CriteriaPO/qwen2.5-3b-dpo-coarse
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
df543bcd09939e90d4ffadc25d29f7b182fc9ad9
qwen2.5-3b-dpo-coarse/checkpoint-10000/trainer_state.json
ModelHub XC df543bcd09 初始化项目，由ModelHub XC社区提供模型 
Model: CriteriaPO/qwen2.5-3b-dpo-coarse
Source: Original Platform
2026-09-06 05:38:18 +08:00

15081 lines
445 KiB
JSON
Raw Blame History

{
"best_metric": 0.23313412070274353,
"best_model_checkpoint": "models/qwen2.5-3b-dpo-coarse/checkpoint-10000",
"epoch": 0.7523888345496953,
"eval_steps": 5000,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 7.523888345496953e-05,
"grad_norm": 2.0655642984991194,
"learning_rate": 3.7593984962406016e-10,
"logits/chosen": -2.03125,
"logits/rejected": -1.53125,
"logps/chosen": -121.0,
"logps/rejected": -146.0,
"loss": 0.6914,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.0007523888345496953,
"grad_norm": 1.9538445459115994,
"learning_rate": 3.759398496240601e-09,
"logits/chosen": -2.390625,
"logits/rejected": -2.09375,
"logps/chosen": -149.0,
"logps/rejected": -147.0,
"loss": 0.692,
"rewards/accuracies": 0.2986111044883728,
"rewards/chosen": 0.0013580322265625,
"rewards/margins": 0.00238037109375,
"rewards/rejected": -0.0010223388671875,
"step": 10
},
{
"epoch": 0.0015047776690993906,
"grad_norm": 1.7482239733475604,
"learning_rate": 7.518796992481202e-09,
"logits/chosen": -2.390625,
"logits/rejected": -2.0,
"logps/chosen": -178.0,
"logps/rejected": -159.0,
"loss": 0.6927,
"rewards/accuracies": 0.375,
"rewards/chosen": 0.00087738037109375,
"rewards/margins": 0.0001277923583984375,
"rewards/rejected": 0.000751495361328125,
"step": 20
},
{
"epoch": 0.0022571665036490857,
"grad_norm": 1.6937945346625471,
"learning_rate": 1.1278195488721804e-08,
"logits/chosen": -2.40625,
"logits/rejected": -2.0,
"logps/chosen": -173.0,
"logps/rejected": -171.0,
"loss": 0.6927,
"rewards/accuracies": 0.40625,
"rewards/chosen": 0.00151824951171875,
"rewards/margins": 0.0020599365234375,
"rewards/rejected": -0.000545501708984375,
"step": 30
},
{
"epoch": 0.003009555338198781,
"grad_norm": 1.8569983758522628,
"learning_rate": 1.5037593984962404e-08,
"logits/chosen": -2.34375,
"logits/rejected": -2.078125,
"logps/chosen": -161.0,
"logps/rejected": -161.0,
"loss": 0.6927,
"rewards/accuracies": 0.3499999940395355,
"rewards/chosen": 0.000705718994140625,
"rewards/margins": -0.000499725341796875,
"rewards/rejected": 0.0012054443359375,
"step": 40
},
{
"epoch": 0.0037619441727484763,
"grad_norm": 1.955956889927452,
"learning_rate": 1.8796992481203004e-08,
"logits/chosen": -2.375,
"logits/rejected": -2.03125,
"logps/chosen": -154.0,
"logps/rejected": -155.0,
"loss": 0.6927,
"rewards/accuracies": 0.3687500059604645,
"rewards/chosen": -0.00151824951171875,
"rewards/margins": 0.00048828125,
"rewards/rejected": -0.0019989013671875,
"step": 50
},
{
"epoch": 0.004514333007298171,
"grad_norm": 1.8154990127537094,
"learning_rate": 2.2556390977443608e-08,
"logits/chosen": -2.40625,
"logits/rejected": -2.015625,
"logps/chosen": -161.0,
"logps/rejected": -147.0,
"loss": 0.6924,
"rewards/accuracies": 0.32499998807907104,
"rewards/chosen": -0.000499725341796875,
"rewards/margins": -0.000720977783203125,
"rewards/rejected": 0.00021839141845703125,
"step": 60
},
{
"epoch": 0.005266721841847867,
"grad_norm": 1.987488235998793,
"learning_rate": 2.6315789473684208e-08,
"logits/chosen": -2.40625,
"logits/rejected": -1.890625,
"logps/chosen": -169.0,
"logps/rejected": -171.0,
"loss": 0.6926,
"rewards/accuracies": 0.375,
"rewards/chosen": 0.00022125244140625,
"rewards/margins": 0.00021839141845703125,
"rewards/rejected": -1.5273690223693848e-06,
"step": 70
},
{
"epoch": 0.006019110676397562,
"grad_norm": 1.7715247772447917,
"learning_rate": 3.007518796992481e-08,
"logits/chosen": -2.390625,
"logits/rejected": -2.03125,
"logps/chosen": -157.0,
"logps/rejected": -148.0,
"loss": 0.6928,
"rewards/accuracies": 0.36250001192092896,
"rewards/chosen": 0.00032806396484375,
"rewards/margins": 0.0004253387451171875,
"rewards/rejected": -9.059906005859375e-05,
"step": 80
},
{
"epoch": 0.006771499510947258,
"grad_norm": 1.5425426980374812,
"learning_rate": 3.383458646616541e-08,
"logits/chosen": -2.453125,
"logits/rejected": -2.15625,
"logps/chosen": -171.0,
"logps/rejected": -151.0,
"loss": 0.6926,
"rewards/accuracies": 0.375,
"rewards/chosen": -0.000644683837890625,
"rewards/margins": 7.82012939453125e-05,
"rewards/rejected": -0.000720977783203125,
"step": 90
},
{
"epoch": 0.0075238883454969525,
"grad_norm": 1.8781269521393504,
"learning_rate": 3.759398496240601e-08,
"logits/chosen": -2.375,
"logits/rejected": -2.015625,
"logps/chosen": -163.0,
"logps/rejected": -147.0,
"loss": 0.6922,
"rewards/accuracies": 0.29374998807907104,
"rewards/chosen": -0.00089263916015625,
"rewards/margins": -0.00010251998901367188,
"rewards/rejected": -0.000789642333984375,
"step": 100
},
{
"epoch": 0.008276277180046648,
"grad_norm": 1.8425671870697207,
"learning_rate": 4.1353383458646615e-08,
"logits/chosen": -2.21875,
"logits/rejected": -1.953125,
"logps/chosen": -159.0,
"logps/rejected": -154.0,
"loss": 0.6927,
"rewards/accuracies": 0.36250001192092896,
"rewards/chosen": 0.00018787384033203125,
"rewards/margins": -0.000438690185546875,
"rewards/rejected": 0.0006256103515625,
"step": 110
},
{
"epoch": 0.009028666014596343,
"grad_norm": 1.8232396339646215,
"learning_rate": 4.5112781954887216e-08,
"logits/chosen": -2.359375,
"logits/rejected": -1.953125,
"logps/chosen": -158.0,
"logps/rejected": -139.0,
"loss": 0.6923,
"rewards/accuracies": 0.3125,
"rewards/chosen": 0.00189208984375,
"rewards/margins": -0.0004062652587890625,
"rewards/rejected": 0.0023040771484375,
"step": 120
},
{
"epoch": 0.009781054849146039,
"grad_norm": 2.0676564160851814,
"learning_rate": 4.8872180451127816e-08,
"logits/chosen": -2.40625,
"logits/rejected": -2.015625,
"logps/chosen": -137.0,
"logps/rejected": -147.0,
"loss": 0.692,
"rewards/accuracies": 0.36250001192092896,
"rewards/chosen": 0.000797271728515625,
"rewards/margins": -3.0994415283203125e-05,
"rewards/rejected": 0.000827789306640625,
"step": 130
},
{
"epoch": 0.010533443683695734,
"grad_norm": 1.9395451342559986,
"learning_rate": 5.2631578947368416e-08,
"logits/chosen": -2.3125,
"logits/rejected": -2.046875,
"logps/chosen": -159.0,
"logps/rejected": -147.0,
"loss": 0.692,
"rewards/accuracies": 0.375,
"rewards/chosen": 0.001190185546875,
"rewards/margins": -1.5974044799804688e-05,
"rewards/rejected": 0.0012054443359375,
"step": 140
},
{
"epoch": 0.01128583251824543,
"grad_norm": 1.9021298256381785,
"learning_rate": 5.6390977443609016e-08,
"logits/chosen": -2.375,
"logits/rejected": -1.9140625,
"logps/chosen": -175.0,
"logps/rejected": -148.0,
"loss": 0.6921,
"rewards/accuracies": 0.42500001192092896,
"rewards/chosen": 0.00177001953125,
"rewards/margins": 0.00213623046875,
"rewards/rejected": -0.000370025634765625,
"step": 150
},
{
"epoch": 0.012038221352795125,
"grad_norm": 1.9054050326686525,
"learning_rate": 6.015037593984962e-08,
"logits/chosen": -2.40625,
"logits/rejected": -2.0625,
"logps/chosen": -160.0,
"logps/rejected": -145.0,
"loss": 0.6917,
"rewards/accuracies": 0.4375,
"rewards/chosen": 0.0025482177734375,
"rewards/margins": 0.0022430419921875,
"rewards/rejected": 0.0003108978271484375,
"step": 160
},
{
"epoch": 0.01279061018734482,
"grad_norm": 1.926853611550257,
"learning_rate": 6.390977443609022e-08,
"logits/chosen": -2.296875,
"logits/rejected": -1.859375,
"logps/chosen": -165.0,
"logps/rejected": -151.0,
"loss": 0.6912,
"rewards/accuracies": 0.5249999761581421,
"rewards/chosen": 0.0035552978515625,
"rewards/margins": 0.003875732421875,
"rewards/rejected": -0.0003299713134765625,
"step": 170
},
{
"epoch": 0.013542999021894516,
"grad_norm": 1.9234881642608486,
"learning_rate": 6.766917293233082e-08,
"logits/chosen": -2.359375,
"logits/rejected": -2.03125,
"logps/chosen": -145.0,
"logps/rejected": -154.0,
"loss": 0.6915,
"rewards/accuracies": 0.4937500059604645,
"rewards/chosen": 0.00628662109375,
"rewards/margins": 0.00213623046875,
"rewards/rejected": 0.004150390625,
"step": 180
},
{
"epoch": 0.01429538785644421,
"grad_norm": 1.9784096184868571,
"learning_rate": 7.142857142857142e-08,
"logits/chosen": -2.5,
"logits/rejected": -2.109375,
"logps/chosen": -163.0,
"logps/rejected": -142.0,
"loss": 0.6911,
"rewards/accuracies": 0.5,
"rewards/chosen": 0.006256103515625,
"rewards/margins": 0.0028839111328125,
"rewards/rejected": 0.0033721923828125,
"step": 190
},
{
"epoch": 0.015047776690993905,
"grad_norm": 1.631724548633527,
"learning_rate": 7.518796992481202e-08,
"logits/chosen": -2.46875,
"logits/rejected": -2.09375,
"logps/chosen": -162.0,
"logps/rejected": -157.0,
"loss": 0.691,
"rewards/accuracies": 0.5249999761581421,
"rewards/chosen": 0.00616455078125,
"rewards/margins": 0.0031280517578125,
"rewards/rejected": 0.003021240234375,
"step": 200
},
{
"epoch": 0.0158001655255436,
"grad_norm": 1.7460592334442049,
"learning_rate": 7.894736842105262e-08,
"logits/chosen": -2.421875,
"logits/rejected": -2.09375,
"logps/chosen": -158.0,
"logps/rejected": -153.0,
"loss": 0.6906,
"rewards/accuracies": 0.4749999940395355,
"rewards/chosen": 0.00665283203125,
"rewards/margins": 0.002532958984375,
"rewards/rejected": 0.004119873046875,
"step": 210
},
{
"epoch": 0.016552554360093296,
"grad_norm": 1.864652158743688,
"learning_rate": 8.270676691729323e-08,
"logits/chosen": -2.46875,
"logits/rejected": -2.140625,
"logps/chosen": -169.0,
"logps/rejected": -157.0,
"loss": 0.6898,
"rewards/accuracies": 0.5562499761581421,
"rewards/chosen": 0.010498046875,
"rewards/margins": 0.00445556640625,
"rewards/rejected": 0.00604248046875,
"step": 220
},
{
"epoch": 0.017304943194642992,
"grad_norm": 1.8807523922343017,
"learning_rate": 8.646616541353382e-08,
"logits/chosen": -2.40625,
"logits/rejected": -2.03125,
"logps/chosen": -150.0,
"logps/rejected": -166.0,
"loss": 0.6892,
"rewards/accuracies": 0.550000011920929,
"rewards/chosen": 0.01275634765625,
"rewards/margins": 0.007049560546875,
"rewards/rejected": 0.005706787109375,
"step": 230
},
{
"epoch": 0.018057332029192685,
"grad_norm": 1.8543547888579985,
"learning_rate": 9.022556390977443e-08,
"logits/chosen": -2.453125,
"logits/rejected": -2.125,
"logps/chosen": -160.0,
"logps/rejected": -153.0,
"loss": 0.6882,
"rewards/accuracies": 0.637499988079071,
"rewards/chosen": 0.016357421875,
"rewards/margins": 0.0087890625,
"rewards/rejected": 0.007537841796875,
"step": 240
},
{
"epoch": 0.01880972086374238,
"grad_norm": 2.002695069307674,
"learning_rate": 9.398496240601502e-08,
"logits/chosen": -2.359375,
"logits/rejected": -1.953125,
"logps/chosen": -138.0,
"logps/rejected": -145.0,
"loss": 0.6882,
"rewards/accuracies": 0.606249988079071,
"rewards/chosen": 0.017822265625,
"rewards/margins": 0.01043701171875,
"rewards/rejected": 0.00738525390625,
"step": 250
},
{
"epoch": 0.019562109698292078,
"grad_norm": 1.694938582256644,
"learning_rate": 9.774436090225563e-08,
"logits/chosen": -2.390625,
"logits/rejected": -2.0625,
"logps/chosen": -143.0,
"logps/rejected": -154.0,
"loss": 0.6874,
"rewards/accuracies": 0.581250011920929,
"rewards/chosen": 0.0152587890625,
"rewards/margins": 0.00787353515625,
"rewards/rejected": 0.007354736328125,
"step": 260
},
{
"epoch": 0.02031449853284177,
"grad_norm": 1.6267392659283901,
"learning_rate": 1.0150375939849622e-07,
"logits/chosen": -2.484375,
"logits/rejected": -2.140625,
"logps/chosen": -173.0,
"logps/rejected": -155.0,
"loss": 0.6872,
"rewards/accuracies": 0.637499988079071,
"rewards/chosen": 0.0145263671875,
"rewards/margins": 0.0120849609375,
"rewards/rejected": 0.0024261474609375,
"step": 270
},
{
"epoch": 0.021066887367391467,
"grad_norm": 1.8009833530928583,
"learning_rate": 1.0526315789473683e-07,
"logits/chosen": -2.359375,
"logits/rejected": -2.25,
"logps/chosen": -169.0,
"logps/rejected": -151.0,
"loss": 0.6859,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": 0.01953125,
"rewards/margins": 0.0123291015625,
"rewards/rejected": 0.00714111328125,
"step": 280
},
{
"epoch": 0.021819276201941164,
"grad_norm": 1.751273107274032,
"learning_rate": 1.0902255639097744e-07,
"logits/chosen": -2.40625,
"logits/rejected": -2.0625,
"logps/chosen": -151.0,
"logps/rejected": -156.0,
"loss": 0.686,
"rewards/accuracies": 0.6312500238418579,
"rewards/chosen": 0.01953125,
"rewards/margins": 0.01397705078125,
"rewards/rejected": 0.005462646484375,
"step": 290
},
{
"epoch": 0.02257166503649086,
"grad_norm": 1.8215393575640535,
"learning_rate": 1.1278195488721803e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.0625,
"logps/chosen": -148.0,
"logps/rejected": -142.0,
"loss": 0.6845,
"rewards/accuracies": 0.668749988079071,
"rewards/chosen": 0.02099609375,
"rewards/margins": 0.0157470703125,
"rewards/rejected": 0.005279541015625,
"step": 300
},
{
"epoch": 0.023324053871040553,
"grad_norm": 1.7387394745988194,
"learning_rate": 1.1654135338345864e-07,
"logits/chosen": -2.328125,
"logits/rejected": -2.09375,
"logps/chosen": -152.0,
"logps/rejected": -127.0,
"loss": 0.6844,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": 0.017333984375,
"rewards/margins": 0.0174560546875,
"rewards/rejected": -0.00016117095947265625,
"step": 310
},
{
"epoch": 0.02407644270559025,
"grad_norm": 1.7127796622071032,
"learning_rate": 1.2030075187969923e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.09375,
"logps/chosen": -158.0,
"logps/rejected": -154.0,
"loss": 0.6838,
"rewards/accuracies": 0.65625,
"rewards/chosen": 0.0218505859375,
"rewards/margins": 0.016357421875,
"rewards/rejected": 0.005523681640625,
"step": 320
},
{
"epoch": 0.024828831540139946,
"grad_norm": 1.8326169180769538,
"learning_rate": 1.2406015037593983e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.015625,
"logps/chosen": -159.0,
"logps/rejected": -174.0,
"loss": 0.6806,
"rewards/accuracies": 0.706250011920929,
"rewards/chosen": 0.0296630859375,
"rewards/margins": 0.030029296875,
"rewards/rejected": -0.0003986358642578125,
"step": 330
},
{
"epoch": 0.02558122037468964,
"grad_norm": 1.7066967555396413,
"learning_rate": 1.2781954887218045e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.125,
"logps/chosen": -165.0,
"logps/rejected": -162.0,
"loss": 0.6797,
"rewards/accuracies": 0.706250011920929,
"rewards/chosen": 0.0230712890625,
"rewards/margins": 0.0279541015625,
"rewards/rejected": -0.0048828125,
"step": 340
},
{
"epoch": 0.026333609209239335,
"grad_norm": 1.735716628750598,
"learning_rate": 1.3157894736842104e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.03125,
"logps/chosen": -170.0,
"logps/rejected": -168.0,
"loss": 0.6781,
"rewards/accuracies": 0.71875,
"rewards/chosen": 0.029541015625,
"rewards/margins": 0.03857421875,
"rewards/rejected": -0.00909423828125,
"step": 350
},
{
"epoch": 0.02708599804378903,
"grad_norm": 2.113635572426321,
"learning_rate": 1.3533834586466163e-07,
"logits/chosen": -2.40625,
"logits/rejected": -1.984375,
"logps/chosen": -165.0,
"logps/rejected": -153.0,
"loss": 0.6773,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": 0.0218505859375,
"rewards/margins": 0.02783203125,
"rewards/rejected": -0.005889892578125,
"step": 360
},
{
"epoch": 0.027838386878338724,
"grad_norm": 1.7305039899173027,
"learning_rate": 1.3909774436090225e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.078125,
"logps/chosen": -147.0,
"logps/rejected": -144.0,
"loss": 0.6743,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": 0.0341796875,
"rewards/margins": 0.039794921875,
"rewards/rejected": -0.005645751953125,
"step": 370
},
{
"epoch": 0.02859077571288842,
"grad_norm": 1.7424950093564118,
"learning_rate": 1.4285714285714285e-07,
"logits/chosen": -2.40625,
"logits/rejected": -2.109375,
"logps/chosen": -151.0,
"logps/rejected": -144.0,
"loss": 0.6739,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": 0.0274658203125,
"rewards/margins": 0.050048828125,
"rewards/rejected": -0.0225830078125,
"step": 380
},
{
"epoch": 0.029343164547438117,
"grad_norm": 1.8220095077196594,
"learning_rate": 1.4661654135338344e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.265625,
"logps/chosen": -177.0,
"logps/rejected": -157.0,
"loss": 0.6736,
"rewards/accuracies": 0.7437499761581421,
"rewards/chosen": 0.016357421875,
"rewards/margins": 0.04541015625,
"rewards/rejected": -0.0291748046875,
"step": 390
},
{
"epoch": 0.03009555338198781,
"grad_norm": 1.9063850709586252,
"learning_rate": 1.5037593984962403e-07,
"logits/chosen": -2.40625,
"logits/rejected": -2.109375,
"logps/chosen": -157.0,
"logps/rejected": -151.0,
"loss": 0.6705,
"rewards/accuracies": 0.668749988079071,
"rewards/chosen": 0.01025390625,
"rewards/margins": 0.0380859375,
"rewards/rejected": -0.02783203125,
"step": 400
},
{
"epoch": 0.030847942216537506,
"grad_norm": 1.6795215046369691,
"learning_rate": 1.5413533834586465e-07,
"logits/chosen": -2.484375,
"logits/rejected": -2.125,
"logps/chosen": -162.0,
"logps/rejected": -154.0,
"loss": 0.6681,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": 0.01104736328125,
"rewards/margins": 0.056396484375,
"rewards/rejected": -0.045166015625,
"step": 410
},
{
"epoch": 0.0316003310510872,
"grad_norm": 2.091341380928474,
"learning_rate": 1.5789473684210525e-07,
"logits/chosen": -2.4375,
"logits/rejected": -2.09375,
"logps/chosen": -153.0,
"logps/rejected": -152.0,
"loss": 0.6638,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": -0.00799560546875,
"rewards/margins": 0.056396484375,
"rewards/rejected": -0.064453125,
"step": 420
},
{
"epoch": 0.0323527198856369,
"grad_norm": 2.083815342302955,
"learning_rate": 1.6165413533834584e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.15625,
"logps/chosen": -171.0,
"logps/rejected": -159.0,
"loss": 0.6609,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -0.004791259765625,
"rewards/margins": 0.07470703125,
"rewards/rejected": -0.0791015625,
"step": 430
},
{
"epoch": 0.03310510872018659,
"grad_norm": 2.1697428767427125,
"learning_rate": 1.6541353383458646e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.28125,
"logps/chosen": -197.0,
"logps/rejected": -177.0,
"loss": 0.6559,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": -0.0189208984375,
"rewards/margins": 0.0888671875,
"rewards/rejected": -0.10791015625,
"step": 440
},
{
"epoch": 0.033857497554736285,
"grad_norm": 2.1298018090255297,
"learning_rate": 1.6917293233082705e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.3125,
"logps/chosen": -179.0,
"logps/rejected": -157.0,
"loss": 0.6531,
"rewards/accuracies": 0.6937500238418579,
"rewards/chosen": -0.0400390625,
"rewards/margins": 0.07080078125,
"rewards/rejected": -0.11083984375,
"step": 450
},
{
"epoch": 0.034609886389285985,
"grad_norm": 2.1836244309606507,
"learning_rate": 1.7293233082706765e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.28125,
"logps/chosen": -167.0,
"logps/rejected": -182.0,
"loss": 0.6425,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -0.0615234375,
"rewards/margins": 0.11376953125,
"rewards/rejected": -0.17578125,
"step": 460
},
{
"epoch": 0.03536227522383568,
"grad_norm": 2.4213655673685968,
"learning_rate": 1.7669172932330824e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.21875,
"logps/chosen": -172.0,
"logps/rejected": -192.0,
"loss": 0.6384,
"rewards/accuracies": 0.7875000238418579,
"rewards/chosen": -0.0830078125,
"rewards/margins": 0.130859375,
"rewards/rejected": -0.2138671875,
"step": 470
},
{
"epoch": 0.03611466405838537,
"grad_norm": 1.944004930995184,
"learning_rate": 1.8045112781954886e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -176.0,
"logps/rejected": -174.0,
"loss": 0.6312,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -0.12451171875,
"rewards/margins": 0.1494140625,
"rewards/rejected": -0.2734375,
"step": 480
},
{
"epoch": 0.03686705289293507,
"grad_norm": 2.191837059576488,
"learning_rate": 1.8421052631578946e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.3125,
"logps/chosen": -188.0,
"logps/rejected": -201.0,
"loss": 0.63,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": -0.1689453125,
"rewards/margins": 0.1669921875,
"rewards/rejected": -0.3359375,
"step": 490
},
{
"epoch": 0.03761944172748476,
"grad_norm": 2.2196701470400395,
"learning_rate": 1.8796992481203005e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.265625,
"logps/chosen": -187.0,
"logps/rejected": -185.0,
"loss": 0.6251,
"rewards/accuracies": 0.6937500238418579,
"rewards/chosen": -0.2314453125,
"rewards/margins": 0.138671875,
"rewards/rejected": -0.369140625,
"step": 500
},
{
"epoch": 0.038371830562034456,
"grad_norm": 2.3001313944863644,
"learning_rate": 1.9172932330827067e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.34375,
"logps/chosen": -174.0,
"logps/rejected": -182.0,
"loss": 0.6189,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.255859375,
"rewards/margins": 0.2001953125,
"rewards/rejected": -0.45703125,
"step": 510
},
{
"epoch": 0.039124219396584156,
"grad_norm": 2.1503105403846376,
"learning_rate": 1.9548872180451126e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.21875,
"logps/chosen": -185.0,
"logps/rejected": -190.0,
"loss": 0.6124,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": -0.30859375,
"rewards/margins": 0.1826171875,
"rewards/rejected": -0.4921875,
"step": 520
},
{
"epoch": 0.03987660823113385,
"grad_norm": 2.330665502460398,
"learning_rate": 1.9924812030075186e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.296875,
"logps/chosen": -185.0,
"logps/rejected": -208.0,
"loss": 0.6012,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.333984375,
"rewards/margins": 0.25390625,
"rewards/rejected": -0.58984375,
"step": 530
},
{
"epoch": 0.04062899706568354,
"grad_norm": 3.5479816270273523,
"learning_rate": 2.0300751879699245e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.234375,
"logps/chosen": -197.0,
"logps/rejected": -206.0,
"loss": 0.5986,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": -0.416015625,
"rewards/margins": 0.2353515625,
"rewards/rejected": -0.65234375,
"step": 540
},
{
"epoch": 0.04138138590023324,
"grad_norm": 2.227973578985842,
"learning_rate": 2.0676691729323307e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.40625,
"logps/chosen": -194.0,
"logps/rejected": -204.0,
"loss": 0.6006,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": -0.439453125,
"rewards/margins": 0.2275390625,
"rewards/rejected": -0.66796875,
"step": 550
},
{
"epoch": 0.042133774734782935,
"grad_norm": 2.3382178358022494,
"learning_rate": 2.1052631578947366e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.546875,
"logps/chosen": -203.0,
"logps/rejected": -231.0,
"loss": 0.5962,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.451171875,
"rewards/margins": 0.25390625,
"rewards/rejected": -0.703125,
"step": 560
},
{
"epoch": 0.042886163569332635,
"grad_norm": 2.3431494680035843,
"learning_rate": 2.1428571428571426e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.453125,
"logps/chosen": -216.0,
"logps/rejected": -224.0,
"loss": 0.5836,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": -0.51171875,
"rewards/margins": 0.27734375,
"rewards/rejected": -0.7890625,
"step": 570
},
{
"epoch": 0.04363855240388233,
"grad_norm": 2.5522490577618884,
"learning_rate": 2.1804511278195488e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.546875,
"logps/chosen": -223.0,
"logps/rejected": -230.0,
"loss": 0.5791,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": -0.546875,
"rewards/margins": 0.306640625,
"rewards/rejected": -0.85546875,
"step": 580
},
{
"epoch": 0.04439094123843202,
"grad_norm": 2.8512898992256646,
"learning_rate": 2.2180451127819547e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5,
"logps/chosen": -217.0,
"logps/rejected": -237.0,
"loss": 0.5741,
"rewards/accuracies": 0.7875000238418579,
"rewards/chosen": -0.62890625,
"rewards/margins": 0.36328125,
"rewards/rejected": -0.9921875,
"step": 590
},
{
"epoch": 0.04514333007298172,
"grad_norm": 4.011523561431844,
"learning_rate": 2.2556390977443606e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.734375,
"logps/chosen": -245.0,
"logps/rejected": -266.0,
"loss": 0.5703,
"rewards/accuracies": 0.7437499761581421,
