| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.9957884885353299, |
| "eval_steps": 500, |
| "global_step": 133, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.007487131492746842, |
| "grad_norm": 183.32981872558594, |
| "learning_rate": 1.9849624060150372e-06, |
| "logits/chosen": 0.32421875, |
| "logits/rejected": 0.2470703125, |
| "logps/chosen": -26.875, |
| "logps/rejected": -25.75, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.014974262985493684, |
| "grad_norm": 162.7139129638672, |
| "learning_rate": 1.969924812030075e-06, |
| "logits/chosen": 0.2734375, |
| "logits/rejected": 0.232421875, |
| "logps/chosen": -27.0, |
| "logps/rejected": -22.25, |
| "loss": 0.6931, |
| "rewards/accuracies": 0.34375, |
| "rewards/chosen": 0.0007781982421875, |
| "rewards/margins": -0.00098419189453125, |
| "rewards/rejected": 0.0017547607421875, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.022461394478240523, |
| "grad_norm": 448.287353515625, |
| "learning_rate": 1.9548872180451127e-06, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.259765625, |
| "logps/chosen": -29.625, |
| "logps/rejected": -232.0, |
| "loss": 0.6897, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.00078582763671875, |
| "rewards/margins": 0.010009765625, |
| "rewards/rejected": -0.00921630859375, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.029948525970987367, |
| "grad_norm": 194.5895538330078, |
| "learning_rate": 1.9398496240601504e-06, |
| "logits/chosen": 0.1767578125, |
| "logits/rejected": 0.267578125, |
| "logps/chosen": -29.75, |
| "logps/rejected": -20.375, |
| "loss": 0.681, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.01153564453125, |
| "rewards/margins": 0.02490234375, |
| "rewards/rejected": -0.0133056640625, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.03743565746373421, |
| "grad_norm": 147.04005432128906, |
| "learning_rate": 1.9248120300751877e-06, |
| "logits/chosen": 0.349609375, |
| "logits/rejected": 0.275390625, |
| "logps/chosen": -25.375, |
| "logps/rejected": -25.875, |
| "loss": 0.6776, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.0242919921875, |
| "rewards/margins": 0.0311279296875, |
| "rewards/rejected": -0.0068359375, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.044922788956481045, |
| "grad_norm": 146.45497131347656, |
| "learning_rate": 1.9097744360902255e-06, |
| "logits/chosen": 0.216796875, |
| "logits/rejected": 0.240234375, |
| "logps/chosen": -28.875, |
| "logps/rejected": -24.5, |
| "loss": 0.6782, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.0211181640625, |
| "rewards/margins": 0.0299072265625, |
| "rewards/rejected": -0.0087890625, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.05240992044922789, |
| "grad_norm": 167.77880859375, |
| "learning_rate": 1.894736842105263e-06, |
| "logits/chosen": 0.33203125, |
| "logits/rejected": 0.275390625, |
| "logps/chosen": -28.0, |
| "logps/rejected": -19.875, |
| "loss": 0.6752, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.03173828125, |
| "rewards/margins": 0.03662109375, |
| "rewards/rejected": -0.00506591796875, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.059897051941974734, |
| "grad_norm": 264.6982727050781, |
| "learning_rate": 1.8796992481203007e-06, |
| "logits/chosen": 0.21875, |
| "logits/rejected": 0.18359375, |
| "logps/chosen": -32.0, |
| "logps/rejected": -87.5, |
| "loss": 0.6571, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06640625, |
| "rewards/margins": 0.0771484375, |
| "rewards/rejected": -0.0103759765625, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.06738418343472158, |
| "grad_norm": 267.3852233886719, |
| "learning_rate": 1.8646616541353382e-06, |
| "logits/chosen": 0.205078125, |
| "logits/rejected": 0.1259765625, |
| "logps/chosen": -28.125, |
| "logps/rejected": -118.5, |
| "loss": 0.6501, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.06982421875, |
| "rewards/margins": 0.09228515625, |
| "rewards/rejected": -0.0225830078125, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.07487131492746842, |
| "grad_norm": 167.42930603027344, |
| "learning_rate": 1.849624060150376e-06, |
| "logits/chosen": 0.203125, |
| "logits/rejected": 0.1416015625, |
| "logps/chosen": -27.875, |
| "logps/rejected": -210.0, |
| "loss": 0.6128, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.07275390625, |
| "rewards/margins": 0.1962890625, |
| "rewards/rejected": -0.12353515625, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.08235844642021525, |
| "grad_norm": 173.68484497070312, |
| "learning_rate": 1.8345864661654133e-06, |
| "logits/chosen": 0.3125, |
| "logits/rejected": 0.2158203125, |
| "logps/chosen": -24.625, |
| "logps/rejected": -199.0, |
| "loss": 0.6319, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05712890625, |
| "rewards/margins": 0.1689453125, |
| "rewards/rejected": -0.1123046875, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.08984557791296209, |
| "grad_norm": 147.03286743164062, |
| "learning_rate": 1.819548872180451e-06, |
| "logits/chosen": 0.29296875, |
| "logits/rejected": 0.150390625, |
| "logps/chosen": -26.25, |
| "logps/rejected": -28.0, |
| "loss": 0.6515, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.06787109375, |
| "rewards/margins": 0.091796875, |
| "rewards/rejected": -0.0240478515625, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.09733270940570894, |
| "grad_norm": 163.2376251220703, |
| "learning_rate": 1.8045112781954885e-06, |
| "logits/chosen": 0.2001953125, |
| "logits/rejected": 0.2734375, |
| "logps/chosen": -28.125, |
| "logps/rejected": -143.0, |
| "loss": 0.6138, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.0927734375, |
| "rewards/margins": 0.208984375, |
| "rewards/rejected": -0.11669921875, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.10481984089845578, |
| "grad_norm": 149.04415893554688, |
| "learning_rate": 1.7894736842105262e-06, |
| "logits/chosen": 0.3125, |
| "logits/rejected": 0.1279296875, |
| "logps/chosen": -23.5, |
| "logps/rejected": -21.875, |
| "loss": 0.633, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.11181640625, |
| "rewards/margins": 0.12890625, |
| "rewards/rejected": -0.017333984375, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.11230697239120262, |
| "grad_norm": 144.06060791015625, |
| "learning_rate": 1.7744360902255638e-06, |
| "logits/chosen": 0.2314453125, |
| "logits/rejected": 0.23828125, |
| "logps/chosen": -27.0, |
| "logps/rejected": -21.5, |
| "loss": 0.616, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.142578125, |
| "rewards/margins": 0.1640625, |
| "rewards/rejected": -0.022216796875, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.11979410388394947, |
| "grad_norm": 138.48106384277344, |
| "learning_rate": 1.7593984962406015e-06, |
| "logits/chosen": 0.296875, |
| "logits/rejected": 0.1923828125, |
| "logps/chosen": -27.0, |
| "logps/rejected": -151.0, |
| "loss": 0.5917, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.111328125, |
| "rewards/margins": 0.30859375, |
| "rewards/rejected": -0.197265625, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.1272812353766963, |
| "grad_norm": 125.52348327636719, |
| "learning_rate": 1.744360902255639e-06, |
