies": 0.981249988079071, |
|
"rewards/generated": -1.1868383884429932, |
|
"rewards/margins": 1.1978209018707275, |
|
"rewards/real": 0.010982497595250607, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 36.5, |
|
"learning_rate": 7.549857549857549e-08, |
|
"logits/generated": 0.21627910435199738, |
|
"logits/real": -0.31892699003219604, |
|
"logps/generated": -590.7273559570312, |
|
"logps/real": -460.1044921875, |
|
"loss": 0.3361, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.1508667469024658, |
|
"rewards/margins": 1.1213773488998413, |
|
"rewards/real": -0.029489323496818542, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 30.5, |
|
"learning_rate": 7.193732193732194e-08, |
|
"logits/generated": 0.2537713944911957, |
|
"logits/real": -0.2855969965457916, |
|
"logps/generated": -577.8758544921875, |
|
"logps/real": -475.7315368652344, |
|
"loss": 0.3277, |
|
"rewards/accuracies": 0.9125000238418579, |
|
"rewards/generated": -1.1186813116073608, |
|
"rewards/margins": 1.1131179332733154, |
|
"rewards/real": -0.005563369486480951, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 31.25, |
|
"learning_rate": 6.837606837606839e-08, |
|
"logits/generated": 0.20594389736652374, |
|
"logits/real": -0.31517493724823, |
|
"logps/generated": -553.8355712890625, |
|
"logps/real": -434.3343811035156, |
|
"loss": 0.3286, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.1672719717025757, |
|
"rewards/margins": 1.2395508289337158, |
|
"rewards/real": 0.07227896898984909, |
|
"step": 685 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 32.0, |
|
"learning_rate": 6.481481481481481e-08, |
|
"logits/generated": 0.205632284283638, |
|
"logits/real": -0.255433589220047, |
|
"logps/generated": -570.3898315429688, |
|
"logps/real": -464.691650390625, |
|
"loss": 0.3308, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.0862834453582764, |
|
"rewards/margins": 1.1157293319702148, |
|
"rewards/real": 0.029445823282003403, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 33.75, |
|
"learning_rate": 6.125356125356125e-08, |
|
"logits/generated": 0.23636643588542938, |
|
"logits/real": -0.2590656280517578, |
|
"logps/generated": -598.99755859375, |
|
"logps/real": -467.8699645996094, |
|
"loss": 0.3365, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.104642629623413, |
|
"rewards/margins": 1.1386440992355347, |
|
"rewards/real": 0.034001342952251434, |
|
"step": 695 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 33.5, |
|
"learning_rate": 5.7692307692307695e-08, |
|
"logits/generated": 0.25425681471824646, |
|
"logits/real": -0.2653648853302002, |
|
"logps/generated": -569.2490844726562, |
|
"logps/real": -475.88763427734375, |
|
"loss": 0.3354, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1180486679077148, |
|
"rewards/margins": 1.0991861820220947, |
|
"rewards/real": -0.018862556666135788, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 30.875, |
|
"learning_rate": 5.413105413105413e-08, |
|
"logits/generated": 0.22857370972633362, |
|
"logits/real": -0.22129984200000763, |
|
"logps/generated": -563.016357421875, |
|
"logps/real": -403.68597412109375, |
|
"loss": 0.3157, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.0614380836486816, |
|
"rewards/margins": 1.1257336139678955, |
|
"rewards/real": 0.06429564207792282, |
|
"step": 705 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 36.75, |
|
"learning_rate": 5.056980056980057e-08, |
|
"logits/generated": 0.21650013327598572, |
|
"logits/real": -0.3323083817958832, |
|
"logps/generated": -581.8919677734375, |
|
"logps/real": -487.620361328125, |
|
"loss": 0.3226, |
|
"rewards/accuracies": 0.981249988079071, |
|
"rewards/generated": -1.2059048414230347, |
|
"rewards/margins": 1.1841565370559692, |
|
"rewards/real": -0.021748319268226624, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 30.875, |
|
"learning_rate": 4.7008547008547005e-08, |
|
"logits/generated": 0.2445484846830368, |
|
"logits/real": -0.31186121702194214, |
|
"logps/generated": -557.0134887695312, |
|
"logps/real": -493.17230224609375, |
|
"loss": 0.3212, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1671897172927856, |
|
"rewards/margins": 1.1586859226226807, |
|
"rewards/real": -0.00850372202694416, |
|
"step": 715 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 34.25, |
|
"learning_rate": 4.3447293447293445e-08, |
|
"logits/generated": 0.21253076195716858, |
|
"logits/real": -0.2138686180114746, |
|
"logps/generated": -586.7265625, |
|
"logps/real": -408.2170104980469, |
|
"loss": 0.3419, |
|
"rewards/accuracies": 0.9375, |
|
"rewards/generated": -1.1229650974273682, |
|
"rewards/margins": 1.1265299320220947, |
|
"rewards/real": 0.0035647773183882236, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 29.75, |
|
"learning_rate": 3.9886039886039886e-08, |
|
"logits/generated": 0.18125857412815094, |
|
"logits/real": -0.32901662588119507, |
|
"logps/generated": -537.9093017578125, |
|
"logps/real": -448.9685974121094, |
|
"loss": 0.3366, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.0713820457458496, |
|
"rewards/margins": 1.105197548866272, |
|
"rewards/real": 0.033815376460552216, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 32.25, |
|
"learning_rate": 3.632478632478633e-08, |
|
"logits/generated": 0.21634677052497864, |
|
"logits/real": -0.30626875162124634, |
|