"rewards/chosen": -0.703125,
"rewards/margins": 0.369140625,
"rewards/rejected": -1.0703125,
"step": 600
},
{
"epoch": 0.04589571890753141,
"grad_norm": 3.41561545490801,
"learning_rate": 2.2932330827067666e-07,
"logits/chosen": -2.984375,
"logits/rejected": -2.671875,
"logps/chosen": -208.0,
"logps/rejected": -260.0,
"loss": 0.548,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": -0.6953125,
"rewards/margins": 0.484375,
"rewards/rejected": -1.1796875,
"step": 610
},
{
"epoch": 0.046648107742081106,
"grad_norm": 4.4201440287646205,
"learning_rate": 2.3308270676691728e-07,
"logits/chosen": -3.1875,
"logits/rejected": -2.765625,
"logps/chosen": -224.0,
"logps/rejected": -278.0,
"loss": 0.5474,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": -0.78515625,
"rewards/margins": 0.478515625,
"rewards/rejected": -1.265625,
"step": 620
},
{
"epoch": 0.047400496576630806,
"grad_norm": 4.695382679355146,
"learning_rate": 2.3684210526315787e-07,
"logits/chosen": -3.03125,
"logits/rejected": -2.703125,
"logps/chosen": -249.0,
"logps/rejected": -278.0,
"loss": 0.5397,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": -0.80078125,
"rewards/margins": 0.458984375,
"rewards/rejected": -1.2578125,
"step": 630
},
{
"epoch": 0.0481528854111805,
"grad_norm": 4.146268880636602,
"learning_rate": 2.4060150375939847e-07,
"logits/chosen": -3.03125,
"logits/rejected": -2.671875,
"logps/chosen": -234.0,
"logps/rejected": -290.0,
"loss": 0.5382,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -0.76171875,
"rewards/margins": 0.59765625,
"rewards/rejected": -1.359375,
"step": 640
},
{
"epoch": 0.04890527424573019,
"grad_norm": 5.032383387570057,
"learning_rate": 2.443609022556391e-07,
"logits/chosen": -3.140625,
"logits/rejected": -2.734375,
"logps/chosen": -237.0,
"logps/rejected": -300.0,
"loss": 0.548,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -0.85546875,
"rewards/margins": 0.55078125,
"rewards/rejected": -1.40625,
"step": 650
},
{
"epoch": 0.04965766308027989,
"grad_norm": 3.68940996433093,
"learning_rate": 2.4812030075187965e-07,
"logits/chosen": -3.140625,
"logits/rejected": -2.84375,
"logps/chosen": -228.0,
"logps/rejected": -280.0,
"loss": 0.5468,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -0.83984375,
"rewards/margins": 0.4921875,
"rewards/rejected": -1.328125,
"step": 660
},
{
"epoch": 0.050410051914829584,
"grad_norm": 3.876718495059869,
"learning_rate": 2.518796992481203e-07,
"logits/chosen": -3.046875,
"logits/rejected": -2.8125,
"logps/chosen": -258.0,
"logps/rejected": -294.0,
"loss": 0.5255,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": -0.8828125,
"rewards/margins": 0.5,
"rewards/rejected": -1.3828125,
"step": 670
},
{
"epoch": 0.05116244074937928,
"grad_norm": 4.75321383421562,
"learning_rate": 2.556390977443609e-07,
"logits/chosen": -3.21875,
"logits/rejected": -2.9375,
"logps/chosen": -260.0,
"logps/rejected": -320.0,
"loss": 0.5196,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -0.875,
"rewards/margins": 0.7421875,
"rewards/rejected": -1.6171875,
"step": 680
},
{
"epoch": 0.05191482958392898,
"grad_norm": 7.793668826360735,
"learning_rate": 2.593984962406015e-07,
"logits/chosen": -3.1875,
"logits/rejected": -3.03125,
"logps/chosen": -266.0,
"logps/rejected": -318.0,
"loss": 0.5229,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": -0.9609375,
"rewards/margins": 0.63671875,
"rewards/rejected": -1.6015625,
"step": 690
},
{
"epoch": 0.05266721841847867,
"grad_norm": 4.6452917151128785,
"learning_rate": 2.631578947368421e-07,
"logits/chosen": -3.109375,
"logits/rejected": -2.875,
"logps/chosen": -266.0,
"logps/rejected": -324.0,
"loss": 0.5161,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": -0.96875,
"rewards/margins": 0.69921875,
"rewards/rejected": -1.6640625,
"step": 700
},
{
"epoch": 0.05341960725302836,
"grad_norm": 7.056235535697858,
"learning_rate": 2.669172932330827e-07,
"logits/chosen": -3.203125,
"logits/rejected": -2.828125,
"logps/chosen": -276.0,
"logps/rejected": -348.0,
"loss": 0.5158,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.99609375,
"rewards/margins": 0.69921875,
"rewards/rejected": -1.6953125,
"step": 710
},
{
"epoch": 0.05417199608757806,
"grad_norm": 7.272525055737868,
"learning_rate": 2.7067669172932327e-07,
"logits/chosen": -3.109375,
"logits/rejected": -2.90625,
"logps/chosen": -274.0,
"logps/rejected": -348.0,
"loss": 0.4925,
"rewards/accuracies": 0.78125,
"rewards/chosen": -1.046875,
"rewards/margins": 0.8046875,
"rewards/rejected": -1.8515625,
"step": 720
},
{
"epoch": 0.054924384922127756,
"grad_norm": 8.479125069828605,
"learning_rate": 2.744360902255639e-07,
"logits/chosen": -3.03125,
"logits/rejected": -2.9375,
"logps/chosen": -274.0,
"logps/rejected": -326.0,
"loss": 0.5078,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": -1.1640625,
"rewards/margins": 0.7890625,
"rewards/rejected": -1.953125,
"step": 730
},
{
"epoch": 0.05567677375667745,
"grad_norm": 7.1392220553921,
"learning_rate": 2.781954887218045e-07,
"logits/chosen": -3.046875,
"logits/rejected": -2.828125,
"logps/chosen": -264.0,
"logps/rejected": -320.0,
"loss": 0.5066,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.015625,
"rewards/margins": 0.7109375,
"rewards/rejected": -1.7265625,
"step": 740
},
{
"epoch": 0.05642916259122715,
"grad_norm": 8.584081832995,
"learning_rate": 2.8195488721804513e-07,
"logits/chosen": -3.109375,
"logits/rejected": -2.78125,
"logps/chosen": -282.0,
"logps/rejected": -352.0,
"loss": 0.4837,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -1.15625,
"rewards/margins": 0.82421875,
"rewards/rejected": -1.984375,
"step": 750
},
{
"epoch": 0.05718155142577684,
"grad_norm": 7.646861159193735,
"learning_rate": 2.857142857142857e-07,
"logits/chosen": -2.984375,
"logits/rejected": -2.734375,
"logps/chosen": -286.0,
"logps/rejected": -340.0,
"loss": 0.4889,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": -1.2109375,
"rewards/margins": 0.72265625,
"rewards/rejected": -1.9375,
"step": 760
},
{
"epoch": 0.057933940260326534,
"grad_norm": 6.103871181309778,
"learning_rate": 2.894736842105263e-07,
"logits/chosen": -3.015625,
"logits/rejected": -2.84375,
"logps/chosen": -300.0,
"logps/rejected": -368.0,
"loss": 0.4781,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.21875,
"rewards/margins": 0.94140625,
"rewards/rejected": -2.15625,
"step": 770
},
{
"epoch": 0.058686329094876234,
"grad_norm": 7.815641392476333,
"learning_rate": 2.932330827067669e-07,
"logits/chosen": -2.953125,
"logits/rejected": -2.765625,
"logps/chosen": -272.0,
"logps/rejected": -358.0,
"loss": 0.4766,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -1.2109375,
"rewards/margins": 0.91796875,
"rewards/rejected": -2.125,
"step": 780
},
{
"epoch": 0.05943871792942593,
"grad_norm": 13.713331235769063,
"learning_rate": 2.969924812030075e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.75,
"logps/chosen": -304.0,
"logps/rejected": -356.0,
"loss": 0.4681,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.2890625,
"rewards/margins": 0.80078125,
"rewards/rejected": -2.09375,
"step": 790
},
{
"epoch": 0.06019110676397562,
"grad_norm": 6.54533749931702,
"learning_rate": 3.0075187969924807e-07,
"logits/chosen": -3.0,
"logits/rejected": -2.671875,
"logps/chosen": -282.0,
"logps/rejected": -358.0,
"loss": 0.4805,
"rewards/accuracies": 0.8125,
"rewards/chosen": -1.1640625,
"rewards/margins": 0.78515625,
"rewards/rejected": -1.9453125,
"step": 800
},
{
"epoch": 0.06094349559852532,
"grad_norm": 7.465470155476823,
"learning_rate": 3.0451127819548874e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.625,
"logps/chosen": -312.0,
"logps/rejected": -394.0,
"loss": 0.4693,
"rewards/accuracies": 0.737500011920929,
"rewards/chosen": -1.3984375,
"rewards/margins": 0.9140625,
"rewards/rejected": -2.3125,
"step": 810
},
{
"epoch": 0.06169588443307501,
"grad_norm": 9.535546928152705,
"learning_rate": 3.082706766917293e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.671875,
"logps/chosen": -300.0,
"logps/rejected": -396.0,
"loss": 0.4545,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.3125,
"rewards/margins": 1.0390625,
"rewards/rejected": -2.359375,
"step": 820
},
{
"epoch": 0.062448273267624706,
"grad_norm": 9.078214464965928,
"learning_rate": 3.1203007518796993e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.609375,
"logps/chosen": -288.0,
"logps/rejected": -394.0,
"loss": 0.4864,
"rewards/accuracies": 0.7875000238418579,
"rewards/chosen": -1.2890625,
"rewards/margins": 0.828125,
"rewards/rejected": -2.109375,
"step": 830
},
{
"epoch": 0.0632006621021744,
"grad_norm": 8.079147706507966,
"learning_rate": 3.157894736842105e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.546875,
"logps/chosen": -292.0,
"logps/rejected": -404.0,
"loss": 0.4627,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -1.3125,
"rewards/margins": 1.171875,
"rewards/rejected": -2.484375,
"step": 840
},
{
"epoch": 0.0639530509367241,
"grad_norm": 6.623406076716393,
"learning_rate": 3.195488721804511e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.640625,
"logps/chosen": -294.0,
"logps/rejected": -384.0,
"loss": 0.463,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.3125,
"rewards/margins": 1.0078125,
"rewards/rejected": -2.3125,
"step": 850
},
{
"epoch": 0.0647054397712738,
"grad_norm": 10.955204890181086,
"learning_rate": 3.233082706766917e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.71875,
"logps/chosen": -316.0,
"logps/rejected": -414.0,
"loss": 0.4563,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.40625,
"rewards/margins": 1.1171875,
"rewards/rejected": -2.53125,
"step": 860
},
{
"epoch": 0.06545782860582348,
"grad_norm": 7.434705610127428,
"learning_rate": 3.270676691729323e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.625,
"logps/chosen": -300.0,
"logps/rejected": -388.0,
"loss": 0.4604,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -1.359375,
"rewards/margins": 1.0625,
"rewards/rejected": -2.421875,
"step": 870
},
{
"epoch": 0.06621021744037318,
"grad_norm": 8.609722009537244,
"learning_rate": 3.308270676691729e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.671875,
"logps/chosen": -334.0,
"logps/rejected": -454.0,
"loss": 0.4257,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -1.765625,
"rewards/margins": 1.1640625,
"rewards/rejected": -2.9375,
"step": 880
},
{
"epoch": 0.06696260627492288,
"grad_norm": 11.999385437912036,
"learning_rate": 3.3458646616541354e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.65625,
"logps/chosen": -316.0,
"logps/rejected": -426.0,
"loss": 0.4347,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.5625,
"rewards/margins": 1.1953125,
"rewards/rejected": -2.75,
"step": 890
},
{
"epoch": 0.06771499510947257,
"grad_norm": 9.118722433060201,
"learning_rate": 3.383458646616541e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.671875,
"logps/chosen": -294.0,
"logps/rejected": -400.0,
"loss": 0.4424,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -1.5234375,
"rewards/margins": 1.1015625,
"rewards/rejected": -2.625,
"step": 900
},
{
"epoch": 0.06846738394402227,
"grad_norm": 8.90229000330093,
"learning_rate": 3.4210526315789473e-07,
"logits/chosen": -2.96875,
"logits/rejected": -2.625,
"logps/chosen": -314.0,
"logps/rejected": -450.0,
"loss": 0.4658,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -1.671875,
"rewards/margins": 1.1796875,
"rewards/rejected": -2.84375,
"step": 910
},
{
"epoch": 0.06921977277857197,
"grad_norm": 12.740993275502067,
"learning_rate": 3.458646616541353e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.671875,
"logps/chosen": -346.0,
"logps/rejected": -430.0,
"loss": 0.4558,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -1.703125,
"rewards/margins": 1.125,
"rewards/rejected": -2.828125,
"step": 920
},
{
"epoch": 0.06997216161312166,
"grad_norm": 9.335082174627217,
"learning_rate": 3.496240601503759e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.703125,
"logps/chosen": -332.0,
"logps/rejected": -458.0,
"loss": 0.4054,
"rewards/accuracies": 0.84375,
"rewards/chosen": -1.7578125,
"rewards/margins": 1.40625,
"rewards/rejected": -3.171875,
"step": 930
},
{
"epoch": 0.07072455044767136,
"grad_norm": 8.239119204332837,
"learning_rate": 3.533834586466165e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -312.0,
"logps/rejected": -446.0,
"loss": 0.4412,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.5703125,
"rewards/margins": 1.3515625,
"rewards/rejected": -2.90625,
"step": 940
},
{
"epoch": 0.07147693928222106,
"grad_norm": 7.440153278177778,
"learning_rate": 3.5714285714285716e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.578125,
"logps/chosen": -294.0,
"logps/rejected": -434.0,
"loss": 0.4392,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -1.4921875,
"rewards/margins": 1.265625,
"rewards/rejected": -2.75,
"step": 950
},
{
"epoch": 0.07222932811677074,
"grad_norm": 9.61612083498033,
"learning_rate": 3.609022556390977e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5,
"logps/chosen": -320.0,
"logps/rejected": -434.0,
"loss": 0.4324,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": -1.5390625,
"rewards/margins": 1.2578125,
"rewards/rejected": -2.796875,
"step": 960
},
{
"epoch": 0.07298171695132044,
"grad_norm": 10.276517044122064,
"learning_rate": 3.6466165413533834e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5625,
"logps/chosen": -356.0,
"logps/rejected": -448.0,
"loss": 0.4271,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -1.6875,
"rewards/margins": 1.265625,
"rewards/rejected": -2.953125,
"step": 970
},
{
"epoch": 0.07373410578587014,
"grad_norm": 7.955238750998195,
"learning_rate": 3.684210526315789e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.578125,
"logps/chosen": -350.0,
"logps/rejected": -476.0,
"loss": 0.4415,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.8515625,
"rewards/margins": 1.3046875,
"rewards/rejected": -3.15625,
"step": 980
},
{
"epoch": 0.07448649462041983,
"grad_norm": 9.228975986129836,
"learning_rate": 3.7218045112781953e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -320.0,
"logps/rejected": -442.0,
"loss": 0.4177,
"rewards/accuracies": 0.8125,
"rewards/chosen": -1.71875,
"rewards/margins": 1.265625,
"rewards/rejected": -2.984375,
"step": 990
},
{
"epoch": 0.07523888345496953,
"grad_norm": 7.6300102127807765,
"learning_rate": 3.759398496240601e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.390625,
"logps/chosen": -318.0,
"logps/rejected": -430.0,
"loss": 0.4256,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -1.546875,
"rewards/margins": 1.125,
"rewards/rejected": -2.671875,
"step": 1000
},
{
"epoch": 0.07599127228951923,
"grad_norm": 11.203017463036868,
"learning_rate": 3.796992481203007e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -322.0,
"logps/rejected": -468.0,
"loss": 0.4073,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -1.5859375,
"rewards/margins": 1.6484375,
"rewards/rejected": -3.234375,
"step": 1010
},
{
"epoch": 0.07674366112406891,
"grad_norm": 10.44396520900687,
"learning_rate": 3.8345864661654134e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -334.0,
"logps/rejected": -470.0,
"loss": 0.4036,
"rewards/accuracies": 0.8125,
"rewards/chosen": -1.71875,
"rewards/margins": 1.3828125,
"rewards/rejected": -3.09375,
"step": 1020
},
{
"epoch": 0.07749604995861861,
"grad_norm": 10.352095048437782,
"learning_rate": 3.8721804511278196e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -302.0,
"logps/rejected": -436.0,
"loss": 0.4089,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -1.484375,
"rewards/margins": 1.46875,
"rewards/rejected": -2.953125,
"step": 1030
},
{
"epoch": 0.07824843879316831,
"grad_norm": 9.027432013217565,
"learning_rate": 3.909774436090225e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.546875,
"logps/chosen": -332.0,
"logps/rejected": -444.0,
"loss": 0.422,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.7109375,
"rewards/margins": 1.1640625,
"rewards/rejected": -2.875,
"step": 1040
},
{
"epoch": 0.079000827627718,
"grad_norm": 8.493987186289845,
"learning_rate": 3.9473684210526315e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.546875,
"logps/chosen": -340.0,
"logps/rejected": -480.0,
"loss": 0.4011,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -1.6484375,
"rewards/margins": 1.4609375,
"rewards/rejected": -3.109375,
"step": 1050
},
{
"epoch": 0.0797532164622677,
"grad_norm": 13.673005824406173,
"learning_rate": 3.984962406015037e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.5,
"logps/chosen": -334.0,
"logps/rejected": -456.0,
"loss": 0.3971,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -1.6484375,
"rewards/margins": 1.3828125,
"rewards/rejected": -3.03125,
"step": 1060
},
{
"epoch": 0.0805056052968174,
"grad_norm": 11.280443761201498,
"learning_rate": 4.0225563909774433e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -344.0,
"logps/rejected": -462.0,
"loss": 0.4124,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -1.7109375,
"rewards/margins": 1.3359375,
"rewards/rejected": -3.046875,
"step": 1070
},
{
"epoch": 0.08125799413136708,
"grad_norm": 11.852849318261235,
"learning_rate": 4.060150375939849e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.5,
"logps/chosen": -388.0,
"logps/rejected": -544.0,
"loss": 0.4084,
"rewards/accuracies": 0.84375,
"rewards/chosen": -2.078125,
"rewards/margins": 1.625,
"rewards/rejected": -3.703125,
"step": 1080
},
{
"epoch": 0.08201038296591678,
"grad_norm": 8.668455285022874,
"learning_rate": 4.0977443609022557e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.59375,
"logps/chosen": -366.0,
"logps/rejected": -516.0,
"loss": 0.4122,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -1.875,
"rewards/margins": 1.609375,
"rewards/rejected": -3.484375,
"step": 1090
},
{
"epoch": 0.08276277180046648,
"grad_norm": 16.385621268125355,
"learning_rate": 4.1353383458646614e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.25,
"logps/chosen": -340.0,
"logps/rejected": -454.0,
"loss": 0.3919,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -1.6640625,
"rewards/margins": 1.40625,
"rewards/rejected": -3.078125,
"step": 1100
},
{
"epoch": 0.08351516063501617,
"grad_norm": 11.29836657890373,
"learning_rate": 4.1729323308270676e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.578125,
"logps/chosen": -368.0,
"logps/rejected": -560.0,
"loss": 0.3525,
"rewards/accuracies": 0.875,
"rewards/chosen": -2.046875,
"rewards/margins": 2.1875,
"rewards/rejected": -4.21875,
"step": 1110
},
{
"epoch": 0.08426754946956587,
"grad_norm": 14.191366008515217,
"learning_rate": 4.2105263157894733e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.6875,
"logps/chosen": -346.0,
"logps/rejected": -468.0,
"loss": 0.4123,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -1.8203125,
"rewards/margins": 1.3515625,
"rewards/rejected": -3.171875,
"step": 1120
},
{
"epoch": 0.08501993830411557,
"grad_norm": 12.571352956931529,
"learning_rate": 4.2481203007518795e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.46875,
"logps/chosen": -372.0,
"logps/rejected": -532.0,
"loss": 0.3899,
"rewards/accuracies": 0.7875000238418579,
"rewards/chosen": -2.0625,
"rewards/margins": 1.6953125,
"rewards/rejected": -3.75,
"step": 1130
},
{
"epoch": 0.08577232713866527,
"grad_norm": 9.584329220428629,
"learning_rate": 4.285714285714285e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.546875,
"logps/chosen": -396.0,
"logps/rejected": -544.0,
"loss": 0.3737,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -2.40625,
"rewards/margins": 1.421875,
"rewards/rejected": -3.828125,
"step": 1140
},
{
"epoch": 0.08652471597321496,
"grad_norm": 9.655313020328894,
"learning_rate": 4.3233082706766913e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5625,
"logps/chosen": -358.0,
"logps/rejected": -520.0,
"loss": 0.3823,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -2.0625,
"rewards/margins": 1.6484375,
"rewards/rejected": -3.703125,
"step": 1150
},
{
"epoch": 0.08727710480776465,
"grad_norm": 8.344975957766751,
"learning_rate": 4.3609022556390975e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.65625,
"logps/chosen": -424.0,
"logps/rejected": -544.0,
"loss": 0.4058,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -2.484375,
"rewards/margins": 1.34375,
"rewards/rejected": -3.828125,
"step": 1160
},
{
"epoch": 0.08802949364231435,
"grad_norm": 17.90436214681741,
"learning_rate": 4.398496240601504e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.46875,
"logps/chosen": -386.0,
"logps/rejected": -536.0,
"loss": 0.3911,
"rewards/accuracies": 0.8125,
"rewards/chosen": -2.359375,
"rewards/margins": 1.625,
"rewards/rejected": -3.984375,
"step": 1170
},
{
"epoch": 0.08878188247686404,
"grad_norm": 10.09502529582343,
"learning_rate": 4.4360902255639094e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.421875,
"logps/chosen": -358.0,
"logps/rejected": -468.0,
"loss": 0.3816,
"rewards/accuracies": 0.768750011920929,
"rewards/chosen": -2.0,
"rewards/margins": 1.203125,
"rewards/rejected": -3.203125,
"step": 1180
},
{
"epoch": 0.08953427131141374,
"grad_norm": 12.723840138089093,
"learning_rate": 4.4736842105263156e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.671875,
"logps/chosen": -386.0,
"logps/rejected": -532.0,
"loss": 0.3603,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.15625,
"rewards/margins": 1.7734375,
"rewards/rejected": -3.9375,
"step": 1190
},
{
"epoch": 0.09028666014596344,
"grad_norm": 13.550815678408945,
"learning_rate": 4.5112781954887213e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.75,
"logps/chosen": -348.0,
"logps/rejected": -510.0,
"loss": 0.3987,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -1.8203125,
"rewards/margins": 1.5390625,
"rewards/rejected": -3.359375,
"step": 1200
},
{
"epoch": 0.09103904898051313,
"grad_norm": 9.008638592876,
"learning_rate": 4.5488721804511275e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -382.0,
"logps/rejected": -576.0,
"loss": 0.3829,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -2.1875,
"rewards/margins": 1.953125,
"rewards/rejected": -4.15625,
"step": 1210
},
{
"epoch": 0.09179143781506283,
"grad_norm": 35.08614510436194,
"learning_rate": 4.586466165413533e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -330.0,
"logps/rejected": -482.0,
"loss": 0.3832,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -1.7734375,
"rewards/margins": 1.6640625,
"rewards/rejected": -3.4375,
"step": 1220
},
{
"epoch": 0.09254382664961253,
"grad_norm": 11.415330910582554,
"learning_rate": 4.62406015037594e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.59375,
"logps/chosen": -388.0,
"logps/rejected": -572.0,
"loss": 0.4032,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -2.15625,
"rewards/margins": 1.875,
"rewards/rejected": -4.03125,
"step": 1230
},
{
"epoch": 0.09329621548416221,
"grad_norm": 9.726227634850884,
"learning_rate": 4.6616541353383456e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.390625,
"logps/chosen": -360.0,
"logps/rejected": -490.0,
"loss": 0.382,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -2.0,
"rewards/margins": 1.4296875,
"rewards/rejected": -3.4375,
"step": 1240
},
{
"epoch": 0.09404860431871191,
"grad_norm": 10.938204411889682,
"learning_rate": 4.699248120300752e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.671875,
"logps/chosen": -402.0,
"logps/rejected": -524.0,
"loss": 0.3515,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.140625,
"rewards/margins": 1.5703125,
"rewards/rejected": -3.71875,
"step": 1250
},
{
"epoch": 0.09480099315326161,
"grad_norm": 11.359648859339549,
"learning_rate": 4.7368421052631574e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.703125,
"logps/chosen": -388.0,
"logps/rejected": -564.0,
"loss": 0.3956,
"rewards/accuracies": 0.84375,
"rewards/chosen": -2.265625,
"rewards/margins": 1.890625,
"rewards/rejected": -4.15625,
"step": 1260
},
{
"epoch": 0.0955533819878113,
"grad_norm": 9.557735064723317,
"learning_rate": 4.774436090225564e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.796875,