| "logits/chosen": 0.265625, |
| "logits/rejected": 0.271484375, |
| "logps/chosen": -26.5, |
| "logps/rejected": -41.75, |
| "loss": 0.6047, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1328125, |
| "rewards/margins": 0.2158203125, |
| "rewards/rejected": -0.08251953125, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.13476836686944316, |
| "grad_norm": 135.62704467773438, |
| "learning_rate": 1.7293233082706765e-06, |
| "logits/chosen": 0.2275390625, |
| "logits/rejected": 0.234375, |
| "logps/chosen": -25.75, |
| "logps/rejected": -116.5, |
| "loss": 0.6024, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1416015625, |
| "rewards/margins": 0.25390625, |
| "rewards/rejected": -0.1123046875, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.14225549836218998, |
| "grad_norm": 135.725830078125, |
| "learning_rate": 1.714285714285714e-06, |
| "logits/chosen": 0.265625, |
| "logits/rejected": 0.2021484375, |
| "logps/chosen": -24.75, |
| "logps/rejected": -78.5, |
| "loss": 0.6002, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1259765625, |
| "rewards/margins": 0.263671875, |
| "rewards/rejected": -0.138671875, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.14974262985493683, |
| "grad_norm": 143.98516845703125, |
| "learning_rate": 1.6992481203007518e-06, |
| "logits/chosen": 0.1904296875, |
| "logits/rejected": 0.14453125, |
| "logps/chosen": -29.0, |
| "logps/rejected": -40.25, |
| "loss": 0.5717, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.193359375, |
| "rewards/margins": 0.27734375, |
| "rewards/rejected": -0.08447265625, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.15722976134768366, |
| "grad_norm": 120.902099609375, |
| "learning_rate": 1.6842105263157893e-06, |
| "logits/chosen": 0.23046875, |
| "logits/rejected": 0.2412109375, |
| "logps/chosen": -25.375, |
| "logps/rejected": -18.875, |
| "loss": 0.6137, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1708984375, |
| "rewards/margins": 0.171875, |
| "rewards/rejected": -0.001556396484375, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.1647168928404305, |
| "grad_norm": 124.2569580078125, |
| "learning_rate": 1.669172932330827e-06, |
| "logits/chosen": 0.337890625, |
| "logits/rejected": 0.171875, |
| "logps/chosen": -25.25, |
| "logps/rejected": -22.75, |
| "loss": 0.6059, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1279296875, |
| "rewards/margins": 0.1943359375, |
| "rewards/rejected": -0.06591796875, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.17220402433317736, |
| "grad_norm": 129.59535217285156, |
| "learning_rate": 1.6541353383458646e-06, |
| "logits/chosen": 0.1884765625, |
| "logits/rejected": 0.12255859375, |
| "logps/chosen": -28.375, |
| "logps/rejected": -26.75, |
| "loss": 0.5615, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2021484375, |
| "rewards/margins": 0.30078125, |
| "rewards/rejected": -0.09912109375, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.17969115582592418, |
| "grad_norm": 119.532958984375, |
| "learning_rate": 1.6390977443609023e-06, |
| "logits/chosen": 0.283203125, |
| "logits/rejected": 0.1279296875, |
| "logps/chosen": -24.25, |
| "logps/rejected": -37.5, |
| "loss": 0.5599, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1689453125, |
| "rewards/margins": 0.322265625, |
| "rewards/rejected": -0.1533203125, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.18717828731867103, |
| "grad_norm": 105.28582763671875, |
| "learning_rate": 1.6240601503759396e-06, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.2451171875, |
| "logps/chosen": -24.75, |
| "logps/rejected": -25.375, |
| "loss": 0.593, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1650390625, |
| "rewards/margins": 0.2431640625, |
| "rewards/rejected": -0.078125, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.19466541881141788, |
| "grad_norm": 123.53404998779297, |
| "learning_rate": 1.6090225563909773e-06, |
| "logits/chosen": 0.275390625, |
| "logits/rejected": 0.1298828125, |
| "logps/chosen": -25.5, |
| "logps/rejected": -160.0, |
| "loss": 0.5258, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2099609375, |
| "rewards/margins": 0.55078125, |
| "rewards/rejected": -0.33984375, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.2021525503041647, |
| "grad_norm": 122.69832611083984, |
| "learning_rate": 1.5939849624060149e-06, |
| "logits/chosen": 0.162109375, |
| "logits/rejected": 0.310546875, |
| "logps/chosen": -28.625, |
| "logps/rejected": -21.0, |
| "loss": 0.595, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1875, |
| "rewards/margins": 0.2119140625, |
| "rewards/rejected": -0.024658203125, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.20963968179691156, |
| "grad_norm": 107.87254333496094, |
| "learning_rate": 1.5789473684210526e-06, |
| "logits/chosen": 0.2490234375, |
| "logits/rejected": 0.11083984375, |
| "logps/chosen": -25.25, |
| "logps/rejected": -127.0, |
| "loss": 0.5614, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1826171875, |
| "rewards/margins": 0.484375, |
| "rewards/rejected": -0.30078125, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.2171268132896584, |
| "grad_norm": 121.16978454589844, |
| "learning_rate": 1.56390977443609e-06, |
| "logits/chosen": 0.2275390625, |
| "logits/rejected": 0.1474609375, |
| "logps/chosen": -26.5, |
| "logps/rejected": -179.0, |
| "loss": 0.4801, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.251953125, |
| "rewards/margins": 0.75, |
| "rewards/rejected": -0.5, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.22461394478240523, |
| "grad_norm": 127.31982421875, |
| "learning_rate": 1.5488721804511278e-06, |
| "logits/chosen": 0.275390625, |
| "logits/rejected": 0.240234375, |
| "logps/chosen": -27.125, |
| "logps/rejected": -84.0, |
| "loss": 0.5083, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.283203125, |
| "rewards/margins": 0.578125, |
| "rewards/rejected": -0.294921875, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.23210107627515209, |
| "grad_norm": 102.35781860351562, |
| "learning_rate": 1.5338345864661654e-06, |
| "logits/chosen": 0.255859375, |
| "logits/rejected": 0.1728515625, |
| "logps/chosen": -25.875, |
| "logps/rejected": -33.0, |
| "loss": 0.5519, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2373046875, |
| "rewards/margins": 0.35546875, |
| "rewards/rejected": -0.11767578125, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.23958820776789894, |
| "grad_norm": 122.3163070678711, |
| "learning_rate": 1.5187969924812029e-06, |
| "logits/chosen": 0.166015625, |
| "logits/rejected": 0.1787109375, |
| "logps/chosen": -29.125, |
| "logps/rejected": -121.5, |
| "loss": 0.4974, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.26171875, |
| "rewards/margins": 0.6328125, |
| "rewards/rejected": -0.373046875, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.24707533926064576, |
| "grad_norm": 114.36420440673828, |
| "learning_rate": 1.5037593984962404e-06, |
| "logits/chosen": 0.2734375, |
| "logits/rejected": 0.263671875, |
| "logps/chosen": -24.5, |
| "logps/rejected": -26.25, |
| "loss": 0.5807, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.173828125, |
| "rewards/margins": 0.26171875, |