"logps/generated": -582.5755004882812, |
|
"logps/real": -460.1416931152344, |
|
"loss": 0.3136, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1917552947998047, |
|
"rewards/margins": 1.2281749248504639, |
|
"rewards/real": 0.03641948103904724, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"grad_norm": 30.625, |
|
"learning_rate": 3.276353276353276e-08, |
|
"logits/generated": 0.2181752622127533, |
|
"logits/real": -0.34954366087913513, |
|
"logps/generated": -566.8768310546875, |
|
"logps/real": -490.39385986328125, |
|
"loss": 0.3149, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.2017440795898438, |
|
"rewards/margins": 1.2231340408325195, |
|
"rewards/real": 0.0213897917419672, |
|
"step": 735 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 30.75, |
|
"learning_rate": 2.92022792022792e-08, |
|
"logits/generated": 0.250019907951355, |
|
"logits/real": -0.23180410265922546, |
|
"logps/generated": -569.7205810546875, |
|
"logps/real": -436.3052673339844, |
|
"loss": 0.3193, |
|
"rewards/accuracies": 0.9437500238418579, |
|
"rewards/generated": -1.0622925758361816, |
|
"rewards/margins": 1.074755311012268, |
|
"rewards/real": 0.012462759390473366, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 32.0, |
|
"learning_rate": 2.564102564102564e-08, |
|
"logits/generated": 0.2107166051864624, |
|
"logits/real": -0.28946200013160706, |
|
"logps/generated": -562.4656982421875, |
|
"logps/real": -423.35076904296875, |
|
"loss": 0.3288, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1599444150924683, |
|
"rewards/margins": 1.1884777545928955, |
|
"rewards/real": 0.02853330597281456, |
|
"step": 745 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 33.25, |
|
"learning_rate": 2.2079772079772077e-08, |
|
"logits/generated": 0.2310904562473297, |
|
"logits/real": -0.2765614092350006, |
|
"logps/generated": -604.2635498046875, |
|
"logps/real": -450.979248046875, |
|
"loss": 0.3074, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.2296929359436035, |
|
"rewards/margins": 1.2464648485183716, |
|
"rewards/real": 0.016771936789155006, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 31.125, |
|
"learning_rate": 1.8518518518518518e-08, |
|
"logits/generated": 0.17430028319358826, |
|
"logits/real": -0.33389025926589966, |
|
"logps/generated": -549.4244995117188, |
|
"logps/real": -464.75653076171875, |
|
"loss": 0.3085, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.1594291925430298, |
|
"rewards/margins": 1.2221739292144775, |
|
"rewards/real": 0.0627446323633194, |
|
"step": 755 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 32.0, |
|
"learning_rate": 1.4957264957264956e-08, |
|
"logits/generated": 0.24278989434242249, |
|
"logits/real": -0.26737385988235474, |
|
"logps/generated": -573.1634521484375, |
|
"logps/real": -469.896484375, |
|
"loss": 0.3263, |
|
"rewards/accuracies": 0.949999988079071, |
|
"rewards/generated": -1.049228310585022, |
|
"rewards/margins": 1.1048952341079712, |
|
"rewards/real": 0.055667001754045486, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 34.5, |
|
"learning_rate": 1.1396011396011397e-08, |
|
"logits/generated": 0.1711110770702362, |
|
"logits/real": -0.35616591572761536, |
|
"logps/generated": -552.7576293945312, |
|
"logps/real": -433.0294494628906, |
|
"loss": 0.3352, |
|
"rewards/accuracies": 0.96875, |
|
"rewards/generated": -1.0972100496292114, |
|
"rewards/margins": 1.1236265897750854, |
|
"rewards/real": 0.02641637995839119, |
|
"step": 765 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 33.75, |
|
"learning_rate": 7.834757834757834e-09, |
|
"logits/generated": 0.3013153672218323, |
|
"logits/real": -0.22638103365898132, |
|
"logps/generated": -591.3074951171875, |
|
"logps/real": -471.1341857910156, |
|
"loss": 0.3269, |
|
"rewards/accuracies": 0.956250011920929, |
|
"rewards/generated": -1.1128944158554077, |
|
"rewards/margins": 1.088680624961853, |
|
"rewards/real": -0.02421378344297409, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 31.0, |
|
"learning_rate": 4.273504273504274e-09, |
|
"logits/generated": 0.2685278058052063, |
|
"logits/real": -0.22944375872612, |
|
"logps/generated": -570.2601928710938, |
|
"logps/real": -418.1946716308594, |
|
"loss": 0.3241, |
|
"rewards/accuracies": 0.9624999761581421, |
|
"rewards/generated": -1.138912320137024, |
|
"rewards/margins": 1.1833699941635132, |
|
"rewards/real": 0.04445788636803627, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 33.25, |
|
"learning_rate": 7.122507122507123e-10, |
|
"logits/generated": 0.23055453598499298, |
|
"logits/real": -0.3259603977203369, |
|
"logps/generated": -538.0984497070312, |
|
"logps/real": -500.5188903808594, |
|
"loss": 0.3448, |
|
"rewards/accuracies": 0.925000011920929, |
|
"rewards/generated": -1.0663647651672363, |
|
"rewards/margins": 1.0742255449295044, |
|
"rewards/real": 0.007860781624913216, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"step": 781, |
|
"total_flos": 0.0, |
|
"train_loss": 0.41590739681687156, |
|
"train_runtime": 9143.8407, |
|
"train_samples_per_second": 5.468, |
|
"train_steps_per_second": 0.085 |
|
} |
|
], |
|
"logging_steps": 5, |
|
"max_steps": 781, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 1, |
|
"save_steps": 500, |
|
"total_flos": 0.0, |
|
"train_batch_size": 4, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|