"logps/chosen": -386.0,
"logps/rejected": -548.0,
"loss": 0.3486,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -2.28125,
"rewards/margins": 1.6796875,
"rewards/rejected": -3.953125,
"step": 1270
},
{
"epoch": 0.096305770822361,
"grad_norm": 12.586703006940098,
"learning_rate": 4.812030075187969e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.90625,
"logps/chosen": -348.0,
"logps/rejected": -516.0,
"loss": 0.3639,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -1.84375,
"rewards/margins": 1.78125,
"rewards/rejected": -3.609375,
"step": 1280
},
{
"epoch": 0.0970581596569107,
"grad_norm": 11.590852438383495,
"learning_rate": 4.849624060150376e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.765625,
"logps/chosen": -396.0,
"logps/rejected": -556.0,
"loss": 0.363,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -2.328125,
"rewards/margins": 1.8515625,
"rewards/rejected": -4.1875,
"step": 1290
},
{
"epoch": 0.09781054849146038,
"grad_norm": 9.666020281314843,
"learning_rate": 4.887218045112782e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.78125,
"logps/chosen": -412.0,
"logps/rejected": -580.0,
"loss": 0.3747,
"rewards/accuracies": 0.8125,
"rewards/chosen": -2.296875,
"rewards/margins": 1.921875,
"rewards/rejected": -4.21875,
"step": 1300
},
{
"epoch": 0.09856293732601008,
"grad_norm": 17.089677017414804,
"learning_rate": 4.924812030075187e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.6875,
"logps/chosen": -432.0,
"logps/rejected": -596.0,
"loss": 0.3642,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -2.546875,
"rewards/margins": 1.9453125,
"rewards/rejected": -4.5,
"step": 1310
},
{
"epoch": 0.09931532616055978,
"grad_norm": 14.284746492926356,
"learning_rate": 4.962406015037593e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.75,
"logps/chosen": -394.0,
"logps/rejected": -536.0,
"loss": 0.356,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -2.34375,
"rewards/margins": 1.6484375,
"rewards/rejected": -3.984375,
"step": 1320
},
{
"epoch": 0.10006771499510947,
"grad_norm": 10.0309975081672,
"learning_rate": 5e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.6875,
"logps/chosen": -432.0,
"logps/rejected": -656.0,
"loss": 0.3597,
"rewards/accuracies": 0.793749988079071,
"rewards/chosen": -2.78125,
"rewards/margins": 2.21875,
"rewards/rejected": -5.0,
"step": 1330
},
{
"epoch": 0.10082010382965917,
"grad_norm": 12.11681402357444,
"learning_rate": 4.999991376679495e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.53125,
"logps/chosen": -434.0,
"logps/rejected": -616.0,
"loss": 0.3456,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -2.6875,
"rewards/margins": 1.921875,
"rewards/rejected": -4.59375,
"step": 1340
},
{
"epoch": 0.10157249266420887,
"grad_norm": 9.241869652249655,
"learning_rate": 4.999965506777466e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.484375,
"logps/chosen": -450.0,
"logps/rejected": -696.0,
"loss": 0.3523,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.875,
"rewards/margins": 2.375,
"rewards/rejected": -5.25,
"step": 1350
},
{
"epoch": 0.10232488149875855,
"grad_norm": 10.700012765427152,
"learning_rate": 4.999922390472384e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.6875,
"logps/chosen": -418.0,
"logps/rejected": -572.0,
"loss": 0.3516,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -2.484375,
"rewards/margins": 1.703125,
"rewards/rejected": -4.1875,
"step": 1360
},
{
"epoch": 0.10307727033330825,
"grad_norm": 10.778973183136214,
"learning_rate": 4.999862028061692e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.640625,
"logps/chosen": -432.0,
"logps/rejected": -584.0,
"loss": 0.3608,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -2.703125,
"rewards/margins": 1.609375,
"rewards/rejected": -4.3125,
"step": 1370
},
{
"epoch": 0.10382965916785795,
"grad_norm": 10.942848875930892,
"learning_rate": 4.99978441996181e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.609375,
"logps/chosen": -424.0,
"logps/rejected": -604.0,
"loss": 0.3596,
"rewards/accuracies": 0.875,
"rewards/chosen": -2.515625,
"rewards/margins": 1.953125,
"rewards/rejected": -4.46875,
"step": 1380
},
{
"epoch": 0.10458204800240764,
"grad_norm": 13.27627553772671,
"learning_rate": 4.999689566708128e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -428.0,
"logps/rejected": -584.0,
"loss": 0.3523,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.609375,
"rewards/margins": 1.7421875,
"rewards/rejected": -4.34375,
"step": 1390
},
{
"epoch": 0.10533443683695734,
"grad_norm": 9.279591495260535,
"learning_rate": 4.999577468955008e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.609375,
"logps/chosen": -456.0,
"logps/rejected": -644.0,
"loss": 0.3669,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.890625,
"rewards/margins": 2.046875,
"rewards/rejected": -4.9375,
"step": 1400
},
{
"epoch": 0.10608682567150704,
"grad_norm": 15.338794234291706,
"learning_rate": 4.999448127475773e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.5625,
"logps/chosen": -420.0,
"logps/rejected": -656.0,
"loss": 0.3322,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.609375,
"rewards/margins": 2.53125,
"rewards/rejected": -5.125,
"step": 1410
},
{
"epoch": 0.10683921450605673,
"grad_norm": 11.07525013808053,
"learning_rate": 4.999301543162706e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.6875,
"logps/chosen": -444.0,
"logps/rejected": -664.0,
"loss": 0.3323,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -2.8125,
"rewards/margins": 2.078125,
"rewards/rejected": -4.875,
"step": 1420
},
{
"epoch": 0.10759160334060643,
"grad_norm": 14.47067636759482,
"learning_rate": 4.999137717027041e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.765625,
"logps/chosen": -452.0,
"logps/rejected": -656.0,
"loss": 0.3365,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.046875,
"rewards/margins": 2.03125,
"rewards/rejected": -5.0625,
"step": 1430
},
{
"epoch": 0.10834399217515613,
"grad_norm": 48.895026311812195,
"learning_rate": 4.998956650198959e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.765625,
"logps/chosen": -420.0,
"logps/rejected": -604.0,
"loss": 0.3544,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -2.703125,
"rewards/margins": 1.8828125,
"rewards/rejected": -4.5625,
"step": 1440
},
{
"epoch": 0.10909638100970581,
"grad_norm": 12.33907739866595,
"learning_rate": 4.998758343927576e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.6875,
"logps/chosen": -464.0,
"logps/rejected": -628.0,
"loss": 0.3275,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.9375,
"rewards/margins": 1.765625,
"rewards/rejected": -4.6875,
"step": 1450
},
{
"epoch": 0.10984876984425551,
"grad_norm": 10.624130772065378,
"learning_rate": 4.998542799580941e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.75,
"logps/chosen": -440.0,
"logps/rejected": -628.0,
"loss": 0.3492,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -2.734375,
"rewards/margins": 2.125,
"rewards/rejected": -4.875,
"step": 1460
},
{
"epoch": 0.11060115867880521,
"grad_norm": 15.790847710499458,
"learning_rate": 4.998310018646021e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.640625,
"logps/chosen": -448.0,
"logps/rejected": -592.0,
"loss": 0.3503,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.828125,
"rewards/margins": 1.65625,
"rewards/rejected": -4.46875,
"step": 1470
},
{
"epoch": 0.1113535475133549,
"grad_norm": 13.668416666518036,
"learning_rate": 4.998060002728689e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.671875,
"logps/chosen": -504.0,
"logps/rejected": -692.0,
"loss": 0.339,
"rewards/accuracies": 0.8062499761581421,
"rewards/chosen": -3.25,
"rewards/margins": 2.125,
"rewards/rejected": -5.375,
"step": 1480
},
{
"epoch": 0.1121059363479046,
"grad_norm": 13.367505264457792,
"learning_rate": 4.99779275355372e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.765625,
"logps/chosen": -442.0,
"logps/rejected": -628.0,
"loss": 0.3514,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -2.640625,
"rewards/margins": 1.8203125,
"rewards/rejected": -4.46875,
"step": 1490
},
{
"epoch": 0.1128583251824543,
"grad_norm": 9.043259055103826,
"learning_rate": 4.997508272964775e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.59375,
"logps/chosen": -480.0,
"logps/rejected": -676.0,
"loss": 0.3178,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.09375,
"rewards/margins": 2.046875,
"rewards/rejected": -5.125,
"step": 1500
},
{
"epoch": 0.11361071401700398,
"grad_norm": 11.016220353679143,
"learning_rate": 4.997206562924387e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -528.0,
"logps/rejected": -720.0,
"loss": 0.3126,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.5,
"rewards/margins": 2.171875,
"rewards/rejected": -5.6875,
"step": 1510
},
{
"epoch": 0.11436310285155368,
"grad_norm": 9.506781512625706,
"learning_rate": 4.99688762551395e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.53125,
"logps/chosen": -508.0,
"logps/rejected": -676.0,
"loss": 0.3293,
"rewards/accuracies": 0.84375,
"rewards/chosen": -3.5,
"rewards/margins": 1.921875,
"rewards/rejected": -5.40625,
"step": 1520
},
{
"epoch": 0.11511549168610338,
"grad_norm": 9.87913772768346,
"learning_rate": 4.996551462933705e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.609375,
"logps/chosen": -472.0,
"logps/rejected": -664.0,
"loss": 0.3387,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -3.03125,
"rewards/margins": 2.109375,
"rewards/rejected": -5.15625,
"step": 1530
},
{
"epoch": 0.11586788052065307,
"grad_norm": 9.143562630664047,
"learning_rate": 4.996198077502719e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.546875,
"logps/chosen": -452.0,
"logps/rejected": -644.0,
"loss": 0.323,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.96875,
"rewards/margins": 2.078125,
"rewards/rejected": -5.03125,
"step": 1540
},
{
"epoch": 0.11662026935520277,
"grad_norm": 10.38817675345434,
"learning_rate": 4.99582747165888e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.53125,
"logps/chosen": -480.0,
"logps/rejected": -712.0,
"loss": 0.319,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.1875,
"rewards/margins": 2.0625,
"rewards/rejected": -5.25,
"step": 1550
},
{
"epoch": 0.11737265818975247,
"grad_norm": 11.492302524993693,
"learning_rate": 4.995439647958869e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.5625,
"logps/chosen": -470.0,
"logps/rejected": -664.0,
"loss": 0.3247,
"rewards/accuracies": 0.84375,
"rewards/chosen": -3.0,
"rewards/margins": 2.046875,
"rewards/rejected": -5.0625,
"step": 1560
},
{
"epoch": 0.11812504702430215,
"grad_norm": 11.709867997960567,
"learning_rate": 4.995034609078148e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.640625,
"logps/chosen": -446.0,
"logps/rejected": -644.0,
"loss": 0.3327,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.890625,
"rewards/margins": 2.0,
"rewards/rejected": -4.90625,
"step": 1570
},
{
"epoch": 0.11887743585885185,
"grad_norm": 12.146034299644466,
"learning_rate": 4.994612357810942e-07,
"logits/chosen": -2.953125,
"logits/rejected": -2.75,
"logps/chosen": -544.0,
"logps/rejected": -716.0,
"loss": 0.3127,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.5,
"rewards/margins": 2.09375,
"rewards/rejected": -5.625,
"step": 1580
},
{
"epoch": 0.11962982469340155,
"grad_norm": 37.73933841606365,
"learning_rate": 4.994172897070217e-07,
"logits/chosen": -3.015625,
"logits/rejected": -2.78125,
"logps/chosen": -486.0,
"logps/rejected": -680.0,
"loss": 0.359,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.171875,
"rewards/margins": 2.046875,
"rewards/rejected": -5.21875,
"step": 1590
},
{
"epoch": 0.12038221352795124,
"grad_norm": 8.414319765873188,
"learning_rate": 4.99371622988766e-07,
"logits/chosen": -2.96875,
"logits/rejected": -2.84375,
"logps/chosen": -448.0,
"logps/rejected": -644.0,
"loss": 0.3201,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -2.984375,
"rewards/margins": 1.921875,
"rewards/rejected": -4.90625,
"step": 1600
},
{
"epoch": 0.12113460236250094,
"grad_norm": 9.915553480092923,
"learning_rate": 4.993242359413664e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.65625,
"logps/chosen": -478.0,
"logps/rejected": -712.0,
"loss": 0.3145,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -3.203125,
"rewards/margins": 2.296875,
"rewards/rejected": -5.5,
"step": 1610
},
{
"epoch": 0.12188699119705064,
"grad_norm": 12.120046031370594,
"learning_rate": 4.992751288917297e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.796875,
"logps/chosen": -448.0,
"logps/rejected": -688.0,
"loss": 0.3246,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.875,
"rewards/margins": 2.515625,
"rewards/rejected": -5.40625,
"step": 1620
},
{
"epoch": 0.12263938003160033,
"grad_norm": 10.565988140666803,
"learning_rate": 4.992243021786285e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.578125,
"logps/chosen": -498.0,
"logps/rejected": -692.0,
"loss": 0.3196,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -3.421875,
"rewards/margins": 1.9453125,
"rewards/rejected": -5.375,
"step": 1630
},
{
"epoch": 0.12339176886615003,
"grad_norm": 9.873827794682715,
"learning_rate": 4.99171756152699e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.609375,
"logps/chosen": -470.0,
"logps/rejected": -688.0,
"loss": 0.3253,
"rewards/accuracies": 0.875,
"rewards/chosen": -2.8125,
"rewards/margins": 2.421875,
"rewards/rejected": -5.21875,
"step": 1640
},
{
"epoch": 0.12414415770069973,
"grad_norm": 11.240991564793916,
"learning_rate": 4.991174911764381e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.75,
"logps/chosen": -502.0,
"logps/rejected": -700.0,
"loss": 0.3067,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -3.359375,
"rewards/margins": 1.9609375,
"rewards/rejected": -5.3125,
"step": 1650
},
{
"epoch": 0.12489654653524941,
"grad_norm": 10.750766844675857,
"learning_rate": 4.990615076242011e-07,
"logits/chosen": -2.96875,
"logits/rejected": -2.765625,
"logps/chosen": -464.0,
"logps/rejected": -708.0,
"loss": 0.3395,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -2.75,
"rewards/margins": 2.484375,
"rewards/rejected": -5.25,
"step": 1660
},
{
"epoch": 0.12564893536979913,
"grad_norm": 20.008574618115496,
"learning_rate": 4.990038058821995e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.671875,
"logps/chosen": -524.0,
"logps/rejected": -704.0,
"loss": 0.3142,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -3.625,
"rewards/margins": 1.890625,
"rewards/rejected": -5.53125,
"step": 1670
},
{
"epoch": 0.1264013242043488,
"grad_norm": 12.148202383381813,
"learning_rate": 4.989443863484976e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.625,
"logps/chosen": -468.0,
"logps/rejected": -644.0,
"loss": 0.3193,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.078125,
"rewards/margins": 1.9453125,
"rewards/rejected": -5.03125,
"step": 1680
},
{
"epoch": 0.1271537130388985,
"grad_norm": 15.287954256111135,
"learning_rate": 4.988832494330106e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.640625,
"logps/chosen": -482.0,
"logps/rejected": -700.0,
"loss": 0.3325,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -3.15625,
"rewards/margins": 2.015625,
"rewards/rejected": -5.1875,
"step": 1690
},
{
"epoch": 0.1279061018734482,
"grad_norm": 11.463559327588364,
"learning_rate": 4.988203955575006e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.703125,
"logps/chosen": -434.0,
"logps/rejected": -656.0,
"loss": 0.2976,
"rewards/accuracies": 0.875,
"rewards/chosen": -2.640625,
"rewards/margins": 2.234375,
"rewards/rejected": -4.875,
"step": 1700
},
{
"epoch": 0.1286584907079979,
"grad_norm": 12.056679797446584,
"learning_rate": 4.987558251555755e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.5625,
"logps/chosen": -456.0,
"logps/rejected": -656.0,
"loss": 0.317,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -2.921875,
"rewards/margins": 2.078125,
"rewards/rejected": -5.0,
"step": 1710
},
{
"epoch": 0.1294108795425476,
"grad_norm": 9.484624994057103,
"learning_rate": 4.98689538672684e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -454.0,
"logps/rejected": -668.0,
"loss": 0.3096,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.859375,
"rewards/margins": 2.40625,
"rewards/rejected": -5.25,
"step": 1720
},
{
"epoch": 0.1301632683770973,
"grad_norm": 14.197927584727962,
"learning_rate": 4.986215365661137e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.6875,
"logps/chosen": -464.0,
"logps/rejected": -708.0,
"loss": 0.3089,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.015625,
"rewards/margins": 2.40625,
"rewards/rejected": -5.40625,
"step": 1730
},
{
"epoch": 0.13091565721164697,
"grad_norm": 8.583390608989294,
"learning_rate": 4.985518193049879e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.640625,
"logps/chosen": -464.0,
"logps/rejected": -652.0,
"loss": 0.3193,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.765625,
"rewards/margins": 2.15625,
"rewards/rejected": -4.90625,
"step": 1740
},
{
"epoch": 0.13166804604619667,
"grad_norm": 15.994335121139475,
"learning_rate": 4.984803873702619e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.609375,
"logps/chosen": -446.0,
"logps/rejected": -664.0,
"loss": 0.3372,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.78125,
"rewards/margins": 2.25,
"rewards/rejected": -5.03125,
"step": 1750
},
{
"epoch": 0.13242043488074637,
"grad_norm": 21.914213241176697,
"learning_rate": 4.984072412547202e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.671875,
"logps/chosen": -432.0,
"logps/rejected": -640.0,
"loss": 0.3436,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.828125,
"rewards/margins": 2.078125,
"rewards/rejected": -4.90625,
"step": 1760
},
{
"epoch": 0.13317282371529607,
"grad_norm": 11.643305728816616,
"learning_rate": 4.983323814629727e-07,
"logits/chosen": -3.046875,
"logits/rejected": -2.9375,
"logps/chosen": -420.0,
"logps/rejected": -612.0,
"loss": 0.3335,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.5625,
"rewards/margins": 1.9921875,
"rewards/rejected": -4.5625,
"step": 1770
},
{
"epoch": 0.13392521254984577,
"grad_norm": 9.9183144070942,
"learning_rate": 4.982558085114515e-07,
"logits/chosen": -3.015625,
"logits/rejected": -2.9375,
"logps/chosen": -460.0,
"logps/rejected": -644.0,
"loss": 0.349,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.015625,
"rewards/margins": 1.96875,
"rewards/rejected": -5.0,
"step": 1780
},
{
"epoch": 0.13467760138439547,
"grad_norm": 12.622299854386242,
"learning_rate": 4.981775229284068e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.671875,
"logps/chosen": -470.0,
"logps/rejected": -648.0,
"loss": 0.32,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.015625,
"rewards/margins": 1.9140625,
"rewards/rejected": -4.9375,
"step": 1790
},
{
"epoch": 0.13542999021894514,
"grad_norm": 13.726359537170527,
"learning_rate": 4.98097525253904e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.484375,
"logps/chosen": -458.0,
"logps/rejected": -652.0,
"loss": 0.3353,
"rewards/accuracies": 0.84375,
"rewards/chosen": -2.90625,
"rewards/margins": 2.015625,
"rewards/rejected": -4.9375,
"step": 1800
},
{
"epoch": 0.13618237905349484,
"grad_norm": 8.533957196814246,
"learning_rate": 4.980158160398198e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.578125,
"logps/chosen": -442.0,
"logps/rejected": -664.0,
"loss": 0.287,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -2.84375,
"rewards/margins": 2.390625,
"rewards/rejected": -5.21875,
"step": 1810
},
{
"epoch": 0.13693476788804454,
"grad_norm": 9.66875724969042,
"learning_rate": 4.979323958498378e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.8125,
"logps/chosen": -466.0,
"logps/rejected": -680.0,
"loss": 0.3176,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.953125,
"rewards/margins": 2.578125,
"rewards/rejected": -5.53125,
"step": 1820
},
{
"epoch": 0.13768715672259424,
"grad_norm": 11.673654825211232,
"learning_rate": 4.978472652594453e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.765625,
"logps/chosen": -504.0,
"logps/rejected": -760.0,
"loss": 0.3245,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.53125,
"rewards/margins": 2.578125,
"rewards/rejected": -6.125,
"step": 1830
},
{
"epoch": 0.13843954555714394,
"grad_norm": 9.459008565137387,
"learning_rate": 4.977604248559289e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.75,
"logps/chosen": -444.0,
"logps/rejected": -668.0,
"loss": 0.3099,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -2.9375,
"rewards/margins": 2.265625,
"rewards/rejected": -5.1875,
"step": 1840
},
{
"epoch": 0.13919193439169364,
"grad_norm": 10.074612458953805,
"learning_rate": 4.976718752383709e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.671875,
"logps/chosen": -504.0,
"logps/rejected": -736.0,
"loss": 0.3083,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.46875,
"rewards/margins": 2.328125,
"rewards/rejected": -5.78125,
"step": 1850
},
{
"epoch": 0.1399443232262433,
"grad_norm": 9.111393234918845,
"learning_rate": 4.975816170176445e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.5625,
"logps/chosen": -498.0,
"logps/rejected": -728.0,
"loss": 0.289,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.25,
"rewards/margins": 2.453125,
"rewards/rejected": -5.71875,
"step": 1860
},
{
"epoch": 0.140696712060793,
"grad_norm": 11.023262933345384,
"learning_rate": 4.974896508164101e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.65625,
"logps/chosen": -500.0,
"logps/rejected": -724.0,
"loss": 0.3058,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.453125,
"rewards/margins": 2.34375,
"rewards/rejected": -5.78125,
"step": 1870
},
{
"epoch": 0.1414491008953427,
"grad_norm": 9.465827793956825,
"learning_rate": 4.973959772691112e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.71875,
"logps/chosen": -482.0,
"logps/rejected": -776.0,
"loss": 0.3245,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.265625,
"rewards/margins": 2.8125,
"rewards/rejected": -6.0625,
"step": 1880
},
{
"epoch": 0.1422014897298924,
"grad_norm": 9.001035578417973,
"learning_rate": 4.973005970219692e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.6875,
"logps/chosen": -504.0,
"logps/rejected": -696.0,
"loss": 0.294,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.3125,
"rewards/margins": 2.21875,
"rewards/rejected": -5.53125,
"step": 1890
},
{
"epoch": 0.1429538785644421,
"grad_norm": 10.221159879257584,
"learning_rate": 4.972035107329796e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.484375,
"logps/chosen": -524.0,
"logps/rejected": -748.0,
"loss": 0.3002,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.515625,
"rewards/margins": 2.46875,
"rewards/rejected": -5.96875,
"step": 1900
},
{
"epoch": 0.1437062673989918,
"grad_norm": 9.557554250860324,
"learning_rate": 4.971047190719076e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.484375,
"logps/chosen": -420.0,
"logps/rejected": -688.0,
"loss": 0.2758,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -2.78125,
"rewards/margins": 2.328125,
"rewards/rejected": -5.09375,
"step": 1910
},
{
"epoch": 0.14445865623354148,
"grad_norm": 24.124984900386067,
"learning_rate": 4.970042227202828e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.640625,
"logps/chosen": -544.0,
"logps/rejected": -828.0,
"loss": 0.2653,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.9375,
"rewards/margins": 2.8125,
"rewards/rejected": -6.75,
"step": 1920
},
{
"epoch": 0.14521104506809118,
"grad_norm": 9.428412327971445,
"learning_rate": 4.969020223713948e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.578125,
"logps/chosen": -476.0,
"logps/rejected": -728.0,
"loss": 0.3094,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.15625,
"rewards/margins": 2.609375,
"rewards/rejected": -5.78125,
"step": 1930
},
{
"epoch": 0.14596343390264088,
"grad_norm": 9.244732989145508,
"learning_rate": 4.967981187302889e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.703125,
"logps/chosen": -502.0,
"logps/rejected": -760.0,