| "rewards/rejected": -0.08837890625, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.2545624707533926, |
| "grad_norm": 103.6934585571289, |
| "learning_rate": 1.4887218045112781e-06, |
| "logits/chosen": 0.1328125, |
| "logits/rejected": 0.051513671875, |
| "logps/chosen": -27.75, |
| "logps/rejected": -101.5, |
| "loss": 0.4898, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2412109375, |
| "rewards/margins": 0.65625, |
| "rewards/rejected": -0.416015625, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.26204960224613943, |
| "grad_norm": 95.02819061279297, |
| "learning_rate": 1.4736842105263156e-06, |
| "logits/chosen": 0.1884765625, |
| "logits/rejected": 0.11181640625, |
| "logps/chosen": -27.625, |
| "logps/rejected": -29.375, |
| "loss": 0.5312, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.232421875, |
| "rewards/margins": 0.41015625, |
| "rewards/rejected": -0.177734375, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.2695367337388863, |
| "grad_norm": 94.60529327392578, |
| "learning_rate": 1.4586466165413534e-06, |
| "logits/chosen": 0.2373046875, |
| "logits/rejected": 0.1474609375, |
| "logps/chosen": -24.875, |
| "logps/rejected": -38.75, |
| "loss": 0.5326, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.263671875, |
| "rewards/margins": 0.439453125, |
| "rewards/rejected": -0.17578125, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.27702386523163314, |
| "grad_norm": 97.94026947021484, |
| "learning_rate": 1.443609022556391e-06, |
| "logits/chosen": 0.21484375, |
| "logits/rejected": 0.197265625, |
| "logps/chosen": -27.0, |
| "logps/rejected": -90.0, |
| "loss": 0.5326, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.267578125, |
| "rewards/margins": 0.52734375, |
| "rewards/rejected": -0.26171875, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.28451099672437996, |
| "grad_norm": 84.70133209228516, |
| "learning_rate": 1.4285714285714286e-06, |
| "logits/chosen": 0.2109375, |
| "logits/rejected": 0.23046875, |
| "logps/chosen": -25.625, |
| "logps/rejected": -39.5, |
| "loss": 0.5299, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.26171875, |
| "rewards/margins": 0.47265625, |
| "rewards/rejected": -0.2109375, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.29199812821712684, |
| "grad_norm": 108.93930053710938, |
| "learning_rate": 1.413533834586466e-06, |
| "logits/chosen": 0.1904296875, |
| "logits/rejected": 0.0947265625, |
| "logps/chosen": -27.375, |
| "logps/rejected": -27.625, |
| "loss": 0.5089, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.3203125, |
| "rewards/margins": 0.474609375, |
| "rewards/rejected": -0.154296875, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.29948525970987366, |
| "grad_norm": 95.77433013916016, |
| "learning_rate": 1.3984962406015037e-06, |
| "logits/chosen": 0.23828125, |
| "logits/rejected": 0.267578125, |
| "logps/chosen": -27.875, |
| "logps/rejected": -20.75, |
| "loss": 0.5214, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.29296875, |
| "rewards/margins": 0.404296875, |
| "rewards/rejected": -0.11181640625, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.3069723912026205, |
| "grad_norm": 93.18071746826172, |
| "learning_rate": 1.3834586466165412e-06, |
| "logits/chosen": 0.2490234375, |
| "logits/rejected": 0.205078125, |
| "logps/chosen": -24.0, |
| "logps/rejected": -28.25, |
| "loss": 0.5503, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2060546875, |
| "rewards/margins": 0.380859375, |
| "rewards/rejected": -0.173828125, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.3144595226953673, |
| "grad_norm": 109.81500244140625, |
| "learning_rate": 1.368421052631579e-06, |
| "logits/chosen": 0.185546875, |
| "logits/rejected": 0.240234375, |
| "logps/chosen": -26.5, |
| "logps/rejected": -34.25, |
| "loss": 0.495, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.33984375, |
| "rewards/margins": 0.55859375, |
| "rewards/rejected": -0.2177734375, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.3219466541881142, |
| "grad_norm": 101.38113403320312, |
| "learning_rate": 1.3533834586466164e-06, |
| "logits/chosen": 0.28125, |
| "logits/rejected": 0.10693359375, |
| "logps/chosen": -24.0, |
| "logps/rejected": -116.5, |
| "loss": 0.4809, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.25390625, |
| "rewards/margins": 0.7421875, |
| "rewards/rejected": -0.490234375, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.329433785680861, |
| "grad_norm": 81.20439147949219, |
| "learning_rate": 1.3383458646616542e-06, |
| "logits/chosen": 0.279296875, |
| "logits/rejected": 0.283203125, |
| "logps/chosen": -24.625, |
| "logps/rejected": -19.75, |
| "loss": 0.5115, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.328125, |
| "rewards/margins": 0.451171875, |
| "rewards/rejected": -0.12353515625, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.33692091717360784, |
| "grad_norm": 72.38237762451172, |
| "learning_rate": 1.3233082706766917e-06, |
| "logits/chosen": 0.1865234375, |
| "logits/rejected": 0.146484375, |
| "logps/chosen": -26.0, |
| "logps/rejected": -86.0, |
| "loss": 0.4921, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.263671875, |
| "rewards/margins": 0.765625, |
| "rewards/rejected": -0.5, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.3444080486663547, |
| "grad_norm": 75.40729522705078, |
| "learning_rate": 1.3082706766917292e-06, |
| "logits/chosen": 0.1962890625, |
| "logits/rejected": 0.12158203125, |
| "logps/chosen": -26.125, |
| "logps/rejected": -141.0, |
| "loss": 0.4922, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.259765625, |
| "rewards/margins": 0.80078125, |
| "rewards/rejected": -0.54296875, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.35189518015910154, |
| "grad_norm": 80.21080017089844, |
| "learning_rate": 1.2932330827067667e-06, |
| "logits/chosen": 0.251953125, |
| "logits/rejected": 0.1669921875, |
| "logps/chosen": -25.0, |
| "logps/rejected": -57.5, |
| "loss": 0.5186, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2294921875, |
| "rewards/margins": 0.51953125, |
| "rewards/rejected": -0.291015625, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.35938231165184836, |
| "grad_norm": 76.38800048828125, |
| "learning_rate": 1.2781954887218045e-06, |
| "logits/chosen": 0.3203125, |
| "logits/rejected": 0.19921875, |
| "logps/chosen": -22.25, |
| "logps/rejected": -116.5, |
| "loss": 0.5045, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.228515625, |
| "rewards/margins": 0.6796875, |
| "rewards/rejected": -0.451171875, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.36686944314459524, |
| "grad_norm": 84.38673400878906, |
| "learning_rate": 1.263157894736842e-06, |
| "logits/chosen": 0.2197265625, |
| "logits/rejected": 0.125, |
| "logps/chosen": -22.375, |
| "logps/rejected": -91.5, |
| "loss": 0.4797, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.234375, |
| "rewards/margins": 0.7265625, |
| "rewards/rejected": -0.4921875, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.37435657463734207, |
| "grad_norm": 79.1534423828125, |
| "learning_rate": 1.2481203007518797e-06, |
| "logits/chosen": 0.205078125, |
| "logits/rejected": 0.21484375, |
| "logps/chosen": -25.125, |
| "logps/rejected": -128.0, |
| "loss": 0.4809, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.326171875, |
| "rewards/margins": 0.79296875, |
| "rewards/rejected": -0.466796875, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3818437061300889, |
| "grad_norm": 95.26607513427734, |
| "learning_rate": 1.2330827067669172e-06, |
| "logits/chosen": 0.244140625, |
| "logits/rejected": 0.06494140625, |
| "logps/chosen": -24.0, |
| "logps/rejected": -169.0, |
| "loss": 0.414, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3515625, |
| "rewards/margins": 1.2109375, |
| "rewards/rejected": -0.86328125, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.38933083762283577, |
| "grad_norm": 79.70458221435547, |
| "learning_rate": 1.218045112781955e-06, |
| "logits/chosen": 0.185546875, |
| "logits/rejected": 0.050048828125, |
| "logps/chosen": -25.875, |
| "logps/rejected": -27.75, |
| "loss": 0.4711, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.29296875, |
| "rewards/margins": 0.62109375, |
| "rewards/rejected": -0.328125, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.3968179691155826, |
| "grad_norm": 78.77261352539062, |
| "learning_rate": 1.2030075187969923e-06, |
| "logits/chosen": 0.263671875, |
| "logits/rejected": 0.099609375, |
| "logps/chosen": -24.25, |
| "logps/rejected": -32.5, |
| "loss": 0.4703, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2734375, |
| "rewards/margins": 0.61328125, |
| "rewards/rejected": -0.33984375, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.4043051006083294, |
| "grad_norm": 64.24870300292969, |
| "learning_rate": 1.18796992481203e-06, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.2099609375, |
| "logps/chosen": -23.375, |
| "logps/rejected": -24.5, |
| "loss": 0.5525, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1689453125, |
| "rewards/margins": 0.37109375, |
| "rewards/rejected": -0.203125, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.4117922321010763, |
| "grad_norm": 71.2979965209961, |
| "learning_rate": 1.1729323308270675e-06, |
| "logits/chosen": 0.287109375, |
| "logits/rejected": 0.3125, |
| "logps/chosen": -25.625, |
| "logps/rejected": -22.0, |
| "loss": 0.498, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.279296875, |
| "rewards/margins": 0.486328125, |
| "rewards/rejected": -0.2080078125, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.4192793635938231, |
| "grad_norm": 78.7899169921875, |
| "learning_rate": 1.1578947368421053e-06, |
| "logits/chosen": 0.255859375, |
| "logits/rejected": 0.232421875, |
| "logps/chosen": -24.125, |
| "logps/rejected": -98.0, |
| "loss": 0.5037, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.232421875, |
| "rewards/margins": 0.671875, |
| "rewards/rejected": -0.4375, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.42676649508656994, |
| "grad_norm": 74.6025619506836, |
| "learning_rate": 1.1428571428571428e-06, |
| "logits/chosen": 0.23828125, |
| "logits/rejected": 0.224609375, |
| "logps/chosen": -24.25, |
| "logps/rejected": -24.625, |
| "loss": 0.4781, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2890625, |
| "rewards/margins": 0.56640625, |
| "rewards/rejected": -0.279296875, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.4342536265793168, |
| "grad_norm": 77.2677993774414, |
| "learning_rate": 1.1278195488721805e-06, |
| "logits/chosen": 0.205078125, |
| "logits/rejected": 0.1640625, |
| "logps/chosen": -24.75, |
| "logps/rejected": -29.25, |
| "loss": 0.4841, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.349609375, |
| "rewards/margins": 0.62109375, |
| "rewards/rejected": -0.271484375, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.44174075807206364, |
| "grad_norm": 72.21055603027344, |
| "learning_rate": 1.112781954887218e-06, |
| "logits/chosen": 0.212890625, |
| "logits/rejected": 0.126953125, |
| "logps/chosen": -28.875, |
| "logps/rejected": -26.5, |
| "loss": 0.433, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.373046875, |
| "rewards/margins": 0.71875, |
| "rewards/rejected": -0.34375, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.44922788956481047, |
| "grad_norm": 59.05992889404297, |
| "learning_rate": 1.0977443609022555e-06, |
| "logits/chosen": 0.28515625, |
| "logits/rejected": 0.2236328125, |
| "logps/chosen": -23.75, |
| "logps/rejected": -234.0, |
| "loss": 0.4477, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.19921875, |
| "rewards/margins": 1.296875, |
| "rewards/rejected": -1.1015625, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.45671502105755735, |
| "grad_norm": 71.27205657958984, |
| "learning_rate": 1.082706766917293e-06, |
| "logits/chosen": 0.1806640625, |
| "logits/rejected": 0.228515625, |
| "logps/chosen": -23.375, |
| "logps/rejected": -20.625, |
| "loss": 0.5128, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.23828125, |
| "rewards/margins": 0.44921875, |
| "rewards/rejected": -0.2109375, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.46420215255030417, |
| "grad_norm": 66.39558410644531, |
| "learning_rate": 1.0676691729323308e-06, |
| "logits/chosen": 0.294921875, |
| "logits/rejected": 0.29296875, |
| "logps/chosen": -24.125, |
| "logps/rejected": -60.5, |
| "loss": 0.4676, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.28125, |
| "rewards/margins": 0.73046875, |
| "rewards/rejected": -0.44921875, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.471689284043051, |
| "grad_norm": 74.42204284667969, |
| "learning_rate": 1.0526315789473683e-06, |
| "logits/chosen": 0.30078125, |
| "logits/rejected": 0.22265625, |
| "logps/chosen": -24.875, |
| "logps/rejected": -37.5, |
| "loss": 0.4886, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.259765625, |
| "rewards/margins": 0.59375, |
| "rewards/rejected": -0.333984375, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.4791764155357979, |
| "grad_norm": 60.48930358886719, |
| "learning_rate": 1.037593984962406e-06, |
| "logits/chosen": 0.2490234375, |
| "logits/rejected": 0.298828125, |
| "logps/chosen": -27.125, |
| "logps/rejected": -183.0, |
| "loss": 0.4197, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.462890625, |
| "rewards/margins": 1.3046875, |
| "rewards/rejected": -0.84375, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.4866635470285447, |
| "grad_norm": 68.27130126953125, |
| "learning_rate": 1.0225563909774436e-06, |
| "logits/chosen": 0.2138671875, |
| "logits/rejected": 0.25390625, |
| "logps/chosen": -28.125, |
| "logps/rejected": -38.5, |
| "loss": 0.4482, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.373046875, |
| "rewards/margins": 0.7734375, |
| "rewards/rejected": -0.400390625, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.4941506785212915, |
| "grad_norm": 61.67387771606445, |
| "learning_rate": 1.0075187969924813e-06, |
| "logits/chosen": 0.244140625, |
| "logits/rejected": 0.1494140625, |
| "logps/chosen": -25.875, |
| "logps/rejected": -146.0, |
| "loss": 0.4096, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.375, |
| "rewards/margins": 1.3125, |
| "rewards/rejected": -0.9375, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.5016378100140384, |
| "grad_norm": 76.5013656616211, |