"loss": 0.2917,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.34375,
"rewards/margins": 2.640625,
"rewards/rejected": -5.96875,
"step": 1940
},
{
"epoch": 0.14671582273719058,
"grad_norm": 10.135613942587367,
"learning_rate": 4.966925125137607e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.609375,
"logps/chosen": -498.0,
"logps/rejected": -740.0,
"loss": 0.2774,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.390625,
"rewards/margins": 2.515625,
"rewards/rejected": -5.90625,
"step": 1950
},
{
"epoch": 0.14746821157174028,
"grad_norm": 15.320900000215662,
"learning_rate": 4.965852044503512e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.625,
"logps/chosen": -470.0,
"logps/rejected": -732.0,
"loss": 0.31,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.046875,
"rewards/margins": 2.453125,
"rewards/rejected": -5.5,
"step": 1960
},
{
"epoch": 0.14822060040628998,
"grad_norm": 9.157953298675203,
"learning_rate": 4.964761952803416e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.515625,
"logps/chosen": -450.0,
"logps/rejected": -660.0,
"loss": 0.3058,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -2.96875,
"rewards/margins": 2.09375,
"rewards/rejected": -5.0625,
"step": 1970
},
{
"epoch": 0.14897298924083965,
"grad_norm": 9.500153891031406,
"learning_rate": 4.963654857557488e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.46875,
"logps/chosen": -468.0,
"logps/rejected": -652.0,
"loss": 0.3027,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.03125,
"rewards/margins": 2.078125,
"rewards/rejected": -5.09375,
"step": 1980
},
{
"epoch": 0.14972537807538935,
"grad_norm": 7.779021609791956,
"learning_rate": 4.962530766403199e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.546875,
"logps/chosen": -476.0,
"logps/rejected": -712.0,
"loss": 0.2961,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.21875,
"rewards/margins": 2.390625,
"rewards/rejected": -5.625,
"step": 1990
},
{
"epoch": 0.15047776690993905,
"grad_norm": 10.341517042412354,
"learning_rate": 4.961389687095267e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.671875,
"logps/chosen": -484.0,
"logps/rejected": -696.0,
"loss": 0.2755,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -3.203125,
"rewards/margins": 2.25,
"rewards/rejected": -5.4375,
"step": 2000
},
{
"epoch": 0.15123015574448875,
"grad_norm": 10.775662545927513,
"learning_rate": 4.960231627505606e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.53125,
"logps/chosen": -490.0,
"logps/rejected": -716.0,
"loss": 0.2881,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.15625,
"rewards/margins": 2.46875,
"rewards/rejected": -5.625,
"step": 2010
},
{
"epoch": 0.15198254457903845,
"grad_norm": 10.755771111022861,
"learning_rate": 4.959056595623271e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.40625,
"logps/chosen": -500.0,
"logps/rejected": -708.0,
"loss": 0.2946,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.234375,
"rewards/margins": 2.265625,
"rewards/rejected": -5.5,
"step": 2020
},
{
"epoch": 0.15273493341358815,
"grad_norm": 8.940161113038451,
"learning_rate": 4.957864599554404e-07,
"logits/chosen": -2.421875,
"logits/rejected": -2.1875,
"logps/chosen": -440.0,
"logps/rejected": -680.0,
"loss": 0.3193,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.796875,
"rewards/margins": 2.390625,
"rewards/rejected": -5.1875,
"step": 2030
},
{
"epoch": 0.15348732224813783,
"grad_norm": 10.0217606927789,
"learning_rate": 4.956655647522176e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.40625,
"logps/chosen": -498.0,
"logps/rejected": -748.0,
"loss": 0.2761,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.1875,
"rewards/margins": 2.765625,
"rewards/rejected": -5.96875,
"step": 2040
},
{
"epoch": 0.15423971108268753,
"grad_norm": 8.538140666623187,
"learning_rate": 4.95542974786673e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.4375,
"logps/chosen": -516.0,
"logps/rejected": -752.0,
"loss": 0.2786,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.5,
"rewards/margins": 2.515625,
"rewards/rejected": -6.03125,
"step": 2050
},
{
"epoch": 0.15499209991723722,
"grad_norm": 11.092066592354108,
"learning_rate": 4.954186909045129e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.5,
"logps/chosen": -476.0,
"logps/rejected": -692.0,
"loss": 0.2978,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.15625,
"rewards/margins": 2.296875,
"rewards/rejected": -5.4375,
"step": 2060
},
{
"epoch": 0.15574448875178692,
"grad_norm": 8.810015279238543,
"learning_rate": 4.95292713963129e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.5625,
"logps/chosen": -452.0,
"logps/rejected": -684.0,
"loss": 0.3037,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -2.984375,
"rewards/margins": 2.40625,
"rewards/rejected": -5.40625,
"step": 2070
},
{
"epoch": 0.15649687758633662,
"grad_norm": 10.304083443176951,
"learning_rate": 4.951650448315929e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.53125,
"logps/chosen": -500.0,
"logps/rejected": -720.0,
"loss": 0.2742,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.359375,
"rewards/margins": 2.34375,
"rewards/rejected": -5.6875,
"step": 2080
},
{
"epoch": 0.15724926642088632,
"grad_norm": 11.038842624924929,
"learning_rate": 4.950356843906501e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.578125,
"logps/chosen": -482.0,
"logps/rejected": -736.0,
"loss": 0.3043,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.265625,
"rewards/margins": 2.5625,
"rewards/rejected": -5.8125,
"step": 2090
},
{
"epoch": 0.158001655255436,
"grad_norm": 16.086119459510908,
"learning_rate": 4.949046335327138e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.40625,
"logps/chosen": -466.0,
"logps/rejected": -684.0,
"loss": 0.2645,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -2.953125,
"rewards/margins": 2.3125,
"rewards/rejected": -5.25,
"step": 2100
},
{
"epoch": 0.1587540440899857,
"grad_norm": 10.440581591351142,
"learning_rate": 4.94771893161859e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -490.0,
"logps/rejected": -752.0,
"loss": 0.3073,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.25,
"rewards/margins": 2.734375,
"rewards/rejected": -6.0,
"step": 2110
},
{
"epoch": 0.1595064329245354,
"grad_norm": 8.118566889874467,
"learning_rate": 4.946374641938157e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.5625,
"logps/chosen": -478.0,
"logps/rejected": -684.0,
"loss": 0.3307,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.203125,
"rewards/margins": 2.1875,
"rewards/rejected": -5.375,
"step": 2120
},
{
"epoch": 0.1602588217590851,
"grad_norm": 12.613630152927428,
"learning_rate": 4.945013475559632e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.484375,
"logps/chosen": -516.0,
"logps/rejected": -728.0,
"loss": 0.2839,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.359375,
"rewards/margins": 2.296875,
"rewards/rejected": -5.65625,
"step": 2130
},
{
"epoch": 0.1610112105936348,
"grad_norm": 10.910342487887005,
"learning_rate": 4.943635441873235e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.546875,
"logps/chosen": -452.0,
"logps/rejected": -684.0,
"loss": 0.2975,
"rewards/accuracies": 0.875,
"rewards/chosen": -2.90625,
"rewards/margins": 2.359375,
"rewards/rejected": -5.28125,
"step": 2140
},
{
"epoch": 0.1617635994281845,
"grad_norm": 7.615912452628361,
"learning_rate": 4.942240550385547e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.609375,
"logps/chosen": -432.0,
"logps/rejected": -684.0,
"loss": 0.2741,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -2.71875,
"rewards/margins": 2.5625,
"rewards/rejected": -5.28125,
"step": 2150
},
{
"epoch": 0.16251598826273417,
"grad_norm": 12.776860043962529,
"learning_rate": 4.940828810719444e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.640625,
"logps/chosen": -484.0,
"logps/rejected": -740.0,
"loss": 0.2956,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.21875,
"rewards/margins": 2.6875,
"rewards/rejected": -5.90625,
"step": 2160
},
{
"epoch": 0.16326837709728387,
"grad_norm": 7.715958635368018,
"learning_rate": 4.939400232614033e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.46875,
"logps/chosen": -440.0,
"logps/rejected": -676.0,
"loss": 0.3187,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -2.78125,
"rewards/margins": 2.390625,
"rewards/rejected": -5.15625,
"step": 2170
},
{
"epoch": 0.16402076593183357,
"grad_norm": 10.909720452071825,
"learning_rate": 4.937954825924585e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.46875,
"logps/chosen": -458.0,
"logps/rejected": -656.0,
"loss": 0.2798,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.015625,
"rewards/margins": 2.140625,
"rewards/rejected": -5.15625,
"step": 2180
},
{
"epoch": 0.16477315476638327,
"grad_norm": 11.056113501762782,
"learning_rate": 4.936492600622464e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.40625,
"logps/chosen": -552.0,
"logps/rejected": -756.0,
"loss": 0.3351,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.6875,
"rewards/margins": 2.328125,
"rewards/rejected": -6.03125,
"step": 2190
},
{
"epoch": 0.16552554360093297,
"grad_norm": 9.181288831576246,
"learning_rate": 4.935013566795058e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.5625,
"logps/chosen": -494.0,
"logps/rejected": -728.0,
"loss": 0.2665,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.421875,
"rewards/margins": 2.3125,
"rewards/rejected": -5.71875,
"step": 2200
},
{
"epoch": 0.16627793243548267,
"grad_norm": 15.421194070464846,
"learning_rate": 4.933517734645714e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.671875,
"logps/chosen": -482.0,
"logps/rejected": -816.0,
"loss": 0.2725,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.390625,
"rewards/margins": 3.1875,
"rewards/rejected": -6.59375,
"step": 2210
},
{
"epoch": 0.16703032127003234,
"grad_norm": 7.732850963579106,
"learning_rate": 4.932005114493665e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.640625,
"logps/chosen": -532.0,
"logps/rejected": -808.0,
"loss": 0.2775,
"rewards/accuracies": 0.84375,
"rewards/chosen": -3.703125,
"rewards/margins": 2.8125,
"rewards/rejected": -6.5,
"step": 2220
},
{
"epoch": 0.16778271010458204,
"grad_norm": 8.73231439791769,
"learning_rate": 4.930475716773956e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.546875,
"logps/chosen": -502.0,
"logps/rejected": -752.0,
"loss": 0.2924,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.390625,
"rewards/margins": 2.53125,
"rewards/rejected": -5.9375,
"step": 2230
},
{
"epoch": 0.16853509893913174,
"grad_norm": 9.123008878161743,
"learning_rate": 4.928929552037378e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.34375,
"logps/chosen": -492.0,
"logps/rejected": -712.0,
"loss": 0.2833,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.328125,
"rewards/margins": 2.40625,
"rewards/rejected": -5.71875,
"step": 2240
},
{
"epoch": 0.16928748777368144,
"grad_norm": 12.07063346888074,
"learning_rate": 4.927366630950389e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.484375,
"logps/chosen": -468.0,
"logps/rejected": -700.0,
"loss": 0.2693,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.09375,
"rewards/margins": 2.46875,
"rewards/rejected": -5.5625,
"step": 2250
},
{
"epoch": 0.17003987660823114,
"grad_norm": 9.58671682334943,
"learning_rate": 4.925786964295046e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.328125,
"logps/chosen": -568.0,
"logps/rejected": -808.0,
"loss": 0.3029,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -4.03125,
"rewards/margins": 2.578125,
"rewards/rejected": -6.59375,
"step": 2260
},
{
"epoch": 0.17079226544278084,
"grad_norm": 9.336994205772083,
"learning_rate": 4.924190562968926e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.453125,
"logps/chosen": -512.0,
"logps/rejected": -716.0,
"loss": 0.2713,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.4375,
"rewards/margins": 2.296875,
"rewards/rejected": -5.75,
"step": 2270
},
{
"epoch": 0.17154465427733054,
"grad_norm": 11.646865344527923,
"learning_rate": 4.922577437985052e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.625,
"logps/chosen": -498.0,
"logps/rejected": -760.0,
"loss": 0.2462,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.421875,
"rewards/margins": 2.609375,
"rewards/rejected": -6.03125,
"step": 2280
},
{
"epoch": 0.1722970431118802,
"grad_norm": 12.116174831993108,
"learning_rate": 4.920947600471821e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.59375,
"logps/chosen": -496.0,
"logps/rejected": -764.0,
"loss": 0.2828,
"rewards/accuracies": 0.84375,
"rewards/chosen": -3.375,
"rewards/margins": 2.734375,
"rewards/rejected": -6.09375,
"step": 2290
},
{
"epoch": 0.1730494319464299,
"grad_norm": 9.542246025337626,
"learning_rate": 4.919301061672919e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.53125,
"logps/chosen": -488.0,
"logps/rejected": -736.0,
"loss": 0.2701,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.3125,
"rewards/margins": 2.53125,
"rewards/rejected": -5.84375,
"step": 2300
},
{
"epoch": 0.1738018207809796,
"grad_norm": 10.606436086492515,
"learning_rate": 4.917637832947256e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.4375,
"logps/chosen": -572.0,
"logps/rejected": -804.0,
"loss": 0.2781,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.9375,
"rewards/margins": 2.5,
"rewards/rejected": -6.4375,
"step": 2310
},
{
"epoch": 0.1745542096155293,
"grad_norm": 10.611161178350386,
"learning_rate": 4.915957925768871e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.609375,
"logps/chosen": -520.0,
"logps/rejected": -748.0,
"loss": 0.2981,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.734375,
"rewards/margins": 2.265625,
"rewards/rejected": -6.0,
"step": 2320
},
{
"epoch": 0.175306598450079,
"grad_norm": 12.829898396212654,
"learning_rate": 4.914261351726868e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.625,
"logps/chosen": -560.0,
"logps/rejected": -784.0,
"loss": 0.2684,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.78125,
"rewards/margins": 2.421875,
"rewards/rejected": -6.1875,
"step": 2330
},
{
"epoch": 0.1760589872846287,
"grad_norm": 10.222444110704744,
"learning_rate": 4.91254812252533e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -510.0,
"logps/rejected": -904.0,
"loss": 0.2601,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.5625,
"rewards/margins": 3.734375,
"rewards/rejected": -7.3125,
"step": 2340
},
{
"epoch": 0.17681137611917838,
"grad_norm": 9.955713067057467,
"learning_rate": 4.910818249983235e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.65625,
"logps/chosen": -548.0,
"logps/rejected": -828.0,
"loss": 0.283,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.921875,
"rewards/margins": 2.734375,
"rewards/rejected": -6.65625,
"step": 2350
},
{
"epoch": 0.17756376495372808,
"grad_norm": 9.899985953459382,
"learning_rate": 4.909071746034379e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.40625,
"logps/chosen": -588.0,
"logps/rejected": -804.0,
"loss": 0.2834,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.15625,
"rewards/margins": 2.3125,
"rewards/rejected": -6.46875,
"step": 2360
},
{
"epoch": 0.17831615378827778,
"grad_norm": 9.734416734362911,
"learning_rate": 4.907308622727293e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.375,
"logps/chosen": -494.0,
"logps/rejected": -748.0,
"loss": 0.2868,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.4375,
"rewards/margins": 2.625,
"rewards/rejected": -6.0625,
"step": 2370
},
{
"epoch": 0.17906854262282748,
"grad_norm": 8.463086487753493,
"learning_rate": 4.90552889222516e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.4375,
"logps/chosen": -540.0,
"logps/rejected": -824.0,
"loss": 0.2605,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.671875,
"rewards/margins": 2.75,
"rewards/rejected": -6.4375,
"step": 2380
},
{
"epoch": 0.17982093145737718,
"grad_norm": 13.146535023342294,
"learning_rate": 4.903732566805727e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.5625,
"logps/chosen": -560.0,
"logps/rejected": -844.0,
"loss": 0.2833,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.890625,
"rewards/margins": 2.9375,
"rewards/rejected": -6.84375,
"step": 2390
},
{
"epoch": 0.18057332029192688,
"grad_norm": 9.711136432827589,
"learning_rate": 4.901919658861228e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.453125,
"logps/chosen": -496.0,
"logps/rejected": -680.0,
"loss": 0.3008,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -3.296875,
"rewards/margins": 2.078125,
"rewards/rejected": -5.375,
"step": 2400
},
{
"epoch": 0.18132570912647655,
"grad_norm": 12.52993235918333,
"learning_rate": 4.900090180898292e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.4375,
"logps/chosen": -524.0,
"logps/rejected": -792.0,
"loss": 0.2695,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -3.5625,
"rewards/margins": 2.78125,
"rewards/rejected": -6.34375,
"step": 2410
},
{
"epoch": 0.18207809796102625,
"grad_norm": 7.896403216710635,
"learning_rate": 4.898244145537857e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.375,
"logps/chosen": -510.0,
"logps/rejected": -756.0,
"loss": 0.2626,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.515625,
"rewards/margins": 2.484375,
"rewards/rejected": -6.0,
"step": 2420
},
{
"epoch": 0.18283048679557595,
"grad_norm": 8.144329513203688,
"learning_rate": 4.896381565515087e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.5,
"logps/chosen": -500.0,
"logps/rejected": -740.0,
"loss": 0.2379,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.390625,
"rewards/margins": 2.65625,
"rewards/rejected": -6.03125,
"step": 2430
},
{
"epoch": 0.18358287563012565,
"grad_norm": 9.329496514618313,
"learning_rate": 4.894502453679284e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -482.0,
"logps/rejected": -764.0,
"loss": 0.2785,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.203125,
"rewards/margins": 2.90625,
"rewards/rejected": -6.125,
"step": 2440
},
{
"epoch": 0.18433526446467535,
"grad_norm": 12.10939641210324,
"learning_rate": 4.892606822993793e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.578125,
"logps/chosen": -436.0,
"logps/rejected": -708.0,
"loss": 0.299,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -2.8125,
"rewards/margins": 2.640625,
"rewards/rejected": -5.4375,
"step": 2450
},
{
"epoch": 0.18508765329922505,
"grad_norm": 9.857872781155606,
"learning_rate": 4.890694686535918e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.375,
"logps/chosen": -510.0,
"logps/rejected": -784.0,
"loss": 0.2701,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.28125,
"rewards/margins": 2.9375,
"rewards/rejected": -6.21875,
"step": 2460
},
{
"epoch": 0.18584004213377472,
"grad_norm": 17.101449726474733,
"learning_rate": 4.888766057496833e-07,
"logits/chosen": -2.46875,
"logits/rejected": -2.3125,
"logps/chosen": -472.0,
"logps/rejected": -728.0,
"loss": 0.2916,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -3.203125,
"rewards/margins": 2.484375,
"rewards/rejected": -5.71875,
"step": 2470
},
{
"epoch": 0.18659243096832442,
"grad_norm": 8.347297641682713,
"learning_rate": 4.886820949181486e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.5,
"logps/chosen": -438.0,
"logps/rejected": -736.0,
"loss": 0.2462,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -2.90625,
"rewards/margins": 2.96875,
"rewards/rejected": -5.875,
"step": 2480
},
{
"epoch": 0.18734481980287412,
"grad_norm": 9.678661384787706,
"learning_rate": 4.884859375008512e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.515625,
"logps/chosen": -486.0,
"logps/rejected": -776.0,
"loss": 0.2875,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.0625,
"rewards/margins": 2.9375,
"rewards/rejected": -6.0,
"step": 2490
},
{
"epoch": 0.18809720863742382,
"grad_norm": 11.145929450555997,
"learning_rate": 4.882881348510136e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.578125,
"logps/chosen": -556.0,
"logps/rejected": -856.0,
"loss": 0.2611,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.84375,
"rewards/margins": 3.171875,
"rewards/rejected": -7.03125,
"step": 2500
},
{
"epoch": 0.18884959747197352,
"grad_norm": 7.1935463269752065,
"learning_rate": 4.880886883332083e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.296875,
"logps/chosen": -516.0,
"logps/rejected": -848.0,
"loss": 0.2421,
"rewards/accuracies": 0.9375,
"rewards/chosen": -3.4375,
"rewards/margins": 3.375,
"rewards/rejected": -6.8125,
"step": 2510
},
{
"epoch": 0.18960198630652322,
"grad_norm": 10.65990633403162,
"learning_rate": 4.878875993233486e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.34375,
"logps/chosen": -516.0,
"logps/rejected": -776.0,
"loss": 0.2631,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -3.703125,
"rewards/margins": 2.515625,
"rewards/rejected": -6.21875,
"step": 2520
},
{
"epoch": 0.1903543751410729,
"grad_norm": 10.814709273380368,
"learning_rate": 4.876848692086782e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -528.0,
"logps/rejected": -796.0,
"loss": 0.2662,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.515625,
"rewards/margins": 2.71875,
"rewards/rejected": -6.25,
"step": 2530
},
{
"epoch": 0.1911067639756226,
"grad_norm": 8.274113340101351,
"learning_rate": 4.874804993877625e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.328125,
"logps/chosen": -524.0,
"logps/rejected": -768.0,
"loss": 0.2679,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.71875,
"rewards/margins": 2.5,
"rewards/rejected": -6.21875,
"step": 2540
},
{
"epoch": 0.1918591528101723,
"grad_norm": 7.8585668700799465,
"learning_rate": 4.872744912704788e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.421875,
"logps/chosen": -480.0,
"logps/rejected": -768.0,
"loss": 0.2575,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -3.21875,
"rewards/margins": 3.078125,
"rewards/rejected": -6.28125,
"step": 2550
},
{
"epoch": 0.192611541644722,
"grad_norm": 9.359705035781545,
"learning_rate": 4.870668462780062e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.4375,
"logps/chosen": -492.0,
"logps/rejected": -748.0,
"loss": 0.2643,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.34375,
"rewards/margins": 2.671875,
"rewards/rejected": -6.03125,
"step": 2560
},
{
"epoch": 0.1933639304792717,
"grad_norm": 10.260726382149574,
"learning_rate": 4.868575658428163e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -474.0,
"logps/rejected": -704.0,
"loss": 0.2763,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.1875,
"rewards/margins": 2.453125,
"rewards/rejected": -5.625,
"step": 2570
},
{
"epoch": 0.1941163193138214,
"grad_norm": 14.078164696272687,
"learning_rate": 4.866466514086628e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.53125,
"logps/chosen": -480.0,
"logps/rejected": -752.0,
"loss": 0.2753,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.03125,
"rewards/margins": 2.984375,
"rewards/rejected": -6.0,
"step": 2580
},
{
"epoch": 0.19486870814837107,
"grad_norm": 8.453767523010452,
"learning_rate": 4.864341044305719e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.453125,
"logps/chosen": -524.0,
"logps/rejected": -720.0,
"loss": 0.2626,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -3.421875,
"rewards/margins": 2.375,
"rewards/rejected": -5.78125,
"step": 2590
},
{
"epoch": 0.19562109698292077,
"grad_norm": 10.18587974692239,
"learning_rate": 4.862199263748323e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.296875,
"logps/chosen": -536.0,
"logps/rejected": -800.0,
"loss": 0.2657,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.78125,
"rewards/margins": 2.84375,
"rewards/rejected": -6.625,
"step": 2600
},
{
"epoch": 0.19637348581747047,
"grad_norm": 11.079570317722963,
"learning_rate": 4.860041187189846e-07,
"logits/chosen": -2.328125,
"logits/rejected": -2.171875,
"logps/chosen": -504.0,
"logps/rejected": -748.0,
"loss": 0.2942,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.46875,
"rewards/margins": 2.46875,
"rewards/rejected": -5.9375,
"step": 2610
},
{
"epoch": 0.19712587465202017,