| "learning_rate": 9.924812030075186e-07, |
| "logits/chosen": 0.275390625, |
| "logits/rejected": 0.212890625, |
| "logps/chosen": -23.5, |
| "logps/rejected": -41.25, |
| "loss": 0.4666, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.22265625, |
| "rewards/margins": 0.66015625, |
| "rewards/rejected": -0.4375, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.5091249415067852, |
| "grad_norm": 50.46118927001953, |
| "learning_rate": 9.774436090225563e-07, |
| "logits/chosen": 0.306640625, |
| "logits/rejected": 0.15234375, |
| "logps/chosen": -25.5, |
| "logps/rejected": -205.0, |
| "loss": 0.4047, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.32421875, |
| "rewards/margins": 1.609375, |
| "rewards/rejected": -1.2890625, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.516612072999532, |
| "grad_norm": 63.7412223815918, |
| "learning_rate": 9.624060150375939e-07, |
| "logits/chosen": 0.162109375, |
| "logits/rejected": 0.216796875, |
| "logps/chosen": -27.625, |
| "logps/rejected": -30.0, |
| "loss": 0.4645, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.306640625, |
| "rewards/margins": 0.70703125, |
| "rewards/rejected": -0.40234375, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.5240992044922789, |
| "grad_norm": 52.42015838623047, |
| "learning_rate": 9.473684210526315e-07, |
| "logits/chosen": 0.2265625, |
| "logits/rejected": 0.11865234375, |
| "logps/chosen": -25.875, |
| "logps/rejected": -117.5, |
| "loss": 0.4393, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2734375, |
| "rewards/margins": 1.078125, |
| "rewards/rejected": -0.80859375, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.5315863359850257, |
| "grad_norm": 79.23849487304688, |
| "learning_rate": 9.323308270676691e-07, |
| "logits/chosen": 0.275390625, |
| "logits/rejected": 0.1923828125, |
| "logps/chosen": -24.75, |
| "logps/rejected": -160.0, |
| "loss": 0.4218, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.353515625, |
| "rewards/margins": 1.4765625, |
| "rewards/rejected": -1.125, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.5390734674777726, |
| "grad_norm": 55.779964447021484, |
| "learning_rate": 9.172932330827066e-07, |
| "logits/chosen": 0.318359375, |
| "logits/rejected": 0.185546875, |
| "logps/chosen": -24.375, |
| "logps/rejected": -33.75, |
| "loss": 0.43, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2177734375, |
| "rewards/margins": 0.84765625, |
| "rewards/rejected": -0.62890625, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.5465605989705195, |
| "grad_norm": 60.20227813720703, |
| "learning_rate": 9.022556390977443e-07, |
| "logits/chosen": 0.2080078125, |
| "logits/rejected": 0.1767578125, |
| "logps/chosen": -26.375, |
| "logps/rejected": -139.0, |
| "loss": 0.3933, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3828125, |
| "rewards/margins": 1.2578125, |
| "rewards/rejected": -0.875, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.5540477304632663, |
| "grad_norm": 51.512977600097656, |
| "learning_rate": 8.872180451127819e-07, |
| "logits/chosen": 0.23046875, |
| "logits/rejected": 0.09423828125, |
| "logps/chosen": -25.5, |
| "logps/rejected": -75.5, |
| "loss": 0.3989, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.263671875, |
| "rewards/margins": 1.0859375, |
| "rewards/rejected": -0.82421875, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.5615348619560131, |
| "grad_norm": 67.2686538696289, |
| "learning_rate": 8.721804511278195e-07, |
| "logits/chosen": 0.3125, |
| "logits/rejected": 0.2041015625, |
| "logps/chosen": -24.125, |
| "logps/rejected": -256.0, |
| "loss": 0.4315, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.21484375, |
| "rewards/margins": 1.546875, |
| "rewards/rejected": -1.3359375, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.5690219934487599, |
| "grad_norm": 70.67787170410156, |
| "learning_rate": 8.57142857142857e-07, |
| "logits/chosen": 0.216796875, |
| "logits/rejected": 0.1513671875, |
| "logps/chosen": -25.0, |
| "logps/rejected": -109.0, |
| "loss": 0.3992, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.26953125, |
| "rewards/margins": 1.265625, |
| "rewards/rejected": -1.0, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.5765091249415067, |
| "grad_norm": 51.3372917175293, |
| "learning_rate": 8.421052631578947e-07, |
| "logits/chosen": 0.185546875, |
| "logits/rejected": 0.2353515625, |
| "logps/chosen": -25.25, |
| "logps/rejected": -178.0, |
| "loss": 0.4455, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21484375, |
| "rewards/margins": 1.2890625, |
| "rewards/rejected": -1.0703125, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.5839962564342537, |
| "grad_norm": 49.59202194213867, |
| "learning_rate": 8.270676691729323e-07, |
| "logits/chosen": 0.298828125, |
| "logits/rejected": 0.265625, |
| "logps/chosen": -26.0, |
| "logps/rejected": -32.25, |
| "loss": 0.4116, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.37109375, |
| "rewards/margins": 0.87890625, |
| "rewards/rejected": -0.5078125, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.5914833879270005, |
| "grad_norm": 52.285152435302734, |
| "learning_rate": 8.120300751879698e-07, |
| "logits/chosen": 0.2431640625, |
| "logits/rejected": 0.271484375, |
| "logps/chosen": -25.75, |
| "logps/rejected": -86.5, |
| "loss": 0.4358, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.314453125, |
| "rewards/margins": 1.015625, |
| "rewards/rejected": -0.69921875, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.5989705194197473, |
| "grad_norm": 52.016822814941406, |
| "learning_rate": 7.969924812030074e-07, |
| "logits/chosen": 0.25390625, |
| "logits/rejected": 0.1826171875, |
| "logps/chosen": -24.875, |
| "logps/rejected": -54.0, |
| "loss": 0.4305, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.25390625, |
| "rewards/margins": 1.0, |
| "rewards/rejected": -0.7421875, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.6064576509124942, |
| "grad_norm": 50.19512176513672, |
| "learning_rate": 7.81954887218045e-07, |
| "logits/chosen": 0.27734375, |
| "logits/rejected": 0.15625, |
| "logps/chosen": -23.5, |
| "logps/rejected": -95.0, |
| "loss": 0.4684, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.158203125, |
| "rewards/margins": 0.87890625, |
| "rewards/rejected": -0.71875, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.613944782405241, |
| "grad_norm": 85.73662567138672, |
| "learning_rate": 7.669172932330827e-07, |
| "logits/chosen": 0.1982421875, |
| "logits/rejected": 0.2265625, |
| "logps/chosen": -26.75, |
| "logps/rejected": -27.25, |
| "loss": 0.4545, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.361328125, |
| "rewards/margins": 0.7265625, |
| "rewards/rejected": -0.365234375, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.6214319138979878, |
| "grad_norm": 61.057830810546875, |
| "learning_rate": 7.518796992481202e-07, |
| "logits/chosen": 0.19140625, |
| "logits/rejected": 0.1318359375, |
| "logps/chosen": -27.375, |
| "logps/rejected": -32.5, |
| "loss": 0.4299, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2119140625, |
| "rewards/margins": 0.7578125, |
| "rewards/rejected": -0.54296875, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.6289190453907346, |
| "grad_norm": 53.54865264892578, |
| "learning_rate": 7.368421052631578e-07, |
| "logits/chosen": 0.25390625, |
| "logits/rejected": 0.10595703125, |
| "logps/chosen": -25.0, |
| "logps/rejected": -86.5, |
| "loss": 0.3744, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.322265625, |
| "rewards/margins": 1.421875, |
| "rewards/rejected": -1.1015625, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.6364061768834816, |
| "grad_norm": 48.02566146850586, |
| "learning_rate": 7.218045112781954e-07, |
| "logits/chosen": 0.310546875, |
| "logits/rejected": 0.193359375, |
| "logps/chosen": -23.75, |
| "logps/rejected": -114.5, |
| "loss": 0.4171, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.265625, |
| "rewards/margins": 1.171875, |
| "rewards/rejected": -0.91015625, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.6438933083762284, |
| "grad_norm": 42.602638244628906, |
| "learning_rate": 7.06766917293233e-07, |
| "logits/chosen": 0.333984375, |
| "logits/rejected": 0.1318359375, |
| "logps/chosen": -24.25, |
| "logps/rejected": -45.0, |
| "loss": 0.3654, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.322265625, |
| "rewards/margins": 1.1796875, |
| "rewards/rejected": -0.859375, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.6513804398689752, |
| "grad_norm": 50.09026336669922, |
| "learning_rate": 6.917293233082706e-07, |
| "logits/chosen": 0.265625, |
| "logits/rejected": 0.1953125, |
| "logps/chosen": -24.5, |
| "logps/rejected": -31.625, |
| "loss": 0.4452, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1962890625, |
| "rewards/margins": 0.80859375, |
| "rewards/rejected": -0.61328125, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.658867571361722, |
| "grad_norm": 46.998741149902344, |
| "learning_rate": 6.766917293233082e-07, |
| "logits/chosen": 0.330078125, |
| "logits/rejected": 0.2158203125, |
| "logps/chosen": -23.375, |
| "logps/rejected": -27.625, |
| "loss": 0.4262, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.302734375, |
| "rewards/margins": 0.765625, |
| "rewards/rejected": -0.46484375, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.6663547028544688, |
| "grad_norm": 69.94436645507812, |
| "learning_rate": 6.616541353383458e-07, |
| "logits/chosen": 0.1884765625, |
| "logits/rejected": 0.064453125, |
| "logps/chosen": -25.75, |
| "logps/rejected": -53.5, |
| "loss": 0.4273, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.234375, |
| "rewards/margins": 1.03125, |
| "rewards/rejected": -0.796875, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.6738418343472157, |
| "grad_norm": 51.591617584228516, |
| "learning_rate": 6.466165413533834e-07, |
| "logits/chosen": 0.234375, |
| "logits/rejected": 0.169921875, |
| "logps/chosen": -25.375, |
| "logps/rejected": -26.0, |
| "loss": 0.4324, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.216796875, |
| "rewards/margins": 0.69921875, |
| "rewards/rejected": -0.482421875, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.6813289658399626, |
| "grad_norm": 43.72792434692383, |
| "learning_rate": 6.31578947368421e-07, |
| "logits/chosen": 0.32421875, |
| "logits/rejected": 0.27734375, |
| "logps/chosen": -26.0, |
| "logps/rejected": -120.5, |
| "loss": 0.4289, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.306640625, |
| "rewards/margins": 1.109375, |
| "rewards/rejected": -0.80078125, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.6888160973327094, |
| "grad_norm": 48.30878448486328, |
| "learning_rate": 6.165413533834586e-07, |
| "logits/chosen": 0.3203125, |
| "logits/rejected": 0.287109375, |
| "logps/chosen": -22.5, |
| "logps/rejected": -25.375, |
| "loss": 0.4916, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.095703125, |
| "rewards/margins": 0.58984375, |
| "rewards/rejected": -0.494140625, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.6963032288254563, |
| "grad_norm": 46.46865463256836, |
| "learning_rate": 6.015037593984961e-07, |
| "logits/chosen": 0.30078125, |
| "logits/rejected": 0.1279296875, |
| "logps/chosen": -24.5, |
| "logps/rejected": -225.0, |
| "loss": 0.3761, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2001953125, |
| "rewards/margins": 1.796875, |
| "rewards/rejected": -1.6015625, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.7037903603182031, |
| "grad_norm": 36.16524887084961, |
| "learning_rate": 5.864661654135338e-07, |
| "logits/chosen": 0.36328125, |
| "logits/rejected": 0.248046875, |
| "logps/chosen": -22.75, |
| "logps/rejected": -436.0, |
| "loss": 0.3547, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2265625, |
| "rewards/margins": 2.625, |
| "rewards/rejected": -2.390625, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.7112774918109499, |
| "grad_norm": 82.83430480957031, |
| "learning_rate": 5.714285714285714e-07, |
| "logits/chosen": 0.1513671875, |
| "logits/rejected": 0.251953125, |
| "logps/chosen": -28.0, |
| "logps/rejected": -24.625, |
| "loss": 0.5123, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1484375, |
| "rewards/margins": 0.5390625, |
| "rewards/rejected": -0.390625, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.7187646233036967, |
| "grad_norm": 38.68367385864258, |
| "learning_rate": 5.56390977443609e-07, |
| "logits/chosen": 0.248046875, |
| "logits/rejected": 0.1533203125, |
| "logps/chosen": -25.625, |
| "logps/rejected": -140.0, |
| "loss": 0.4303, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1787109375, |
| "rewards/margins": 1.265625, |
| "rewards/rejected": -1.0859375, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.7262517547964437, |
| "grad_norm": 53.11600875854492, |
| "learning_rate": 5.413533834586465e-07, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.27734375, |
| "logps/chosen": -23.75, |
| "logps/rejected": -27.75, |
| "loss": 0.5037, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.13671875, |
| "rewards/margins": 0.59375, |
| "rewards/rejected": -0.458984375, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.7337388862891905, |
| "grad_norm": 49.679344177246094, |
| "learning_rate": 5.263157894736842e-07, |
| "logits/chosen": 0.30859375, |
| "logits/rejected": 0.2060546875, |
| "logps/chosen": -23.625, |
| "logps/rejected": -144.0, |
| "loss": 0.4902, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.07080078125, |
| "rewards/margins": 1.171875, |
| "rewards/rejected": -1.09375, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.7412260177819373, |
| "grad_norm": 72.93644714355469, |
| "learning_rate": 5.112781954887218e-07, |
| "logits/chosen": 0.2421875, |
| "logits/rejected": 0.2138671875, |
| "logps/chosen": -25.75, |
| "logps/rejected": -23.5, |
| "loss": 0.4311, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.337890625, |
| "rewards/margins": 0.703125, |
| "rewards/rejected": -0.36328125, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.7487131492746841, |
| "grad_norm": 40.050777435302734, |
| "learning_rate": 4.962406015037593e-07, |
| "logits/chosen": 0.2578125, |
| "logits/rejected": 0.1982421875, |
| "logps/chosen": -26.0, |
| "logps/rejected": -44.0, |
| "loss": 0.4212, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.345703125, |
| "rewards/margins": 0.98828125, |