"grad_norm": 7.9462869418389745,
"learning_rate": 4.857866829518118e-07,
"logits/chosen": -2.484375,
"logits/rejected": -2.3125,
"logps/chosen": -494.0,
"logps/rejected": -788.0,
"loss": 0.2422,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.4375,
"rewards/margins": 2.734375,
"rewards/rejected": -6.15625,
"step": 2620
},
{
"epoch": 0.19787826348656987,
"grad_norm": 10.695129742443964,
"learning_rate": 4.855676205733287e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.59375,
"logps/chosen": -552.0,
"logps/rejected": -832.0,
"loss": 0.2931,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.828125,
"rewards/margins": 2.890625,
"rewards/rejected": -6.71875,
"step": 2630
},
{
"epoch": 0.19863065232111957,
"grad_norm": 13.909715185144693,
"learning_rate": 4.853469330947712e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.640625,
"logps/chosen": -516.0,
"logps/rejected": -772.0,
"loss": 0.2719,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -3.328125,
"rewards/margins": 2.84375,
"rewards/rejected": -6.1875,
"step": 2640
},
{
"epoch": 0.19938304115566924,
"grad_norm": 7.977293712578769,
"learning_rate": 4.851246220385862e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.421875,
"logps/chosen": -548.0,
"logps/rejected": -776.0,
"loss": 0.2662,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.796875,
"rewards/margins": 2.546875,
"rewards/rejected": -6.34375,
"step": 2650
},
{
"epoch": 0.20013542999021894,
"grad_norm": 7.690656776892857,
"learning_rate": 4.849006889384217e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.25,
"logps/chosen": -572.0,
"logps/rejected": -768.0,
"loss": 0.2768,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.9375,
"rewards/margins": 2.21875,
"rewards/rejected": -6.15625,
"step": 2660
},
{
"epoch": 0.20088781882476864,
"grad_norm": 8.564652490755712,
"learning_rate": 4.84675135339115e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.3125,
"logps/chosen": -494.0,
"logps/rejected": -768.0,
"loss": 0.2813,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.390625,
"rewards/margins": 2.84375,
"rewards/rejected": -6.21875,
"step": 2670
},
{
"epoch": 0.20164020765931834,
"grad_norm": 6.442833431737132,
"learning_rate": 4.84447962796683e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.421875,
"logps/chosen": -580.0,
"logps/rejected": -844.0,
"loss": 0.2707,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.0,
"rewards/margins": 2.796875,
"rewards/rejected": -6.78125,
"step": 2680
},
{
"epoch": 0.20239259649386804,
"grad_norm": 11.556397527239499,
"learning_rate": 4.842191728783107e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.453125,
"logps/chosen": -520.0,
"logps/rejected": -756.0,
"loss": 0.2938,
"rewards/accuracies": 0.84375,
"rewards/chosen": -3.578125,
"rewards/margins": 2.390625,
"rewards/rejected": -5.96875,
"step": 2690
},
{
"epoch": 0.20314498532841774,
"grad_norm": 10.766048526372968,
"learning_rate": 4.839887671623414e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.46875,
"logps/chosen": -478.0,
"logps/rejected": -768.0,
"loss": 0.2471,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.21875,
"rewards/margins": 2.9375,
"rewards/rejected": -6.15625,
"step": 2700
},
{
"epoch": 0.2038973741629674,
"grad_norm": 13.5244223811314,
"learning_rate": 4.837567472382651e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.34375,
"logps/chosen": -536.0,
"logps/rejected": -812.0,
"loss": 0.2909,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.828125,
"rewards/margins": 2.796875,
"rewards/rejected": -6.625,
"step": 2710
},
{
"epoch": 0.2046497629975171,
"grad_norm": 9.738583048600992,
"learning_rate": 4.835231147067072e-07,
"logits/chosen": -2.484375,
"logits/rejected": -2.359375,
"logps/chosen": -516.0,
"logps/rejected": -764.0,
"loss": 0.2766,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -3.609375,
"rewards/margins": 2.453125,
"rewards/rejected": -6.0625,
"step": 2720
},
{
"epoch": 0.2054021518320668,
"grad_norm": 10.261504032860017,
"learning_rate": 4.832878711794185e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.390625,
"logps/chosen": -516.0,
"logps/rejected": -860.0,
"loss": 0.2551,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -3.578125,
"rewards/margins": 3.375,
"rewards/rejected": -6.9375,
"step": 2730
},
{
"epoch": 0.2061545406666165,
"grad_norm": 6.5383786738976415,
"learning_rate": 4.830510182792632e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.453125,
"logps/chosen": -544.0,
"logps/rejected": -824.0,
"loss": 0.2445,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.8125,
"rewards/margins": 2.90625,
"rewards/rejected": -6.71875,
"step": 2740
},
{
"epoch": 0.2069069295011662,
"grad_norm": 10.447397489817225,
"learning_rate": 4.82812557640208e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.515625,
"logps/chosen": -552.0,
"logps/rejected": -784.0,
"loss": 0.2563,
"rewards/accuracies": 0.8374999761581421,
"rewards/chosen": -3.671875,
"rewards/margins": 2.828125,
"rewards/rejected": -6.5,
"step": 2750
},
{
"epoch": 0.2076593183357159,
"grad_norm": 9.348554817009415,
"learning_rate": 4.82572490907311e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.46875,
"logps/chosen": -552.0,
"logps/rejected": -808.0,
"loss": 0.2576,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.765625,
"rewards/margins": 2.90625,
"rewards/rejected": -6.6875,
"step": 2760
},
{
"epoch": 0.20841170717026558,
"grad_norm": 8.70381283312396,
"learning_rate": 4.8233081973671e-07,
"logits/chosen": -2.46875,
"logits/rejected": -2.25,
"logps/chosen": -532.0,
"logps/rejected": -800.0,
"loss": 0.2484,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.8125,
"rewards/margins": 2.640625,
"rewards/rejected": -6.4375,
"step": 2770
},
{
"epoch": 0.20916409600481528,
"grad_norm": 7.4097108054434955,
"learning_rate": 4.820875457956115e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.515625,
"logps/chosen": -548.0,
"logps/rejected": -828.0,
"loss": 0.2774,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.875,
"rewards/margins": 2.875,
"rewards/rejected": -6.75,
"step": 2780
},
{
"epoch": 0.20991648483936498,
"grad_norm": 8.929321008061232,
"learning_rate": 4.818426707622788e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.390625,
"logps/chosen": -494.0,
"logps/rejected": -732.0,
"loss": 0.257,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -3.265625,
"rewards/margins": 2.546875,
"rewards/rejected": -5.8125,
"step": 2790
},
{
"epoch": 0.21066887367391468,
"grad_norm": 8.981128088011406,
"learning_rate": 4.815961963260207e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.46875,
"logps/chosen": -490.0,
"logps/rejected": -764.0,
"loss": 0.2419,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.21875,
"rewards/margins": 2.953125,
"rewards/rejected": -6.1875,
"step": 2800
},
{
"epoch": 0.21142126250846438,
"grad_norm": 10.05509938002628,
"learning_rate": 4.813481241871794e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.421875,
"logps/chosen": -484.0,
"logps/rejected": -780.0,
"loss": 0.2477,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.234375,
"rewards/margins": 2.9375,
"rewards/rejected": -6.15625,
"step": 2810
},
{
"epoch": 0.21217365134301408,
"grad_norm": 11.72370585695034,
"learning_rate": 4.810984560571195e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.328125,
"logps/chosen": -472.0,
"logps/rejected": -772.0,
"loss": 0.2559,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.21875,
"rewards/margins": 3.234375,
"rewards/rejected": -6.4375,
"step": 2820
},
{
"epoch": 0.21292604017756375,
"grad_norm": 9.647642410986215,
"learning_rate": 4.808471936582156e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -472.0,
"logps/rejected": -764.0,
"loss": 0.2374,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.140625,
"rewards/margins": 3.03125,
"rewards/rejected": -6.15625,
"step": 2830
},
{
"epoch": 0.21367842901211345,
"grad_norm": 11.420752468431145,
"learning_rate": 4.805943387238409e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -464.0,
"logps/rejected": -736.0,
"loss": 0.252,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.0,
"rewards/margins": 2.921875,
"rewards/rejected": -5.90625,
"step": 2840
},
{
"epoch": 0.21443081784666315,
"grad_norm": 9.785252828322875,
"learning_rate": 4.803398929983543e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -476.0,
"logps/rejected": -760.0,
"loss": 0.2856,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.21875,
"rewards/margins": 2.90625,
"rewards/rejected": -6.125,
"step": 2850
},
{
"epoch": 0.21518320668121285,
"grad_norm": 11.971762302414026,
"learning_rate": 4.800838582370898e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.34375,
"logps/chosen": -464.0,
"logps/rejected": -684.0,
"loss": 0.2718,
"rewards/accuracies": 0.8125,
"rewards/chosen": -3.09375,
"rewards/margins": 2.234375,
"rewards/rejected": -5.3125,
"step": 2860
},
{
"epoch": 0.21593559551576255,
"grad_norm": 8.477706468059246,
"learning_rate": 4.79826236206343e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.171875,
"logps/chosen": -488.0,
"logps/rejected": -724.0,
"loss": 0.2513,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.25,
"rewards/margins": 2.390625,
"rewards/rejected": -5.625,
"step": 2870
},
{
"epoch": 0.21668798435031225,
"grad_norm": 11.869962119536272,
"learning_rate": 4.795670286833599e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.328125,
"logps/chosen": -512.0,
"logps/rejected": -784.0,
"loss": 0.2611,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.34375,
"rewards/margins": 2.859375,
"rewards/rejected": -6.1875,
"step": 2880
},
{
"epoch": 0.21744037318486192,
"grad_norm": 8.487625004734987,
"learning_rate": 4.79306237456324e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.375,
"logps/chosen": -512.0,
"logps/rejected": -764.0,
"loss": 0.265,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.5625,
"rewards/margins": 2.515625,
"rewards/rejected": -6.0625,
"step": 2890
},
{
"epoch": 0.21819276201941162,
"grad_norm": 11.615793395248453,
"learning_rate": 4.790438643243447e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -500.0,
"logps/rejected": -752.0,
"loss": 0.2533,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -3.140625,
"rewards/margins": 2.828125,
"rewards/rejected": -5.96875,
"step": 2900
},
{
"epoch": 0.21894515085396132,
"grad_norm": 10.318914547631199,
"learning_rate": 4.787799110974436e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.65625,
"logps/chosen": -516.0,
"logps/rejected": -816.0,
"loss": 0.2414,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.5625,
"rewards/margins": 3.21875,
"rewards/rejected": -6.78125,
"step": 2910
},
{
"epoch": 0.21969753968851102,
"grad_norm": 12.904513672408855,
"learning_rate": 4.785143795965437e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.71875,
"logps/chosen": -528.0,
"logps/rejected": -784.0,
"loss": 0.2577,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": -3.640625,
"rewards/margins": 2.640625,
"rewards/rejected": -6.28125,
"step": 2920
},
{
"epoch": 0.22044992852306072,
"grad_norm": 8.687685973219471,
"learning_rate": 4.782472716534554e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.6875,
"logps/chosen": -576.0,
"logps/rejected": -860.0,
"loss": 0.25,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.96875,
"rewards/margins": 2.875,
"rewards/rejected": -6.84375,
"step": 2930
},
{
"epoch": 0.22120231735761042,
"grad_norm": 8.425580278247608,
"learning_rate": 4.779785891108647e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5625,
"logps/chosen": -540.0,
"logps/rejected": -776.0,
"loss": 0.2682,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -3.734375,
"rewards/margins": 2.546875,
"rewards/rejected": -6.28125,
"step": 2940
},
{
"epoch": 0.2219547061921601,
"grad_norm": 8.92649200504024,
"learning_rate": 4.777083338223202e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -572.0,
"logps/rejected": -840.0,
"loss": 0.2475,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.25,
"rewards/margins": 2.671875,
"rewards/rejected": -6.90625,
"step": 2950
},
{
"epoch": 0.2227070950267098,
"grad_norm": 9.22059724524348,
"learning_rate": 4.774365076522202e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.625,
"logps/chosen": -564.0,
"logps/rejected": -804.0,
"loss": 0.2469,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.765625,
"rewards/margins": 2.765625,
"rewards/rejected": -6.53125,
"step": 2960
},
{
"epoch": 0.2234594838612595,
"grad_norm": 9.454530660849164,
"learning_rate": 4.771631124758e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -520.0,
"logps/rejected": -804.0,
"loss": 0.2499,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.6875,
"rewards/margins": 2.765625,
"rewards/rejected": -6.46875,
"step": 2970
},
{
"epoch": 0.2242118726958092,
"grad_norm": 9.480975232151959,
"learning_rate": 4.76888150179119e-07,
"logits/chosen": -2.984375,
"logits/rejected": -2.671875,
"logps/chosen": -512.0,
"logps/rejected": -828.0,
"loss": 0.2496,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -3.484375,
"rewards/margins": 3.21875,
"rewards/rejected": -6.71875,
"step": 2980
},
{
"epoch": 0.2249642615303589,
"grad_norm": 7.916418722194274,
"learning_rate": 4.7661162265904773e-07,
"logits/chosen": -3.109375,
"logits/rejected": -2.8125,
"logps/chosen": -564.0,
"logps/rejected": -840.0,
"loss": 0.2273,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -3.90625,
"rewards/margins": 2.984375,
"rewards/rejected": -6.875,
"step": 2990
},
{
"epoch": 0.2257166503649086,
"grad_norm": 8.229870586292975,
"learning_rate": 4.7633353182325443e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.578125,
"logps/chosen": -536.0,
"logps/rejected": -880.0,
"loss": 0.2373,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.84375,
"rewards/margins": 3.28125,
"rewards/rejected": -7.125,
"step": 3000
},
{
"epoch": 0.22646903919945827,
"grad_norm": 10.226118184963164,
"learning_rate": 4.760538795901923e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.609375,
"logps/chosen": -548.0,
"logps/rejected": -848.0,
"loss": 0.2677,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -3.828125,
"rewards/margins": 3.203125,
"rewards/rejected": -7.03125,
"step": 3010
},
{
"epoch": 0.22722142803400797,
"grad_norm": 8.49693960463905,
"learning_rate": 4.757726678890859e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -506.0,
"logps/rejected": -800.0,
"loss": 0.254,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.40625,
"rewards/margins": 3.140625,
"rewards/rejected": -6.5625,
"step": 3020
},
{
"epoch": 0.22797381686855767,
"grad_norm": 9.5155953716937,
"learning_rate": 4.754898986599182e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.46875,
"logps/chosen": -484.0,
"logps/rejected": -716.0,
"loss": 0.2568,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.234375,
"rewards/margins": 2.46875,
"rewards/rejected": -5.71875,
"step": 3030
},
{
"epoch": 0.22872620570310737,
"grad_norm": 9.050451730524532,
"learning_rate": 4.75205573853417e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.328125,
"logps/chosen": -564.0,
"logps/rejected": -832.0,
"loss": 0.2664,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.03125,
"rewards/margins": 2.890625,
"rewards/rejected": -6.9375,
"step": 3040
},
{
"epoch": 0.22947859453765707,
"grad_norm": 11.230060400251109,
"learning_rate": 4.749196954310414e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.515625,
"logps/chosen": -580.0,
"logps/rejected": -880.0,
"loss": 0.2462,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -4.09375,
"rewards/margins": 3.171875,
"rewards/rejected": -7.25,
"step": 3050
},
{
"epoch": 0.23023098337220677,
"grad_norm": 8.780783249418834,
"learning_rate": 4.746322653649683e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.59375,
"logps/chosen": -564.0,
"logps/rejected": -856.0,
"loss": 0.252,
"rewards/accuracies": 0.84375,
"rewards/chosen": -4.09375,
"rewards/margins": 2.890625,
"rewards/rejected": -6.96875,
"step": 3060
},
{
"epoch": 0.23098337220675647,
"grad_norm": 9.12154565614575,
"learning_rate": 4.7434328563807913e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -528.0,
"logps/rejected": -872.0,
"loss": 0.2577,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.734375,
"rewards/margins": 3.296875,
"rewards/rejected": -7.03125,
"step": 3070
},
{
"epoch": 0.23173576104130614,
"grad_norm": 14.056073227705715,
"learning_rate": 4.7405275824394574e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.578125,
"logps/chosen": -568.0,
"logps/rejected": -796.0,
"loss": 0.2442,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.984375,
"rewards/margins": 2.484375,
"rewards/rejected": -6.46875,
"step": 3080
},
{
"epoch": 0.23248814987585584,
"grad_norm": 9.397850408648406,
"learning_rate": 4.737606851868167e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.578125,
"logps/chosen": -536.0,
"logps/rejected": -812.0,
"loss": 0.2361,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.546875,
"rewards/margins": 3.1875,
"rewards/rejected": -6.75,
"step": 3090
},
{
"epoch": 0.23324053871040554,
"grad_norm": 8.73361335302379,
"learning_rate": 4.734670684816037e-07,
"logits/chosen": -2.4375,
"logits/rejected": -2.3125,
"logps/chosen": -510.0,
"logps/rejected": -820.0,
"loss": 0.2494,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -3.203125,
"rewards/margins": 3.15625,
"rewards/rejected": -6.34375,
"step": 3100
},
{
"epoch": 0.23399292754495524,
"grad_norm": 10.750693728458632,
"learning_rate": 4.7317191015386744e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.53125,
"logps/chosen": -510.0,
"logps/rejected": -808.0,
"loss": 0.2487,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.671875,
"rewards/margins": 2.984375,
"rewards/rejected": -6.625,
"step": 3110
},
{
"epoch": 0.23474531637950494,
"grad_norm": 7.196445794814489,
"learning_rate": 4.7287521223980395e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.46875,
"logps/chosen": -506.0,
"logps/rejected": -756.0,
"loss": 0.269,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -3.28125,
"rewards/margins": 2.6875,
"rewards/rejected": -6.0,
"step": 3120
},
{
"epoch": 0.23549770521405464,
"grad_norm": 8.778673968279726,
"learning_rate": 4.725769767862301e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -544.0,
"logps/rejected": -784.0,
"loss": 0.2344,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.796875,
"rewards/margins": 2.578125,
"rewards/rejected": -6.375,
"step": 3130
},
{
"epoch": 0.2362500940486043,
"grad_norm": 9.278296782454238,
"learning_rate": 4.7227720585056986e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.625,
"logps/chosen": -588.0,
"logps/rejected": -876.0,
"loss": 0.2251,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.1875,
"rewards/margins": 3.28125,
"rewards/rejected": -7.46875,
"step": 3140
},
{
"epoch": 0.237002482883154,
"grad_norm": 11.778051631603764,
"learning_rate": 4.7197590150083986e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.484375,
"logps/chosen": -524.0,
"logps/rejected": -848.0,
"loss": 0.2454,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -3.515625,
"rewards/margins": 3.328125,
"rewards/rejected": -6.84375,
"step": 3150
},
{
"epoch": 0.2377548717177037,
"grad_norm": 10.909061617481207,
"learning_rate": 4.716730658156354e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -544.0,
"logps/rejected": -832.0,
"loss": 0.2329,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -3.859375,
"rewards/margins": 2.9375,
"rewards/rejected": -6.8125,
"step": 3160
},
{
"epoch": 0.2385072605522534,
"grad_norm": 9.24366980179484,
"learning_rate": 4.7136870088411564e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -552.0,
"logps/rejected": -860.0,
"loss": 0.2298,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -3.921875,
"rewards/margins": 3.265625,
"rewards/rejected": -7.1875,
"step": 3170
},
{
"epoch": 0.2392596493868031,
"grad_norm": 10.949962350431266,
"learning_rate": 4.710628088059898e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.515625,
"logps/chosen": -540.0,
"logps/rejected": -828.0,
"loss": 0.2271,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.8125,
"rewards/margins": 2.84375,
"rewards/rejected": -6.65625,
"step": 3180
},
{
"epoch": 0.2400120382213528,
"grad_norm": 11.993941852243937,
"learning_rate": 4.707553916915022e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.5625,
"logps/chosen": -536.0,
"logps/rejected": -816.0,
"loss": 0.2549,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -3.875,
"rewards/margins": 2.890625,
"rewards/rejected": -6.78125,
"step": 3190
},
{
"epoch": 0.24076442705590248,
"grad_norm": 10.660695498977127,
"learning_rate": 4.704464516614178e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -552.0,
"logps/rejected": -864.0,
"loss": 0.2583,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.921875,
"rewards/margins": 3.125,
"rewards/rejected": -7.03125,
"step": 3200
},
{
"epoch": 0.24151681589045218,
"grad_norm": 8.141720298029627,
"learning_rate": 4.7013599084700787e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.5,
"logps/chosen": -568.0,
"logps/rejected": -816.0,
"loss": 0.2528,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.03125,
"rewards/margins": 2.703125,
"rewards/rejected": -6.75,
"step": 3210
},
{
"epoch": 0.24226920472500188,
"grad_norm": 12.508230218250487,
"learning_rate": 4.698240113900348e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.5,
"logps/chosen": -572.0,
"logps/rejected": -912.0,
"loss": 0.2322,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.21875,
"rewards/margins": 3.40625,
"rewards/rejected": -7.625,
"step": 3220
},
{
"epoch": 0.24302159355955158,
"grad_norm": 8.835584611496607,
"learning_rate": 4.6951051544273765e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -532.0,
"logps/rejected": -908.0,
"loss": 0.2373,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -3.6875,
"rewards/margins": 3.625,
"rewards/rejected": -7.3125,
"step": 3230
},
{
"epoch": 0.24377398239410128,
"grad_norm": 8.181209169447907,
"learning_rate": 4.6919550516781723e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.40625,
"logps/chosen": -532.0,
"logps/rejected": -844.0,
"loss": 0.2355,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -3.765625,
"rewards/margins": 3.15625,
"rewards/rejected": -6.90625,
"step": 3240
},
{
"epoch": 0.24452637122865098,
"grad_norm": 9.207106143916054,
"learning_rate": 4.688789827384213e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.546875,
"logps/chosen": -580.0,
"logps/rejected": -848.0,
"loss": 0.2627,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.25,
"rewards/margins": 2.703125,
"rewards/rejected": -6.9375,
"step": 3250
},
{
"epoch": 0.24527876006320065,
"grad_norm": 12.041595328496067,
"learning_rate": 4.6856095033812927e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.46875,
"logps/chosen": -520.0,
"logps/rejected": -816.0,
"loss": 0.2378,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -3.546875,
"rewards/margins": 3.046875,
"rewards/rejected": -6.59375,
"step": 3260
},
{
"epoch": 0.24603114889775035,
"grad_norm": 8.162952877767816,
"learning_rate": 4.682414101609374e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.578125,
"logps/chosen": -580.0,
"logps/rejected": -884.0,
"loss": 0.2397,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.25,
"rewards/margins": 3.015625,
"rewards/rejected": -7.28125,
"step": 3270
},
{
"epoch": 0.24678353773230005,
"grad_norm": 7.661645701725578,
"learning_rate": 4.679203644112436e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -592.0,
"logps/rejected": -892.0,
"loss": 0.2426,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.3125,
"rewards/margins": 2.984375,
"rewards/rejected": -7.3125,
"step": 3280
},
{
"epoch": 0.24753592656684975,
"grad_norm": 9.39559412819907,
"learning_rate": 4.6759781530383214e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.390625,
"logps/chosen": -576.0,
"logps/rejected": -868.0,
"loss": 0.2572,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.25,
"rewards/margins": 3.015625,
"rewards/rejected": -7.25,
"step": 3290
},
{