| "rewards/rejected": -0.640625, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.756200280767431, |
| "grad_norm": 45.1883544921875, |
| "learning_rate": 4.812030075187969e-07, |
| "logits/chosen": 0.26171875, |
| "logits/rejected": 0.310546875, |
| "logps/chosen": -26.5, |
| "logps/rejected": -22.375, |
| "loss": 0.4283, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.3515625, |
| "rewards/margins": 0.71875, |
| "rewards/rejected": -0.3671875, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.7636874122601778, |
| "grad_norm": 40.10118103027344, |
| "learning_rate": 4.6616541353383456e-07, |
| "logits/chosen": 0.28125, |
| "logits/rejected": 0.208984375, |
| "logps/chosen": -26.375, |
| "logps/rejected": -219.0, |
| "loss": 0.3421, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.310546875, |
| "rewards/margins": 2.078125, |
| "rewards/rejected": -1.765625, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.7711745437529247, |
| "grad_norm": 107.4455795288086, |
| "learning_rate": 4.5112781954887213e-07, |
| "logits/chosen": 0.1884765625, |
| "logits/rejected": 0.15625, |
| "logps/chosen": -24.75, |
| "logps/rejected": -182.0, |
| "loss": 0.4027, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2236328125, |
| "rewards/margins": 1.53125, |
| "rewards/rejected": -1.3125, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.7786616752456715, |
| "grad_norm": 51.714324951171875, |
| "learning_rate": 4.3609022556390975e-07, |
| "logits/chosen": 0.251953125, |
| "logits/rejected": 0.1318359375, |
| "logps/chosen": -24.625, |
| "logps/rejected": -38.0, |
| "loss": 0.391, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.25390625, |
| "rewards/margins": 1.0390625, |
| "rewards/rejected": -0.78515625, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.7861488067384184, |
| "grad_norm": 54.54036331176758, |
| "learning_rate": 4.2105263157894733e-07, |
| "logits/chosen": 0.1640625, |
| "logits/rejected": 0.037109375, |
| "logps/chosen": -26.25, |
| "logps/rejected": -56.0, |
| "loss": 0.3871, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2158203125, |
| "rewards/margins": 1.34375, |
| "rewards/rejected": -1.1328125, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.7936359382311652, |
| "grad_norm": 47.15876007080078, |
| "learning_rate": 4.060150375939849e-07, |
| "logits/chosen": 0.2392578125, |
| "logits/rejected": 0.1875, |
| "logps/chosen": -26.625, |
| "logps/rejected": -32.25, |
| "loss": 0.3612, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.43359375, |
| "rewards/margins": 1.078125, |
| "rewards/rejected": -0.64453125, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.801123069723912, |
| "grad_norm": 54.331844329833984, |
| "learning_rate": 3.909774436090225e-07, |
| "logits/chosen": 0.3046875, |
| "logits/rejected": 0.158203125, |
| "logps/chosen": -24.875, |
| "logps/rejected": -37.5, |
| "loss": 0.3862, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.318359375, |
| "rewards/margins": 1.078125, |
| "rewards/rejected": -0.76171875, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.8086102012166588, |
| "grad_norm": 39.809444427490234, |
| "learning_rate": 3.759398496240601e-07, |
| "logits/chosen": 0.2197265625, |
| "logits/rejected": 0.16796875, |
| "logps/chosen": -27.125, |
| "logps/rejected": -103.5, |
| "loss": 0.4545, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1484375, |
| "rewards/margins": 1.2734375, |
| "rewards/rejected": -1.125, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.8160973327094057, |
| "grad_norm": 49.4395751953125, |
| "learning_rate": 3.609022556390977e-07, |
| "logits/chosen": 0.2060546875, |
| "logits/rejected": 0.162109375, |
| "logps/chosen": -26.125, |
| "logps/rejected": -39.0, |
| "loss": 0.3786, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.416015625, |
| "rewards/margins": 1.15625, |
| "rewards/rejected": -0.73828125, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.8235844642021526, |
| "grad_norm": 68.54365539550781, |
| "learning_rate": 3.458646616541353e-07, |
| "logits/chosen": 0.166015625, |
| "logits/rejected": 0.1005859375, |
| "logps/chosen": -26.125, |
| "logps/rejected": -30.0, |
| "loss": 0.3824, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.287109375, |
| "rewards/margins": 0.96484375, |
| "rewards/rejected": -0.6796875, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.8310715956948994, |
| "grad_norm": 57.101627349853516, |
| "learning_rate": 3.308270676691729e-07, |
| "logits/chosen": 0.28125, |
| "logits/rejected": 0.302734375, |
| "logps/chosen": -25.0, |
| "logps/rejected": -96.5, |
| "loss": 0.4199, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.341796875, |
| "rewards/margins": 1.0703125, |
| "rewards/rejected": -0.7265625, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.8385587271876462, |
| "grad_norm": 42.954647064208984, |
| "learning_rate": 3.157894736842105e-07, |
| "logits/chosen": 0.21875, |
| "logits/rejected": 0.2373046875, |
| "logps/chosen": -26.5, |
| "logps/rejected": -24.375, |
| "loss": 0.4437, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.267578125, |
| "rewards/margins": 0.71484375, |
| "rewards/rejected": -0.44921875, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.8460458586803931, |
| "grad_norm": 44.198177337646484, |
| "learning_rate": 3.0075187969924807e-07, |
| "logits/chosen": 0.287109375, |
| "logits/rejected": 0.1875, |
| "logps/chosen": -26.125, |
| "logps/rejected": -30.25, |
| "loss": 0.4037, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.25390625, |
| "rewards/margins": 0.90625, |
| "rewards/rejected": -0.65234375, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.8535329901731399, |
| "grad_norm": 45.43602752685547, |
| "learning_rate": 2.857142857142857e-07, |
| "logits/chosen": 0.224609375, |
| "logits/rejected": 0.1474609375, |
| "logps/chosen": -25.875, |
| "logps/rejected": -189.0, |
| "loss": 0.3862, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.27734375, |
| "rewards/margins": 1.8046875, |
| "rewards/rejected": -1.5234375, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.8610201216658867, |
| "grad_norm": 44.5158805847168, |
| "learning_rate": 2.7067669172932327e-07, |
| "logits/chosen": 0.208984375, |
| "logits/rejected": 0.09716796875, |
| "logps/chosen": -25.5, |
| "logps/rejected": -33.25, |
| "loss": 0.436, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.25, |
| "rewards/margins": 0.87890625, |
| "rewards/rejected": -0.62890625, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.8685072531586336, |
| "grad_norm": 60.82008743286133, |
| "learning_rate": 2.556390977443609e-07, |
| "logits/chosen": 0.283203125, |
| "logits/rejected": 0.2431640625, |
| "logps/chosen": -23.5, |
| "logps/rejected": -50.0, |
| "loss": 0.4736, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2158203125, |
| "rewards/margins": 0.76953125, |
| "rewards/rejected": -0.5546875, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.8759943846513805, |
| "grad_norm": 65.28529357910156, |
| "learning_rate": 2.4060150375939847e-07, |
| "logits/chosen": 0.134765625, |
| "logits/rejected": 0.0751953125, |
| "logps/chosen": -26.25, |
| "logps/rejected": -35.0, |
| "loss": 0.3799, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.306640625, |
| "rewards/margins": 1.1015625, |
| "rewards/rejected": -0.79296875, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.8834815161441273, |
| "grad_norm": 54.721885681152344, |
| "learning_rate": 2.2556390977443606e-07, |
| "logits/chosen": 0.236328125, |
| "logits/rejected": 0.12158203125, |
| "logps/chosen": -24.125, |
| "logps/rejected": -27.625, |
| "loss": 0.4465, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.2333984375, |
| "rewards/margins": 0.7421875, |
| "rewards/rejected": -0.5078125, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.8909686476368741, |
| "grad_norm": 47.69539260864258, |
| "learning_rate": 2.1052631578947366e-07, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.1953125, |
| "logps/chosen": -24.25, |
| "logps/rejected": -30.25, |
| "loss": 0.4815, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1015625, |
| "rewards/margins": 0.625, |
| "rewards/rejected": -0.5234375, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.8984557791296209, |
| "grad_norm": 37.75003433227539, |
| "learning_rate": 1.9548872180451126e-07, |
| "logits/chosen": 0.29296875, |
| "logits/rejected": 0.146484375, |
| "logps/chosen": -24.875, |
| "logps/rejected": -112.5, |
| "loss": 0.4232, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1650390625, |
| "rewards/margins": 1.1875, |
| "rewards/rejected": -1.0234375, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.9059429106223678, |
| "grad_norm": 65.35391235351562, |
| "learning_rate": 1.8045112781954886e-07, |
| "logits/chosen": 0.330078125, |
| "logits/rejected": 0.205078125, |
| "logps/chosen": -24.375, |
| "logps/rejected": -36.25, |
| "loss": 0.3961, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.201171875, |
| "rewards/margins": 0.9921875, |
| "rewards/rejected": -0.79296875, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.9134300421151147, |
| "grad_norm": 41.64967727661133, |
| "learning_rate": 1.6541353383458646e-07, |
| "logits/chosen": 0.2353515625, |
| "logits/rejected": 0.2216796875, |
| "logps/chosen": -24.25, |
| "logps/rejected": -78.5, |
| "loss": 0.3987, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2236328125, |
| "rewards/margins": 1.3125, |
| "rewards/rejected": -1.09375, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.9209171736078615, |
| "grad_norm": 45.184326171875, |
| "learning_rate": 1.5037593984962403e-07, |
| "logits/chosen": 0.267578125, |
| "logits/rejected": 0.1630859375, |
| "logps/chosen": -25.25, |
| "logps/rejected": -99.0, |
| "loss": 0.4022, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.212890625, |
| "rewards/margins": 1.3359375, |
| "rewards/rejected": -1.1171875, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.9284043051006083, |
| "grad_norm": 37.19181823730469, |
| "learning_rate": 1.3533834586466163e-07, |
| "logits/chosen": 0.2314453125, |
| "logits/rejected": 0.14453125, |
| "logps/chosen": -26.375, |
| "logps/rejected": -190.0, |
| "loss": 0.3883, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.302734375, |
| "rewards/margins": 1.90625, |
| "rewards/rejected": -1.609375, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.9358914365933552, |
| "grad_norm": 48.21459197998047, |
| "learning_rate": 1.2030075187969923e-07, |
| "logits/chosen": 0.26953125, |
| "logits/rejected": 0.2216796875, |
| "logps/chosen": -24.625, |
| "logps/rejected": -23.125, |
| "loss": 0.4731, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.134765625, |
| "rewards/margins": 0.5625, |
| "rewards/rejected": -0.42578125, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.943378568086102, |
| "grad_norm": 46.0123176574707, |
| "learning_rate": 1.0526315789473683e-07, |
| "logits/chosen": 0.2177734375, |
| "logits/rejected": 0.1728515625, |
| "logps/chosen": -25.0, |
| "logps/rejected": -28.25, |
| "loss": 0.3912, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.1787109375, |
| "rewards/margins": 0.83984375, |
| "rewards/rejected": -0.66015625, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.9508656995788488, |
| "grad_norm": 52.23374938964844, |
| "learning_rate": 9.022556390977443e-08, |
| "logits/chosen": 0.2314453125, |
| "logits/rejected": 0.1796875, |
| "logps/chosen": -24.25, |
| "logps/rejected": -86.0, |
| "loss": 0.4142, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.28515625, |
| "rewards/margins": 1.3125, |
| "rewards/rejected": -1.03125, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.9583528310715957, |
| "grad_norm": 39.65555953979492, |
| "learning_rate": 7.518796992481202e-08, |
| "logits/chosen": 0.1796875, |
| "logits/rejected": 0.1103515625, |
| "logps/chosen": -28.0, |
| "logps/rejected": -130.0, |
| "loss": 0.3215, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.40234375, |
| "rewards/margins": 2.046875, |
| "rewards/rejected": -1.6328125, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.9658399625643426, |
| "grad_norm": 41.55595016479492, |
| "learning_rate": 6.015037593984962e-08, |
| "logits/chosen": 0.267578125, |
| "logits/rejected": 0.251953125, |
| "logps/chosen": -24.25, |
| "logps/rejected": -27.25, |
| "loss": 0.4264, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.220703125, |
| "rewards/margins": 0.84765625, |
| "rewards/rejected": -0.625, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.9733270940570894, |
| "grad_norm": 44.8392333984375, |
| "learning_rate": 4.5112781954887216e-08, |
| "logits/chosen": 0.353515625, |
| "logits/rejected": 0.27734375, |
| "logps/chosen": -22.375, |
| "logps/rejected": -177.0, |
| "loss": 0.4437, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.146484375, |
| "rewards/margins": 1.4765625, |
| "rewards/rejected": -1.328125, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.9808142255498362, |
| "grad_norm": 46.03443145751953, |
| "learning_rate": 3.007518796992481e-08, |
| "logits/chosen": 0.296875, |
| "logits/rejected": 0.216796875, |
| "logps/chosen": -25.0, |
| "logps/rejected": -26.375, |
| "loss": 0.4178, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2890625, |
| "rewards/margins": 0.83203125, |
| "rewards/rejected": -0.54296875, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.988301357042583, |
| "grad_norm": 39.30998611450195, |
| "learning_rate": 1.5037593984962404e-08, |
| "logits/chosen": 0.326171875, |
| "logits/rejected": 0.2451171875, |
| "logps/chosen": -23.25, |
| "logps/rejected": -100.5, |
| "loss": 0.4292, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.181640625, |
| "rewards/margins": 1.1484375, |
| "rewards/rejected": -0.96484375, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.9957884885353299, |
| "grad_norm": 35.90797424316406, |
| "learning_rate": 0.0, |
| "logits/chosen": 0.1376953125, |
| "logits/rejected": 0.1220703125, |
| "logps/chosen": -29.0, |
| "logps/rejected": -190.0, |
| "loss": 0.3593, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.34765625, |
| "rewards/margins": 2.046875, |
| "rewards/rejected": -1.6953125, |
| "step": 133 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 133, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "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": 0.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