"epoch": 0.24828831540139945,
"grad_norm": 8.664281653841188,
"learning_rate": 4.672737650638586e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.609375,
"logps/chosen": -604.0,
"logps/rejected": -856.0,
"loss": 0.2428,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.375,
"rewards/margins": 2.828125,
"rewards/rejected": -7.1875,
"step": 3300
},
{
"epoch": 0.24904070423594915,
"grad_norm": 9.341469048911085,
"learning_rate": 4.669482159268341e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -588.0,
"logps/rejected": -912.0,
"loss": 0.2357,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.375,
"rewards/margins": 3.3125,
"rewards/rejected": -7.6875,
"step": 3310
},
{
"epoch": 0.24979309307049882,
"grad_norm": 8.73158517562442,
"learning_rate": 4.666211701386103e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.390625,
"logps/chosen": -612.0,
"logps/rejected": -872.0,
"loss": 0.2537,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.53125,
"rewards/margins": 2.859375,
"rewards/rejected": -7.40625,
"step": 3320
},
{
"epoch": 0.25054548190504855,
"grad_norm": 9.193942328899496,
"learning_rate": 4.662926299553638e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.53125,
"logps/chosen": -624.0,
"logps/rejected": -912.0,
"loss": 0.2391,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.03125,
"rewards/rejected": -7.625,
"step": 3330
},
{
"epoch": 0.25129787073959825,
"grad_norm": 9.383374978565644,
"learning_rate": 4.6596259764358037e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.46875,
"logps/chosen": -624.0,
"logps/rejected": -980.0,
"loss": 0.2206,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.59375,
"rewards/margins": 3.546875,
"rewards/rejected": -8.125,
"step": 3340
},
{
"epoch": 0.2520502595741479,
"grad_norm": 8.09815946238127,
"learning_rate": 4.6563107548003967e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.578125,
"logps/chosen": -600.0,
"logps/rejected": -936.0,
"loss": 0.2575,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.4375,
"rewards/margins": 3.4375,
"rewards/rejected": -7.875,
"step": 3350
},
{
"epoch": 0.2528026484086976,
"grad_norm": 9.28524948295606,
"learning_rate": 4.6529806575179895e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.46875,
"logps/chosen": -560.0,
"logps/rejected": -812.0,
"loss": 0.2532,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.09375,
"rewards/margins": 2.609375,
"rewards/rejected": -6.71875,
"step": 3360
},
{
"epoch": 0.2535550372432473,
"grad_norm": 8.372693128057215,
"learning_rate": 4.6496357075617816e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -624.0,
"logps/rejected": -920.0,
"loss": 0.2312,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.4375,
"rewards/margins": 3.125,
"rewards/rejected": -7.5625,
"step": 3370
},
{
"epoch": 0.254307426077797,
"grad_norm": 8.973796350635121,
"learning_rate": 4.646275928007431e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.59375,
"logps/chosen": -576.0,
"logps/rejected": -836.0,
"loss": 0.2511,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": -4.21875,
"rewards/margins": 2.671875,
"rewards/rejected": -6.875,
"step": 3380
},
{
"epoch": 0.2550598149123467,
"grad_norm": 15.043288873372585,
"learning_rate": 4.642901342032905e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.484375,
"logps/chosen": -580.0,
"logps/rejected": -868.0,
"loss": 0.2644,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -4.25,
"rewards/margins": 2.984375,
"rewards/rejected": -7.21875,
"step": 3390
},
{
"epoch": 0.2558122037468964,
"grad_norm": 8.60829166534939,
"learning_rate": 4.639511972918311e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.46875,
"logps/chosen": -624.0,
"logps/rejected": -932.0,
"loss": 0.2515,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.78125,
"rewards/margins": 3.109375,
"rewards/rejected": -7.875,
"step": 3400
},
{
"epoch": 0.2565645925814461,
"grad_norm": 9.286255016844617,
"learning_rate": 4.636107844045742e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.40625,
"logps/chosen": -612.0,
"logps/rejected": -916.0,
"loss": 0.2498,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.59375,
"rewards/margins": 2.953125,
"rewards/rejected": -7.53125,
"step": 3410
},
{
"epoch": 0.2573169814159958,
"grad_norm": 7.683783123714642,
"learning_rate": 4.632688978899114e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.484375,
"logps/chosen": -588.0,
"logps/rejected": -884.0,
"loss": 0.2387,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.25,
"rewards/margins": 2.9375,
"rewards/rejected": -7.1875,
"step": 3420
},
{
"epoch": 0.2580693702505455,
"grad_norm": 10.260751299912629,
"learning_rate": 4.629255401064004e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.6875,
"logps/chosen": -592.0,
"logps/rejected": -888.0,
"loss": 0.2323,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.34375,
"rewards/margins": 3.09375,
"rewards/rejected": -7.4375,
"step": 3430
},
{
"epoch": 0.2588217590850952,
"grad_norm": 7.726491373550179,
"learning_rate": 4.625807134227483e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.6875,
"logps/chosen": -528.0,
"logps/rejected": -804.0,
"loss": 0.2666,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -3.6875,
"rewards/margins": 2.90625,
"rewards/rejected": -6.59375,
"step": 3440
},
{
"epoch": 0.2595741479196449,
"grad_norm": 9.175160897527569,
"learning_rate": 4.622344202177961e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.46875,
"logps/chosen": -528.0,
"logps/rejected": -816.0,
"loss": 0.2426,
"rewards/accuracies": 0.9375,
"rewards/chosen": -3.59375,
"rewards/margins": 2.8125,
"rewards/rejected": -6.40625,
"step": 3450
},
{
"epoch": 0.2603265367541946,
"grad_norm": 10.620774123065072,
"learning_rate": 4.6188666288050163e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -588.0,
"logps/rejected": -888.0,
"loss": 0.2542,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.21875,
"rewards/margins": 3.15625,
"rewards/rejected": -7.375,
"step": 3460
},
{
"epoch": 0.26107892558874424,
"grad_norm": 8.48922670158365,
"learning_rate": 4.615374438099232e-07,
"logits/chosen": -2.46875,
"logits/rejected": -2.25,
"logps/chosen": -592.0,
"logps/rejected": -832.0,
"loss": 0.2499,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -4.34375,
"rewards/margins": 2.46875,
"rewards/rejected": -6.8125,
"step": 3470
},
{
"epoch": 0.26183131442329394,
"grad_norm": 9.380206108680339,
"learning_rate": 4.611867654152033e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.359375,
"logps/chosen": -620.0,
"logps/rejected": -896.0,
"loss": 0.2377,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.625,
"rewards/margins": 2.796875,
"rewards/rejected": -7.4375,
"step": 3480
},
{
"epoch": 0.26258370325784364,
"grad_norm": 8.630455909850124,
"learning_rate": 4.608346301155516e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -548.0,
"logps/rejected": -856.0,
"loss": 0.2446,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -3.734375,
"rewards/margins": 3.1875,
"rewards/rejected": -6.9375,
"step": 3490
},
{
"epoch": 0.26333609209239334,
"grad_norm": 8.010936942529533,
"learning_rate": 4.604810403402285e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.4375,
"logps/chosen": -596.0,
"logps/rejected": -852.0,
"loss": 0.2511,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.4375,
"rewards/margins": 2.765625,
"rewards/rejected": -7.1875,
"step": 3500
},
{
"epoch": 0.26408848092694304,
"grad_norm": 8.594226664704044,
"learning_rate": 4.601259985285284e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.5,
"logps/chosen": -640.0,
"logps/rejected": -916.0,
"loss": 0.2225,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -4.6875,
"rewards/margins": 2.984375,
"rewards/rejected": -7.65625,
"step": 3510
},
{
"epoch": 0.26484086976149274,
"grad_norm": 7.3686969444976,
"learning_rate": 4.5976950712976286e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.484375,
"logps/chosen": -608.0,
"logps/rejected": -916.0,
"loss": 0.2473,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.5,
"rewards/margins": 3.109375,
"rewards/rejected": -7.625,
"step": 3520
},
{
"epoch": 0.26559325859604244,
"grad_norm": 9.006304280429053,
"learning_rate": 4.5941156860324345e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.359375,
"logps/chosen": -628.0,
"logps/rejected": -952.0,
"loss": 0.2532,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.5,
"rewards/margins": 3.546875,
"rewards/rejected": -8.0,
"step": 3530
},
{
"epoch": 0.26634564743059214,
"grad_norm": 8.731170425293595,
"learning_rate": 4.590521854182651e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.484375,
"logps/chosen": -640.0,
"logps/rejected": -944.0,
"loss": 0.2361,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.5,
"rewards/margins": 3.3125,
"rewards/rejected": -7.8125,
"step": 3540
},
{
"epoch": 0.26709803626514184,
"grad_norm": 10.35512022707896,
"learning_rate": 4.5869136005408893e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.53125,
"logps/chosen": -652.0,
"logps/rejected": -944.0,
"loss": 0.255,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.5,
"rewards/margins": 3.359375,
"rewards/rejected": -7.875,
"step": 3550
},
{
"epoch": 0.26785042509969154,
"grad_norm": 8.593676807068588,
"learning_rate": 4.5832909499992514e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.53125,
"logps/chosen": -584.0,
"logps/rejected": -876.0,
"loss": 0.2537,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.28125,
"rewards/margins": 3.0,
"rewards/rejected": -7.3125,
"step": 3560
},
{
"epoch": 0.26860281393424124,
"grad_norm": 7.229452609142653,
"learning_rate": 4.579653927549159e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.578125,
"logps/chosen": -648.0,
"logps/rejected": -944.0,
"loss": 0.2514,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.875,
"rewards/margins": 3.171875,
"rewards/rejected": -8.0625,
"step": 3570
},
{
"epoch": 0.26935520276879094,
"grad_norm": 9.484733316900101,
"learning_rate": 4.57600255828118e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.578125,
"logps/chosen": -604.0,
"logps/rejected": -940.0,
"loss": 0.2488,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.46875,
"rewards/margins": 3.28125,
"rewards/rejected": -7.71875,
"step": 3580
},
{
"epoch": 0.2701075916033406,
"grad_norm": 7.616886332053109,
"learning_rate": 4.572336867384856e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.546875,
"logps/chosen": -604.0,
"logps/rejected": -872.0,
"loss": 0.2532,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.3125,
"rewards/margins": 2.90625,
"rewards/rejected": -7.21875,
"step": 3590
},
{
"epoch": 0.2708599804378903,
"grad_norm": 9.424387046076045,
"learning_rate": 4.5686568801485306e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -580.0,
"logps/rejected": -864.0,
"loss": 0.219,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.28125,
"rewards/margins": 2.96875,
"rewards/rejected": -7.25,
"step": 3600
},
{
"epoch": 0.27161236927244,
"grad_norm": 12.547475129287186,
"learning_rate": 4.5649626219591685e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.578125,
"logps/chosen": -620.0,
"logps/rejected": -932.0,
"loss": 0.2504,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.59375,
"rewards/margins": 3.1875,
"rewards/rejected": -7.78125,
"step": 3610
},
{
"epoch": 0.2723647581069897,
"grad_norm": 10.17739824717397,
"learning_rate": 4.5612541183021905e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -620.0,
"logps/rejected": -972.0,
"loss": 0.2157,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.6875,
"rewards/margins": 3.578125,
"rewards/rejected": -8.3125,
"step": 3620
},
{
"epoch": 0.2731171469415394,
"grad_norm": 12.055309990415822,
"learning_rate": 4.5575313947612875e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.328125,
"logps/chosen": -628.0,
"logps/rejected": -968.0,
"loss": 0.2546,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.75,
"rewards/margins": 3.390625,
"rewards/rejected": -8.125,
"step": 3630
},
{
"epoch": 0.2738695357760891,
"grad_norm": 9.381452162285955,
"learning_rate": 4.553794477018251e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.34375,
"logps/chosen": -648.0,
"logps/rejected": -940.0,
"loss": 0.2405,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.78125,
"rewards/margins": 3.0625,
"rewards/rejected": -7.84375,
"step": 3640
},
{
"epoch": 0.2746219246106388,
"grad_norm": 9.717179832719665,
"learning_rate": 4.550043390852791e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.4375,
"logps/chosen": -608.0,
"logps/rejected": -956.0,
"loss": 0.2198,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.5,
"rewards/margins": 3.46875,
"rewards/rejected": -7.96875,
"step": 3650
},
{
"epoch": 0.2753743134451885,
"grad_norm": 9.596874231784692,
"learning_rate": 4.546278162142364e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.65625,
"logps/chosen": -580.0,
"logps/rejected": -888.0,
"loss": 0.2173,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.21875,
"rewards/margins": 3.34375,
"rewards/rejected": -7.5625,
"step": 3660
},
{
"epoch": 0.2761267022797382,
"grad_norm": 7.719511977487069,
"learning_rate": 4.542498816861988e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.59375,
"logps/chosen": -596.0,
"logps/rejected": -900.0,
"loss": 0.2265,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.375,
"rewards/margins": 3.234375,
"rewards/rejected": -7.625,
"step": 3670
},
{
"epoch": 0.2768790911142879,
"grad_norm": 11.043176918667355,
"learning_rate": 4.538705381084067e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.46875,
"logps/chosen": -620.0,
"logps/rejected": -900.0,
"loss": 0.2592,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -4.6875,
"rewards/margins": 3.0,
"rewards/rejected": -7.6875,
"step": 3680
},
{
"epoch": 0.2776314799488376,
"grad_norm": 11.332378533299988,
"learning_rate": 4.5348978809782123e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.375,
"logps/chosen": -608.0,
"logps/rejected": -864.0,
"loss": 0.2477,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.40625,
"rewards/margins": 2.75,
"rewards/rejected": -7.15625,
"step": 3690
},
{
"epoch": 0.2783838687833873,
"grad_norm": 9.797098424450917,
"learning_rate": 4.531076342811059e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.46875,
"logps/chosen": -580.0,
"logps/rejected": -880.0,
"loss": 0.2161,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.0625,
"rewards/margins": 3.140625,
"rewards/rejected": -7.1875,
"step": 3700
},
{
"epoch": 0.2791362576179369,
"grad_norm": 12.677047667665043,
"learning_rate": 4.5272407929460846e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.53125,
"logps/chosen": -624.0,
"logps/rejected": -896.0,
"loss": 0.2478,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.46875,
"rewards/margins": 3.09375,
"rewards/rejected": -7.5625,
"step": 3710
},
{
"epoch": 0.2798886464524866,
"grad_norm": 7.7470550829514995,
"learning_rate": 4.5233912578434297e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.390625,
"logps/chosen": -664.0,
"logps/rejected": -976.0,
"loss": 0.2311,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 3.0625,
"rewards/rejected": -8.0,
"step": 3720
},
{
"epoch": 0.2806410352870363,
"grad_norm": 9.903990733335885,
"learning_rate": 4.5195277640597165e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.40625,
"logps/chosen": -612.0,
"logps/rejected": -908.0,
"loss": 0.2546,
"rewards/accuracies": 0.84375,
"rewards/chosen": -4.4375,
"rewards/margins": 3.046875,
"rewards/rejected": -7.46875,
"step": 3730
},
{
"epoch": 0.281393424121586,
"grad_norm": 14.923608246368913,
"learning_rate": 4.5156503382478583e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -608.0,
"logps/rejected": -876.0,
"loss": 0.2426,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.3125,
"rewards/margins": 3.0,
"rewards/rejected": -7.3125,
"step": 3740
},
{
"epoch": 0.2821458129561357,
"grad_norm": 10.290171086499644,
"learning_rate": 4.511759007156886e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5,
"logps/chosen": -644.0,
"logps/rejected": -940.0,
"loss": 0.2252,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 3.0625,
"rewards/rejected": -7.90625,
"step": 3750
},
{
"epoch": 0.2828982017906854,
"grad_norm": 9.12773732569798,
"learning_rate": 4.507853797631753e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.390625,
"logps/chosen": -628.0,
"logps/rejected": -908.0,
"loss": 0.2214,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.625,
"rewards/margins": 2.90625,
"rewards/rejected": -7.53125,
"step": 3760
},
{
"epoch": 0.2836505906252351,
"grad_norm": 8.54967309370874,
"learning_rate": 4.503934736613161e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.484375,
"logps/chosen": -632.0,
"logps/rejected": -936.0,
"loss": 0.2154,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.296875,
"rewards/rejected": -7.90625,
"step": 3770
},
{
"epoch": 0.2844029794597848,
"grad_norm": 8.658423249282029,
"learning_rate": 4.500001851137363e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.3125,
"logps/chosen": -612.0,
"logps/rejected": -952.0,
"loss": 0.2313,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.625,
"rewards/margins": 3.546875,
"rewards/rejected": -8.1875,
"step": 3780
},
{
"epoch": 0.2851553682943345,
"grad_norm": 9.32583793890318,
"learning_rate": 4.496055168335986e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.265625,
"logps/chosen": -600.0,
"logps/rejected": -908.0,
"loss": 0.2481,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.4375,
"rewards/margins": 2.875,
"rewards/rejected": -7.34375,
"step": 3790
},
{
"epoch": 0.2859077571288842,
"grad_norm": 12.386535228494425,
"learning_rate": 4.4920947154358384e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.34375,
"logps/chosen": -636.0,
"logps/rejected": -964.0,
"loss": 0.2265,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.34375,
"rewards/rejected": -7.9375,
"step": 3800
},
{
"epoch": 0.2866601459634339,
"grad_norm": 7.1901796846029296,
"learning_rate": 4.4881205197587236e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.3125,
"logps/chosen": -592.0,
"logps/rejected": -884.0,
"loss": 0.2304,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.21875,
"rewards/margins": 3.1875,
"rewards/rejected": -7.375,
"step": 3810
},
{
"epoch": 0.2874125347979836,
"grad_norm": 8.02806199551109,
"learning_rate": 4.484132608721252e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.390625,
"logps/chosen": -516.0,
"logps/rejected": -844.0,
"loss": 0.2235,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.71875,
"rewards/margins": 3.1875,
"rewards/rejected": -6.90625,
"step": 3820
},
{
"epoch": 0.2881649236325333,
"grad_norm": 10.603106664851868,
"learning_rate": 4.480131009834651e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.390625,
"logps/chosen": -564.0,
"logps/rejected": -888.0,
"loss": 0.2457,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.09375,
"rewards/margins": 3.3125,
"rewards/rejected": -7.40625,
"step": 3830
},
{
"epoch": 0.28891731246708297,
"grad_norm": 6.45000355010823,
"learning_rate": 4.476115750704578e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.46875,
"logps/chosen": -564.0,
"logps/rejected": -860.0,
"loss": 0.2465,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -3.953125,
"rewards/margins": 3.234375,
"rewards/rejected": -7.1875,
"step": 3840
},
{
"epoch": 0.28966970130163266,
"grad_norm": 8.929005375270732,
"learning_rate": 4.472086859030926e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.546875,
"logps/chosen": -544.0,
"logps/rejected": -876.0,
"loss": 0.2149,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.0,
"rewards/margins": 3.203125,
"rewards/rejected": -7.21875,
"step": 3850
},
{
"epoch": 0.29042209013618236,
"grad_norm": 8.024874951607805,
"learning_rate": 4.468044362607633e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.40625,
"logps/chosen": -588.0,
"logps/rejected": -944.0,
"loss": 0.2153,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.21875,
"rewards/margins": 3.484375,
"rewards/rejected": -7.71875,
"step": 3860
},
{
"epoch": 0.29117447897073206,
"grad_norm": 9.5901959856907,
"learning_rate": 4.4639882893224924e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.375,
"logps/chosen": -608.0,
"logps/rejected": -900.0,
"loss": 0.2361,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.375,
"rewards/margins": 3.1875,
"rewards/rejected": -7.5625,
"step": 3870
},
{
"epoch": 0.29192686780528176,
"grad_norm": 9.101424498045278,
"learning_rate": 4.4599186671569623e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.359375,
"logps/chosen": -620.0,
"logps/rejected": -888.0,
"loss": 0.2144,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 2.765625,
"rewards/rejected": -7.3125,
"step": 3880
},
{
"epoch": 0.29267925663983146,
"grad_norm": 8.652186114303964,
"learning_rate": 4.4558355241859655e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -580.0,
"logps/rejected": -904.0,
"loss": 0.2362,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.375,
"rewards/margins": 3.296875,
"rewards/rejected": -7.65625,
"step": 3890
},
{
"epoch": 0.29343164547438116,
"grad_norm": 11.241719704363485,
"learning_rate": 4.451738888577705e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.578125,
"logps/chosen": -596.0,
"logps/rejected": -892.0,
"loss": 0.2483,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.34375,
"rewards/margins": 3.046875,
"rewards/rejected": -7.40625,
"step": 3900
},
{
"epoch": 0.29418403430893086,
"grad_norm": 7.151893594323524,
"learning_rate": 4.44762878859346e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.625,
"logps/chosen": -636.0,
"logps/rejected": -860.0,
"loss": 0.2198,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -4.625,
"rewards/margins": 2.484375,
"rewards/rejected": -7.125,
"step": 3910
},
{
"epoch": 0.29493642314348056,
"grad_norm": 6.694485106549714,
"learning_rate": 4.443505252587399e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.71875,
"logps/chosen": -676.0,
"logps/rejected": -1024.0,
"loss": 0.2236,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.125,
"rewards/margins": 3.71875,
"rewards/rejected": -8.8125,
"step": 3920
},
{
"epoch": 0.29568881197803026,
"grad_norm": 9.938005153509135,
"learning_rate": 4.43936830900638e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.53125,
"logps/chosen": -564.0,
"logps/rejected": -936.0,
"loss": 0.2042,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.3125,
"rewards/margins": 3.609375,
"rewards/rejected": -7.90625,
"step": 3930
},
{
"epoch": 0.29644120081257996,
"grad_norm": 16.682092053114292,
"learning_rate": 4.4352179863897556e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.65625,
"logps/chosen": -672.0,
"logps/rejected": -972.0,
"loss": 0.241,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.125,
"rewards/margins": 3.171875,
"rewards/rejected": -8.3125,
"step": 3940
},
{
"epoch": 0.29719358964712966,
"grad_norm": 7.512938135048488,
"learning_rate": 4.4310543133691746e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.484375,
"logps/chosen": -684.0,
"logps/rejected": -992.0,
"loss": 0.2316,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 3.3125,
"rewards/rejected": -8.3125,
"step": 3950
},
{
"epoch": 0.2979459784816793,
"grad_norm": 12.509486795634867,
"learning_rate": 4.426877318668387e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -584.0,
"logps/rejected": -904.0,
"loss": 0.2155,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.34375,
"rewards/margins": 3.265625,
"rewards/rejected": -7.59375,
"step": 3960
},
{
"epoch": 0.298698367316229,
"grad_norm": 7.987377416432433,
"learning_rate": 4.422687031103045e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.390625,
"logps/chosen": -636.0,
"logps/rejected": -944.0,
"loss": 0.262,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -4.5625,
"rewards/margins": 3.4375,
"rewards/rejected": -8.0,
"step": 3970
},
{
"epoch": 0.2994507561507787,
"grad_norm": 8.497924735104682,
"learning_rate": 4.4184834795805016e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.34375,
"logps/chosen": -552.0,
"logps/rejected": -892.0,
"loss": 0.2233,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.0625,
"rewards/margins": 3.421875,
"rewards/rejected": -7.5,
"step": 3980
},
{
"epoch": 0.3002031449853284,
"grad_norm": 12.602114991629081,
"learning_rate": 4.414266693099615e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -624.0,
"logps/rejected": -916.0,
"loss": 0.2254,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.40625,
"rewards/margins": 3.171875,
"rewards/rejected": -7.59375,
"step": 3990
},
{
"epoch": 0.3009555338198781,
"grad_norm": 8.14636861622685,
"learning_rate": 4.4100367007505455e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.71875,
"logps/chosen": -600.0,
"logps/rejected": -944.0,
"loss": 0.2228,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.4375,
"rewards/margins": 3.4375,
"rewards/rejected": -7.875,
"step": 4000
},
{
"epoch": 0.3017079226544278,
"grad_norm": 8.71715153360338,
"learning_rate": 4.405793531714558e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.59375,
"logps/chosen": -584.0,
"logps/rejected": -920.0,
"loss": 0.233,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.25,
"rewards/margins": 3.453125,
"rewards/rejected": -7.71875,
"step": 4010
},
{
"epoch": 0.3024603114889775,
"grad_norm": 7.675931534560127,
"learning_rate": 4.401537215263816e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.4375,
"logps/chosen": -672.0,
"logps/rejected": -960.0,
"loss": 0.2029,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.96875,
"rewards/margins": 3.078125,
"rewards/rejected": -8.0625,
"step": 4020
},
{
"epoch": 0.3032127003235272,
"grad_norm": 10.17405003846414,
"learning_rate": 4.3972677807611866e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.46875,
"logps/chosen": -624.0,
"logps/rejected": -996.0,
"loss": 0.2365,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.5625,
"rewards/margins": 3.59375,
"rewards/rejected": -8.1875,
"step": 4030
},
{
"epoch": 0.3039650891580769,
"grad_norm": 7.09818779533062,
"learning_rate": 4.39298525766003e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.46875,
"logps/chosen": -596.0,
"logps/rejected": -908.0,
"loss": 0.211,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.28125,
"rewards/margins": 3.359375,
"rewards/rejected": -7.65625,
"step": 4040
},
{
"epoch": 0.3047174779926266,
"grad_norm": 9.918785165797798,
"learning_rate": 4.388689675504002e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -656.0,
"logps/rejected": -972.0,
"loss": 0.2303,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.71875,
"rewards/margins": 3.328125,
"rewards/rejected": -8.0625,
"step": 4050
},
{
"epoch": 0.3054698668271763,
"grad_norm": 8.835187248857086,
"learning_rate": 4.3843810639268475e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.5,
"logps/chosen": -596.0,
"logps/rejected": -908.0,
"loss": 0.2353,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.25,
"rewards/margins": 3.421875,
"rewards/rejected": -7.65625,
"step": 4060
},
{
"epoch": 0.306222255661726,
"grad_norm": 8.367596510251957,
"learning_rate": 4.3800594526521983e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.40625,
"logps/chosen": -532.0,
"logps/rejected": -840.0,
"loss": 0.2074,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.859375,
"rewards/margins": 3.109375,
"rewards/rejected": -6.96875,
"step": 4070
},
{
"epoch": 0.30697464449627565,
"grad_norm": 9.414826271664367,
"learning_rate": 4.375724871493365e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.453125,
"logps/chosen": -556.0,
"logps/rejected": -860.0,
"loss": 0.2373,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -3.953125,
"rewards/margins": 3.21875,
"rewards/rejected": -7.15625,
"step": 4080
},
{
"epoch": 0.30772703333082535,
"grad_norm": 8.162028844684041,
"learning_rate": 4.371377350353134e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -536.0,
"logps/rejected": -848.0,
"loss": 0.229,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -3.828125,
"rewards/margins": 3.21875,
"rewards/rejected": -7.0625,
"step": 4090
},
{
"epoch": 0.30847942216537505,
"grad_norm": 7.801467771167628,
"learning_rate": 4.3670169192235586e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.546875,
"logps/chosen": -540.0,
"logps/rejected": -824.0,
"loss": 0.2394,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.859375,
"rewards/margins": 2.9375,
"rewards/rejected": -6.8125,
"step": 4100
},
{
"epoch": 0.30923181099992475,
"grad_norm": 10.277704427965585,
"learning_rate": 4.362643608185757e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.46875,
"logps/chosen": -628.0,
"logps/rejected": -932.0,
"loss": 0.2344,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.46875,
"rewards/margins": 3.140625,
"rewards/rejected": -7.625,
"step": 4110
},
{
"epoch": 0.30998419983447445,
"grad_norm": 13.33467825685964,
"learning_rate": 4.358257447409698e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.4375,
"logps/chosen": -588.0,
"logps/rejected": -928.0,
"loss": 0.2165,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.3125,
"rewards/margins": 3.40625,
"rewards/rejected": -7.71875,
"step": 4120
},
{
"epoch": 0.31073658866902415,
"grad_norm": 8.16160167838467,
"learning_rate": 4.353858467153998e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.515625,
"logps/chosen": -572.0,
"logps/rejected": -916.0,
"loss": 0.2297,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.15625,
"rewards/margins": 3.5625,
"rewards/rejected": -7.71875,
"step": 4130
},
{
"epoch": 0.31148897750357385,
"grad_norm": 7.879945152353835,
"learning_rate": 4.349446697765711e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.53125,
"logps/chosen": -588.0,
"logps/rejected": -924.0,
"loss": 0.227,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.25,
"rewards/margins": 3.5625,
"rewards/rejected": -7.8125,
"step": 4140
},
{
"epoch": 0.31224136633812355,
"grad_norm": 8.616338325133528,
"learning_rate": 4.345022169680117e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.375,
"logps/chosen": -564.0,
"logps/rejected": -900.0,
"loss": 0.2254,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.1875,
"rewards/margins": 3.375,
"rewards/rejected": -7.53125,
"step": 4150
},
{
"epoch": 0.31299375517267325,
"grad_norm": 9.359562218439928,
"learning_rate": 4.3405849134205164e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -608.0,
"logps/rejected": -904.0,
"loss": 0.2257,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.5,
"rewards/margins": 2.9375,
"rewards/rejected": -7.4375,
"step": 4160
},
{
"epoch": 0.31374614400722295,
"grad_norm": 8.969523285372976,
"learning_rate": 4.3361349595980136e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.375,
"logps/chosen": -696.0,
"logps/rejected": -1072.0,
"loss": 0.2182,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.25,
"rewards/margins": 3.78125,
"rewards/rejected": -9.0625,
"step": 4170
},
{
"epoch": 0.31449853284177265,
"grad_norm": 10.138740882360356,
"learning_rate": 4.331672338911313e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.28125,
"logps/chosen": -628.0,
"logps/rejected": -936.0,
"loss": 0.2201,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.6875,
"rewards/margins": 3.171875,
"rewards/rejected": -7.875,
"step": 4180
},
{
"epoch": 0.31525092167632235,
"grad_norm": 9.223656893908142,
"learning_rate": 4.3271970821465e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -988.0,
"loss": 0.2257,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0,
"rewards/margins": 3.1875,
"rewards/rejected": -8.1875,
"step": 4190
},
{
"epoch": 0.316003310510872,
"grad_norm": 11.12549439598949,
"learning_rate": 4.3227092201768346e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.484375,
"logps/chosen": -644.0,
"logps/rejected": -940.0,
"loss": 0.2265,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.90625,
"rewards/margins": 3.078125,
"rewards/rejected": -8.0,
"step": 4200
},
{
"epoch": 0.3167556993454217,
"grad_norm": 9.207865303317924,
"learning_rate": 4.3182087839625326e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.484375,
"logps/chosen": -584.0,
"logps/rejected": -896.0,
"loss": 0.2244,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.25,
"rewards/margins": 3.140625,
"rewards/rejected": -7.40625,
"step": 4210
},
{
"epoch": 0.3175080881799714,
"grad_norm": 8.426670829083685,
"learning_rate": 4.313695804550559e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -628.0,
"logps/rejected": -948.0,
"loss": 0.2203,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.6875,
"rewards/margins": 3.078125,
"rewards/rejected": -7.75,
"step": 4220
},
{
"epoch": 0.3182604770145211,
"grad_norm": 9.764940312491921,
"learning_rate": 4.309170313074408e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.40625,
"logps/chosen": -640.0,
"logps/rejected": -964.0,
"loss": 0.2608,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.046875,
"rewards/rejected": -7.875,
"step": 4230
},
{
"epoch": 0.3190128658490708,
"grad_norm": 7.952069517630966,
"learning_rate": 4.304632340753889e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.328125,
"logps/chosen": -536.0,
"logps/rejected": -848.0,
"loss": 0.2279,
"rewards/accuracies": 0.90625,
"rewards/chosen": -3.796875,
"rewards/margins": 3.1875,
"rewards/rejected": -7.0,
"step": 4240
},
{
"epoch": 0.3197652546836205,
"grad_norm": 8.83343001948812,
"learning_rate": 4.3000819188949145e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.390625,
"logps/chosen": -608.0,
"logps/rejected": -876.0,
"loss": 0.2347,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 2.84375,
"rewards/rejected": -7.40625,
"step": 4250
},
{
"epoch": 0.3205176435181702,
"grad_norm": 8.597957828748827,
"learning_rate": 4.2955190788892827e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.375,
"logps/chosen": -644.0,
"logps/rejected": -932.0,
"loss": 0.2496,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.03125,
"rewards/rejected": -7.84375,
"step": 4260
},
{
"epoch": 0.3212700323527199,
"grad_norm": 7.6696314669829775,
"learning_rate": 4.2909438522144566e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.34375,
"logps/chosen": -584.0,
"logps/rejected": -944.0,
"loss": 0.2319,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.28125,
"rewards/margins": 3.546875,
"rewards/rejected": -7.84375,
"step": 4270
},
{
"epoch": 0.3220224211872696,
"grad_norm": 10.007443872180758,
"learning_rate": 4.2863562704333545e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.34375,
"logps/chosen": -588.0,
"logps/rejected": -892.0,
"loss": 0.2052,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.25,
"rewards/margins": 3.15625,
"rewards/rejected": -7.40625,
"step": 4280
},
{
"epoch": 0.3227748100218193,
"grad_norm": 11.118831248083465,
"learning_rate": 4.2817563651941263e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.40625,
"logps/chosen": -628.0,
"logps/rejected": -932.0,
"loss": 0.2277,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.71875,
"rewards/margins": 3.15625,
"rewards/rejected": -7.875,
"step": 4290
},
{
"epoch": 0.323527198856369,
"grad_norm": 10.345024663416677,
"learning_rate": 4.277144168229939e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.28125,
"logps/chosen": -628.0,
"logps/rejected": -940.0,
"loss": 0.2453,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.71875,
"rewards/margins": 3.0625,
"rewards/rejected": -7.78125,
"step": 4300
},
{
"epoch": 0.3242795876909187,
"grad_norm": 7.051363685111973,
"learning_rate": 4.272519711358753e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.375,
"logps/chosen": -584.0,
"logps/rejected": -896.0,
"loss": 0.2047,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.25,
"rewards/margins": 3.234375,
"rewards/rejected": -7.5,
"step": 4310
},
{
"epoch": 0.32503197652546834,
"grad_norm": 9.476688840579149,
"learning_rate": 4.2678830264831077e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -632.0,
"logps/rejected": -952.0,
"loss": 0.2251,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.28125,
"rewards/rejected": -7.875,
"step": 4320
},
{
"epoch": 0.32578436536001804,
"grad_norm": 7.95698757416428,
"learning_rate": 4.2632341455899e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -628.0,
"logps/rejected": -932.0,
"loss": 0.2135,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -4.78125,
"rewards/margins": 3.109375,
"rewards/rejected": -7.875,
"step": 4330
},
{
"epoch": 0.32653675419456774,
"grad_norm": 6.006840590992643,
"learning_rate": 4.25857310075016e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.5625,
"logps/chosen": -596.0,
"logps/rejected": -904.0,
"loss": 0.2293,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.34375,
"rewards/margins": 3.1875,
"rewards/rejected": -7.53125,
"step": 4340
},
{
"epoch": 0.32728914302911744,
"grad_norm": 10.148748285674634,
"learning_rate": 4.2538999241188374e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.359375,
"logps/chosen": -628.0,
"logps/rejected": -932.0,
"loss": 0.1906,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.71875,
"rewards/margins": 3.328125,
"rewards/rejected": -8.0,
"step": 4350
},
{
"epoch": 0.32804153186366714,
"grad_norm": 11.129120168982654,
"learning_rate": 4.2492146479345693e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.453125,
"logps/chosen": -684.0,
"logps/rejected": -1032.0,
"loss": 0.22,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.28125,
"rewards/margins": 3.5,
"rewards/rejected": -8.8125,
"step": 4360
},
{
"epoch": 0.32879392069821683,
"grad_norm": 9.910453441221708,
"learning_rate": 4.244517304519467e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.3125,
"logps/chosen": -644.0,
"logps/rejected": -1008.0,
"loss": 0.2234,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.96875,
"rewards/margins": 3.46875,
"rewards/rejected": -8.4375,
"step": 4370
},
{
"epoch": 0.32954630953276653,
"grad_norm": 11.052178846558816,
"learning_rate": 4.2398079262788893e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.46875,
"logps/chosen": -648.0,
"logps/rejected": -956.0,
"loss": 0.222,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.125,
"rewards/rejected": -8.0625,
"step": 4380
},
{
"epoch": 0.33029869836731623,
"grad_norm": 9.642152534654413,
"learning_rate": 4.2350865457012184e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.453125,
"logps/chosen": -648.0,
"logps/rejected": -980.0,
"loss": 0.2332,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.875,
"rewards/margins": 3.453125,
"rewards/rejected": -8.3125,
"step": 4390
},
{
"epoch": 0.33105108720186593,
"grad_norm": 9.60330425948754,
"learning_rate": 4.2303531953576366e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.59375,
"logps/chosen": -624.0,
"logps/rejected": -960.0,
"loss": 0.2225,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.59375,
"rewards/margins": 3.421875,
"rewards/rejected": -8.0,
"step": 4400
},
{
"epoch": 0.33180347603641563,
"grad_norm": 7.6600645560452785,
"learning_rate": 4.2256079079019015e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -564.0,
"logps/rejected": -880.0,
"loss": 0.2416,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.15625,
"rewards/margins": 3.171875,
"rewards/rejected": -7.34375,
"step": 4410
},
{
"epoch": 0.33255586487096533,
"grad_norm": 7.850742770509769,
"learning_rate": 4.220850716070121e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -620.0,
"logps/rejected": -912.0,
"loss": 0.2238,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 3.171875,
"rewards/rejected": -7.75,
"step": 4420
},
{
"epoch": 0.33330825370551503,
"grad_norm": 7.909302118744176,
"learning_rate": 4.2160816526805267e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.4375,
"logps/chosen": -664.0,
"logps/rejected": -1016.0,
"loss": 0.2501,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 3.515625,
"rewards/rejected": -8.625,
"step": 4430
},
{
"epoch": 0.3340606425400647,
"grad_norm": 8.94738188436282,
"learning_rate": 4.211300750633248e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.375,
"logps/chosen": -672.0,
"logps/rejected": -988.0,
"loss": 0.2314,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.125,
"rewards/margins": 3.1875,
"rewards/rejected": -8.3125,
"step": 4440
},
{
"epoch": 0.3348130313746144,
"grad_norm": 8.676073044947644,
"learning_rate": 4.2065080429100865e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.359375,
"logps/chosen": -628.0,
"logps/rejected": -924.0,
"loss": 0.2225,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.625,
"rewards/margins": 3.0,
"rewards/rejected": -7.625,
"step": 4450
},
{
"epoch": 0.3355654202091641,
"grad_norm": 9.806473078557882,
"learning_rate": 4.2017035625742846e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -632.0,
"logps/rejected": -964.0,
"loss": 0.2098,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.78125,
"rewards/margins": 3.265625,
"rewards/rejected": -8.0625,
"step": 4460
},
{
"epoch": 0.3363178090437138,
"grad_norm": 7.522444333584003,
"learning_rate": 4.196887342770302e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.4375,
"logps/chosen": -668.0,
"logps/rejected": -1012.0,
"loss": 0.2095,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 3.703125,
"rewards/rejected": -8.625,
"step": 4470
},
{
"epoch": 0.3370701978782635,
"grad_norm": 8.728501659564188,
"learning_rate": 4.1920594167235845e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.484375,
"logps/chosen": -648.0,
"logps/rejected": -960.0,
"loss": 0.234,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.78125,
"rewards/margins": 3.34375,
"rewards/rejected": -8.125,
"step": 4480
},
{
"epoch": 0.3378225867128132,
"grad_norm": 9.465075398652212,
"learning_rate": 4.187219817740336e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.390625,
"logps/chosen": -612.0,
"logps/rejected": -936.0,
"loss": 0.2285,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.46875,
"rewards/margins": 3.15625,
"rewards/rejected": -7.625,
"step": 4490
},
{
"epoch": 0.3385749755473629,
"grad_norm": 7.109319264397478,
"learning_rate": 4.182368579207285e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.3125,
"logps/chosen": -616.0,
"logps/rejected": -936.0,
"loss": 0.2165,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.4375,
"rewards/margins": 3.34375,
"rewards/rejected": -7.78125,
"step": 4500
},
{
"epoch": 0.3393273643819126,
"grad_norm": 10.507873292947329,
"learning_rate": 4.177505734591461e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -608.0,
"logps/rejected": -932.0,
"loss": 0.2096,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.40625,
"rewards/margins": 3.34375,
"rewards/rejected": -7.75,
"step": 4510
},
{
"epoch": 0.3400797532164623,
"grad_norm": 7.3959641126703435,
"learning_rate": 4.172631317439956e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.421875,
"logps/chosen": -616.0,
"logps/rejected": -992.0,
"loss": 0.2317,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.46875,
"rewards/margins": 3.6875,
"rewards/rejected": -8.125,
"step": 4520
},
{
"epoch": 0.340832142051012,
"grad_norm": 8.87302820276429,
"learning_rate": 4.167745361379702e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.359375,
"logps/chosen": -600.0,
"logps/rejected": -928.0,
"loss": 0.2207,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.40625,
"rewards/margins": 3.40625,
"rewards/rejected": -7.8125,
"step": 4530
},
{
"epoch": 0.3415845308855617,
"grad_norm": 9.758571350415783,
"learning_rate": 4.162847900117229e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.203125,
"logps/chosen": -564.0,
"logps/rejected": -896.0,
"loss": 0.2177,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.03125,
"rewards/margins": 3.203125,
"rewards/rejected": -7.25,
"step": 4540
},
{
"epoch": 0.3423369197201114,
"grad_norm": 8.896439051201618,
"learning_rate": 4.1579389674384394e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.234375,
"logps/chosen": -632.0,
"logps/rejected": -908.0,
"loss": 0.2145,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.71875,
"rewards/margins": 2.96875,
"rewards/rejected": -7.6875,
"step": 4550
},
{
"epoch": 0.3430893085546611,
"grad_norm": 7.711661820543145,
"learning_rate": 4.153018597208374e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.3125,
"logps/chosen": -628.0,
"logps/rejected": -980.0,
"loss": 0.2203,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.75,
"rewards/margins": 3.578125,
"rewards/rejected": -8.3125,
"step": 4560
},
{
"epoch": 0.3438416973892107,
"grad_norm": 8.21240547755215,
"learning_rate": 4.1480868233709765e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.359375,
"logps/chosen": -628.0,
"logps/rejected": -984.0,
"loss": 0.219,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.4375,
"rewards/margins": 3.828125,
"rewards/rejected": -8.25,
"step": 4570
},
{
"epoch": 0.3445940862237604,
"grad_norm": 11.90114190092608,
"learning_rate": 4.1431436799488606e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.359375,
"logps/chosen": -644.0,
"logps/rejected": -948.0,
"loss": 0.2127,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 3.171875,
"rewards/rejected": -7.96875,
"step": 4580
},
{
"epoch": 0.3453464750583101,
"grad_norm": 6.8696936160082265,
"learning_rate": 4.1381892010430717e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.296875,
"logps/chosen": -656.0,
"logps/rejected": -960.0,
"loss": 0.2498,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.0,
"rewards/margins": 3.078125,
"rewards/rejected": -8.0625,
"step": 4590
},
{
"epoch": 0.3460988638928598,
"grad_norm": 10.171026913882173,
"learning_rate": 4.13322342083286e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.296875,
"logps/chosen": -688.0,
"logps/rejected": -988.0,
"loss": 0.2332,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.9375,
"rewards/margins": 3.3125,
"rewards/rejected": -8.25,
"step": 4600
},
{
"epoch": 0.3468512527274095,
"grad_norm": 8.63453504033253,
"learning_rate": 4.1282463735754356e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.453125,
"logps/chosen": -628.0,
"logps/rejected": -932.0,
"loss": 0.2289,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.65625,
"rewards/margins": 3.046875,
"rewards/rejected": -7.71875,
"step": 4610
},
{
"epoch": 0.3476036415619592,
"grad_norm": 9.580783720183787,
"learning_rate": 4.123258093605739e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -648.0,
"logps/rejected": -972.0,
"loss": 0.2302,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.71875,
"rewards/margins": 3.46875,
"rewards/rejected": -8.1875,
"step": 4620
},
{
"epoch": 0.3483560303965089,
"grad_norm": 7.120063150875206,
"learning_rate": 4.118258615336199e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.4375,
"logps/chosen": -624.0,
"logps/rejected": -920.0,
"loss": 0.2366,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 3.015625,
"rewards/rejected": -7.5625,
"step": 4630
},
{
"epoch": 0.3491084192310586,
"grad_norm": 5.827687137113672,
"learning_rate": 4.1132479732564964e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.28125,
"logps/chosen": -620.0,
"logps/rejected": -968.0,
"loss": 0.2027,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.53125,
"rewards/margins": 3.34375,
"rewards/rejected": -7.875,
"step": 4640
},
{
"epoch": 0.3498608080656083,
"grad_norm": 10.67088810537178,
"learning_rate": 4.1082262019333315e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.3125,
"logps/chosen": -620.0,
"logps/rejected": -1012.0,
"loss": 0.2132,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.59375,
"rewards/margins": 3.78125,
"rewards/rejected": -8.375,
"step": 4650
},
{
"epoch": 0.350613196900158,
"grad_norm": 10.094322081691317,
"learning_rate": 4.103193336010179e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -684.0,
"logps/rejected": -988.0,
"loss": 0.215,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.03125,
"rewards/margins": 3.5,
"rewards/rejected": -8.5,
"step": 4660
},
{
"epoch": 0.3513655857347077,
"grad_norm": 8.374810404276547,
"learning_rate": 4.098149410207051e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.359375,
"logps/chosen": -648.0,
"logps/rejected": -1012.0,
"loss": 0.2365,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 3.515625,
"rewards/rejected": -8.5,
"step": 4670
},
{
"epoch": 0.3521179745692574,
"grad_norm": 8.392131172676132,
"learning_rate": 4.09309445932026e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.359375,
"logps/chosen": -672.0,
"logps/rejected": -992.0,
"loss": 0.226,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.21875,
"rewards/margins": 3.1875,
"rewards/rejected": -8.375,
"step": 4680
},
{
"epoch": 0.35287036340380706,
"grad_norm": 9.468666629268613,
"learning_rate": 4.088028518222173e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.28125,
"logps/chosen": -620.0,
"logps/rejected": -940.0,
"loss": 0.2211,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.40625,
"rewards/rejected": -8.0,
"step": 4690
},
{
"epoch": 0.35362275223835676,
"grad_norm": 9.71697535829911,
"learning_rate": 4.0829516218609785e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -664.0,
"logps/rejected": -1032.0,
"loss": 0.2035,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.875,
"rewards/margins": 3.859375,
"rewards/rejected": -8.75,
"step": 4700
},
{
"epoch": 0.35437514107290646,
"grad_norm": 8.282815293843191,
"learning_rate": 4.077863805260439e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.625,
"logps/chosen": -640.0,
"logps/rejected": -936.0,
"loss": 0.2176,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.5625,
"rewards/margins": 3.34375,
"rewards/rejected": -7.875,
"step": 4710
},
{
"epoch": 0.35512752990745616,
"grad_norm": 7.901474744472959,
"learning_rate": 4.0727651035196545e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.3125,
"logps/chosen": -600.0,
"logps/rejected": -1000.0,
"loss": 0.2287,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.40625,
"rewards/margins": 3.765625,
"rewards/rejected": -8.1875,
"step": 4720
},
{
"epoch": 0.35587991874200586,
"grad_norm": 10.371105232687958,
"learning_rate": 4.0676555518128157e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.359375,
"logps/chosen": -668.0,
"logps/rejected": -964.0,
"loss": 0.2551,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -5.0625,
"rewards/margins": 3.078125,
"rewards/rejected": -8.125,
"step": 4730
},
{
"epoch": 0.35663230757655556,
"grad_norm": 7.284751062931884,
"learning_rate": 4.062535185388964e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.375,
"logps/chosen": -592.0,
"logps/rejected": -884.0,
"loss": 0.2446,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.25,
"rewards/margins": 3.109375,
"rewards/rejected": -7.375,
"step": 4740
},
{
"epoch": 0.35738469641110526,
"grad_norm": 7.798212343706078,
"learning_rate": 4.0574040395717493e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.46875,
"logps/chosen": -588.0,
"logps/rejected": -916.0,
"loss": 0.199,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.15625,
"rewards/margins": 3.328125,
"rewards/rejected": -7.46875,
"step": 4750
},
{
"epoch": 0.35813708524565496,
"grad_norm": 7.738938259994802,
"learning_rate": 4.0522621497591813e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.421875,
"logps/chosen": -608.0,
"logps/rejected": -936.0,
"loss": 0.2369,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.5625,
"rewards/margins": 3.390625,
"rewards/rejected": -7.9375,
"step": 4760
},
{
"epoch": 0.35888947408020466,
"grad_norm": 8.30773676851323,
"learning_rate": 4.0471095514233933e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.34375,
"logps/chosen": -600.0,
"logps/rejected": -940.0,
"loss": 0.2093,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.4375,
"rewards/margins": 3.328125,
"rewards/rejected": -7.78125,
"step": 4770
},
{
"epoch": 0.35964186291475436,
"grad_norm": 8.691676803479753,
"learning_rate": 4.041946280110389e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.421875,
"logps/chosen": -576.0,
"logps/rejected": -896.0,
"loss": 0.2141,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.21875,
"rewards/margins": 3.3125,
"rewards/rejected": -7.53125,
"step": 4780
},
{
"epoch": 0.36039425174930406,
"grad_norm": 8.607549823897239,
"learning_rate": 4.0367723714398047e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.46875,
"logps/chosen": -656.0,
"logps/rejected": -1024.0,
"loss": 0.1978,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.84375,
"rewards/margins": 3.796875,
"rewards/rejected": -8.625,
"step": 4790
},
{
"epoch": 0.36114664058385376,
"grad_norm": 8.771643431392596,
"learning_rate": 4.031587861104657e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.515625,
"logps/chosen": -676.0,
"logps/rejected": -996.0,
"loss": 0.2288,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.15625,
"rewards/margins": 3.265625,
"rewards/rejected": -8.4375,
"step": 4800
},
{
"epoch": 0.3618990294184034,
"grad_norm": 7.347767311708758,
"learning_rate": 4.0263927848711026e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.359375,
"logps/chosen": -656.0,
"logps/rejected": -956.0,
"loss": 0.2333,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.875,
"rewards/margins": 3.234375,
"rewards/rejected": -8.125,
"step": 4810
},
{
"epoch": 0.3626514182529531,
"grad_norm": 8.949761500130862,
"learning_rate": 4.021187178578187e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.34375,
"logps/chosen": -644.0,
"logps/rejected": -972.0,
"loss": 0.2175,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.875,
"rewards/margins": 3.296875,
"rewards/rejected": -8.1875,
"step": 4820
},
{
"epoch": 0.3634038070875028,
"grad_norm": 10.64662225641983,
"learning_rate": 4.0159710781375994e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.40625,
"logps/chosen": -700.0,
"logps/rejected": -1004.0,
"loss": 0.2273,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.46875,
"rewards/margins": 3.046875,
"rewards/rejected": -8.5,
"step": 4830
},
{
"epoch": 0.3641561959220525,
"grad_norm": 10.534790254504463,
"learning_rate": 4.0107445195334244e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.203125,
"logps/chosen": -704.0,
"logps/rejected": -1024.0,
"loss": 0.2303,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.3125,
"rewards/margins": 3.4375,
"rewards/rejected": -8.75,
"step": 4840
},
{
"epoch": 0.3649085847566022,
"grad_norm": 7.871168672929195,
"learning_rate": 4.005507538821894e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.1875,
"logps/chosen": -692.0,
"logps/rejected": -1012.0,
"loss": 0.2123,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.0625,
"rewards/margins": 3.375,
"rewards/rejected": -8.4375,
"step": 4850
},
{
"epoch": 0.3656609735911519,
"grad_norm": 8.593711676069107,
"learning_rate": 4.0002601721311393e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.25,
"logps/chosen": -680.0,
"logps/rejected": -1012.0,
"loss": 0.1999,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.125,
"rewards/margins": 3.53125,
"rewards/rejected": -8.6875,
"step": 4860
},
{
"epoch": 0.3664133624257016,
"grad_norm": 8.635194837292705,
"learning_rate": 3.995002455660939e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.359375,
"logps/chosen": -680.0,
"logps/rejected": -1064.0,
"loss": 0.2006,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -5.15625,
"rewards/margins": 3.953125,
"rewards/rejected": -9.125,
"step": 4870
},
{
"epoch": 0.3671657512602513,
"grad_norm": 9.553334243561997,
"learning_rate": 3.989734425682473e-07,
"logits/chosen": -2.484375,
"logits/rejected": -2.140625,
"logps/chosen": -740.0,
"logps/rejected": -1064.0,
"loss": 0.2232,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.875,
"rewards/margins": 3.34375,
"rewards/rejected": -9.1875,
"step": 4880
},
{
"epoch": 0.367918140094801,
"grad_norm": 9.264445545978198,
"learning_rate": 3.984456118538071e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.359375,
"logps/chosen": -688.0,
"logps/rejected": -1064.0,
"loss": 0.2264,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 3.6875,
"rewards/rejected": -9.0,
"step": 4890
},
{
"epoch": 0.3686705289293507,
"grad_norm": 7.532750216761137,
"learning_rate": 3.9791675706409593e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.296875,
"logps/chosen": -600.0,
"logps/rejected": -992.0,
"loss": 0.2365,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.46875,
"rewards/margins": 3.921875,
"rewards/rejected": -8.375,
"step": 4900
},
{
"epoch": 0.3694229177639004,
"grad_norm": 6.92523330567057,
"learning_rate": 3.9738688184750125e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.34375,
"logps/chosen": -644.0,
"logps/rejected": -964.0,
"loss": 0.2077,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.65625,
"rewards/margins": 3.359375,
"rewards/rejected": -8.0,
"step": 4910
},
{
"epoch": 0.3701753065984501,
"grad_norm": 10.657909024859954,
"learning_rate": 3.968559898594502e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.28125,
"logps/chosen": -620.0,
"logps/rejected": -948.0,
"loss": 0.2223,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 3.4375,
"rewards/rejected": -8.1875,
"step": 4920
},
{
"epoch": 0.37092769543299975,
"grad_norm": 6.720765504539305,
"learning_rate": 3.9632408476238416e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.3125,
"logps/chosen": -648.0,
"logps/rejected": -940.0,
"loss": 0.2112,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.875,
"rewards/margins": 2.984375,
"rewards/rejected": -7.875,
"step": 4930
},
{
"epoch": 0.37168008426754945,
"grad_norm": 7.963706047021695,
"learning_rate": 3.957911702257337e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.40625,
"logps/chosen": -656.0,
"logps/rejected": -1016.0,
"loss": 0.2229,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.75,
"rewards/margins": 3.71875,
"rewards/rejected": -8.4375,
"step": 4940
},
{
"epoch": 0.37243247310209915,
"grad_norm": 9.314200881993933,
"learning_rate": 3.952572499258929e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.375,
"logps/chosen": -628.0,
"logps/rejected": -952.0,
"loss": 0.2239,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.65625,
"rewards/margins": 3.421875,
"rewards/rejected": -8.0625,
"step": 4950
},
{
"epoch": 0.37318486193664885,
"grad_norm": 8.115661280449446,
"learning_rate": 3.9472232754619474e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.421875,
"logps/chosen": -632.0,
"logps/rejected": -960.0,
"loss": 0.216,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.53125,
"rewards/rejected": -8.125,
"step": 4960
},
{
"epoch": 0.37393725077119855,
"grad_norm": 7.013722897295247,
"learning_rate": 3.9418640677688464e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.3125,
"logps/chosen": -680.0,
"logps/rejected": -960.0,
"loss": 0.2489,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.03125,
"rewards/margins": 3.015625,
"rewards/rejected": -8.0625,
"step": 4970
},
{
"epoch": 0.37468963960574825,
"grad_norm": 9.184692139331514,
"learning_rate": 3.936494913150961e-07,
"logits/chosen": -2.5,
"logits/rejected": -2.328125,
"logps/chosen": -648.0,
"logps/rejected": -912.0,
"loss": 0.198,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.75,
"rewards/margins": 2.875,
"rewards/rejected": -7.625,
"step": 4980
},
{
"epoch": 0.37544202844029795,
"grad_norm": 9.392069646399577,
"learning_rate": 3.931115848648242e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -648.0,
"logps/rejected": -996.0,
"loss": 0.2036,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.875,
"rewards/margins": 3.671875,
"rewards/rejected": -8.5625,
"step": 4990
},
{
"epoch": 0.37619441727484765,
"grad_norm": 9.776594081515965,
"learning_rate": 3.925726911369008e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.640625,
"logps/chosen": -672.0,
"logps/rejected": -1024.0,
"loss": 0.1966,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.0,
"rewards/margins": 3.734375,
"rewards/rejected": -8.75,
"step": 5000
},
{
"epoch": 0.37619441727484765,
"eval_logits/chosen": -2.84375,
"eval_logits/rejected": -2.5625,
"eval_logps/chosen": -688.0,
"eval_logps/rejected": -1032.0,
"eval_loss": 0.25248709321022034,
"eval_rewards/accuracies": 0.888202428817749,
"eval_rewards/chosen": -5.21875,
"eval_rewards/margins": 3.578125,
"eval_rewards/rejected": -8.8125,
"eval_runtime": 3417.1912,
"eval_samples_per_second": 27.652,
"eval_steps_per_second": 0.432,
"step": 5000
},
{
"epoch": 0.37694680610939735,
"grad_norm": 8.537262562794556,
"learning_rate": 3.9203281384896856e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.640625,
"logps/chosen": -640.0,
"logps/rejected": -960.0,
"loss": 0.2204,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 3.34375,
"rewards/rejected": -8.25,
"step": 5010
},
{
"epoch": 0.37769919494394705,
"grad_norm": 8.964447569309238,
"learning_rate": 3.9149195672545547e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1016.0,
"loss": 0.2281,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.15625,
"rewards/margins": 3.390625,
"rewards/rejected": -8.5625,
"step": 5020
},
{
"epoch": 0.37845158377849675,
"grad_norm": 7.045933538630638,
"learning_rate": 3.9095012349754897e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.484375,
"logps/chosen": -688.0,
"logps/rejected": -1020.0,
"loss": 0.2218,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.21875,
"rewards/margins": 3.5625,
"rewards/rejected": -8.75,
"step": 5030
},
{
"epoch": 0.37920397261304645,
"grad_norm": 7.612458090206155,
"learning_rate": 3.904073179031702e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.265625,
"logps/chosen": -660.0,
"logps/rejected": -980.0,
"loss": 0.2149,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.875,
"rewards/margins": 3.1875,
"rewards/rejected": -8.0625,
"step": 5040
},
{
"epoch": 0.3799563614475961,
"grad_norm": 8.655710429881823,
"learning_rate": 3.898635436869485e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.359375,
"logps/chosen": -684.0,
"logps/rejected": -1024.0,
"loss": 0.2207,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.28125,
"rewards/margins": 3.53125,
"rewards/rejected": -8.8125,
"step": 5050
},
{
"epoch": 0.3807087502821458,
"grad_norm": 10.581734628831592,
"learning_rate": 3.8931880460019537e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.328125,
"logps/chosen": -644.0,
"logps/rejected": -928.0,
"loss": 0.2345,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.90625,
"rewards/margins": 2.984375,
"rewards/rejected": -7.875,
"step": 5060
},
{
"epoch": 0.3814611391166955,
"grad_norm": 7.882854152082069,
"learning_rate": 3.887731044008787e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.53125,
"logps/chosen": -656.0,
"logps/rejected": -976.0,
"loss": 0.1997,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 3.171875,
"rewards/rejected": -8.125,
"step": 5070
},
{
"epoch": 0.3822135279512452,
"grad_norm": 9.906732797744418,
"learning_rate": 3.8822644685359655e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -672.0,
"logps/rejected": -1040.0,
"loss": 0.2419,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.0625,
"rewards/margins": 3.71875,
"rewards/rejected": -8.8125,
"step": 5080
},
{
"epoch": 0.3829659167857949,
"grad_norm": 8.193426118810464,
"learning_rate": 3.876788357295516e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.265625,
"logps/chosen": -604.0,
"logps/rejected": -924.0,
"loss": 0.2228,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.53125,
"rewards/margins": 3.203125,
"rewards/rejected": -7.71875,
"step": 5090
},
{
"epoch": 0.3837183056203446,
"grad_norm": 7.671834236636164,
"learning_rate": 3.871302748065246e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.171875,
"logps/chosen": -576.0,
"logps/rejected": -936.0,
"loss": 0.2184,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.34375,
"rewards/margins": 3.25,
"rewards/rejected": -7.625,
"step": 5100
},
{
"epoch": 0.3844706944548943,
"grad_norm": 8.643588666196429,
"learning_rate": 3.8658076786884917e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.375,
"logps/chosen": -648.0,
"logps/rejected": -1024.0,
"loss": 0.1989,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.90625,
"rewards/margins": 3.796875,
"rewards/rejected": -8.6875,
"step": 5110
},
{
"epoch": 0.385223083289444,
"grad_norm": 9.75416624908225,
"learning_rate": 3.860303187073848e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.375,
"logps/chosen": -652.0,
"logps/rejected": -1012.0,
"loss": 0.1922,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.90625,
"rewards/margins": 3.671875,
"rewards/rejected": -8.625,
"step": 5120
},
{
"epoch": 0.3859754721239937,
"grad_norm": 11.127358228566324,
"learning_rate": 3.854789311194909e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -632.0,
"logps/rejected": -968.0,
"loss": 0.2132,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.75,
"rewards/margins": 3.53125,
"rewards/rejected": -8.3125,
"step": 5130
},
{
"epoch": 0.3867278609585434,
"grad_norm": 9.280144402951457,
"learning_rate": 3.849266089090013e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.34375,
"logps/chosen": -612.0,
"logps/rejected": -996.0,
"loss": 0.2038,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.46875,
"rewards/margins": 3.890625,
"rewards/rejected": -8.375,
"step": 5140
},
{
"epoch": 0.3874802497930931,
"grad_norm": 11.614194573881234,
"learning_rate": 3.843733558861969e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.40625,
"logps/chosen": -620.0,
"logps/rejected": -980.0,
"loss": 0.2072,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.625,
"rewards/margins": 3.59375,
"rewards/rejected": -8.25,
"step": 5150
},
{
"epoch": 0.3882326386276428,
"grad_norm": 8.34285723391702,
"learning_rate": 3.8381917586778036e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.359375,
"logps/chosen": -624.0,
"logps/rejected": -952.0,
"loss": 0.2061,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.65625,
"rewards/margins": 3.40625,
"rewards/rejected": -8.0625,
"step": 5160
},
{
"epoch": 0.38898502746219243,
"grad_norm": 6.882057609993868,
"learning_rate": 3.8326407267684916e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.328125,
"logps/chosen": -640.0,
"logps/rejected": -996.0,
"loss": 0.2104,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.625,
"rewards/margins": 3.75,
"rewards/rejected": -8.375,
"step": 5170
},
{
"epoch": 0.38973741629674213,
"grad_norm": 10.713709287777075,
"learning_rate": 3.827080501428695e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.390625,
"logps/chosen": -640.0,
"logps/rejected": -984.0,
"loss": 0.2094,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.59375,
"rewards/rejected": -8.375,
"step": 5180
},
{
"epoch": 0.39048980513129183,
"grad_norm": 7.105350641020761,
"learning_rate": 3.821511121016498e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5,
"logps/chosen": -604.0,
"logps/rejected": -936.0,
"loss": 0.2157,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.5625,
"rewards/margins": 3.359375,
"rewards/rejected": -7.90625,
"step": 5190
},
{
"epoch": 0.39124219396584153,
"grad_norm": 8.837764335158008,
"learning_rate": 3.815932623953142e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -632.0,
"logps/rejected": -960.0,
"loss": 0.2075,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.625,
"rewards/margins": 3.4375,
"rewards/rejected": -8.0625,
"step": 5200
},
{
"epoch": 0.39199458280039123,
"grad_norm": 9.041698753486347,
"learning_rate": 3.810345048722762e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.484375,
"logps/chosen": -628.0,
"logps/rejected": -940.0,
"loss": 0.2086,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.28125,
"rewards/rejected": -7.84375,
"step": 5210
},
{
"epoch": 0.39274697163494093,
"grad_norm": 8.84149882552823,
"learning_rate": 3.804748433872119e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.390625,
"logps/chosen": -600.0,
"logps/rejected": -912.0,
"loss": 0.1973,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.375,
"rewards/margins": 3.234375,
"rewards/rejected": -7.625,
"step": 5220
},
{
"epoch": 0.39349936046949063,
"grad_norm": 10.67174010897917,
"learning_rate": 3.7991428180103375e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.359375,
"logps/chosen": -664.0,
"logps/rejected": -1008.0,
"loss": 0.2128,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 3.640625,
"rewards/rejected": -8.5,
"step": 5230
},
{
"epoch": 0.39425174930404033,
"grad_norm": 9.519966003861786,
"learning_rate": 3.7935282398086335e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.4375,
"logps/chosen": -652.0,
"logps/rejected": -968.0,
"loss": 0.2288,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -4.78125,
"rewards/margins": 3.453125,
"rewards/rejected": -8.25,
"step": 5240
},
{
"epoch": 0.39500413813859003,
"grad_norm": 7.417461935131493,
"learning_rate": 3.787904738000053e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.34375,
"logps/chosen": -620.0,
"logps/rejected": -956.0,
"loss": 0.2207,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.5,
"rewards/margins": 3.328125,
"rewards/rejected": -7.8125,
"step": 5250
},
{
"epoch": 0.39575652697313973,
"grad_norm": 12.351645854988238,
"learning_rate": 3.7822723513792036e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.234375,
"logps/chosen": -620.0,
"logps/rejected": -980.0,
"loss": 0.1946,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.5625,
"rewards/margins": 3.84375,
"rewards/rejected": -8.375,
"step": 5260
},
{
"epoch": 0.39650891580768943,
"grad_norm": 8.905945193049313,
"learning_rate": 3.776631118801985e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -644.0,
"logps/rejected": -992.0,
"loss": 0.1959,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.875,
"rewards/margins": 3.578125,
"rewards/rejected": -8.4375,
"step": 5270
},
{
"epoch": 0.39726130464223913,
"grad_norm": 10.875357145554204,
"learning_rate": 3.7709810791853224e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.421875,
"logps/chosen": -632.0,
"logps/rejected": -1032.0,
"loss": 0.2058,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.6875,
"rewards/margins": 4.0625,
"rewards/rejected": -8.75,
"step": 5280
},
{
"epoch": 0.3980136934767888,
"grad_norm": 10.155296544076947,
"learning_rate": 3.765322271506898e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.296875,
"logps/chosen": -628.0,
"logps/rejected": -968.0,
"loss": 0.1956,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.78125,
"rewards/margins": 3.53125,
"rewards/rejected": -8.3125,
"step": 5290
},
{
"epoch": 0.3987660823113385,
"grad_norm": 8.845714187658283,
"learning_rate": 3.7596547348048806e-07,
"logits/chosen": -2.421875,
"logits/rejected": -2.1875,
"logps/chosen": -692.0,
"logps/rejected": -1024.0,
"loss": 0.1965,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.46875,
"rewards/margins": 3.15625,
"rewards/rejected": -8.625,
"step": 5300
},
{
"epoch": 0.3995184711458882,
"grad_norm": 9.208429894417057,
"learning_rate": 3.753978508177661e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.265625,
"logps/chosen": -720.0,
"logps/rejected": -1072.0,
"loss": 0.19,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.53125,
"rewards/margins": 3.546875,
"rewards/rejected": -9.0625,
"step": 5310
},
{
"epoch": 0.4002708599804379,
"grad_norm": 9.199388800164709,
"learning_rate": 3.7482936307835737e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.453125,
"logps/chosen": -668.0,
"logps/rejected": -1016.0,
"loss": 0.19,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.0,
"rewards/margins": 3.640625,
"rewards/rejected": -8.625,
"step": 5320
},
{
"epoch": 0.4010232488149876,
"grad_norm": 11.99410111092291,
"learning_rate": 3.7426001418406356e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.359375,
"logps/chosen": -660.0,
"logps/rejected": -1064.0,
"loss": 0.2066,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 4.125,
"rewards/rejected": -9.0625,
"step": 5330
},
{
"epoch": 0.4017756376495373,
"grad_norm": 8.94621660275642,
"learning_rate": 3.7368980806262707e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.390625,
"logps/chosen": -664.0,
"logps/rejected": -1000.0,
"loss": 0.1974,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.1875,
"rewards/margins": 3.40625,
"rewards/rejected": -8.625,
"step": 5340
},
{
"epoch": 0.402528026484087,
"grad_norm": 9.30671459047822,
"learning_rate": 3.73118748647704e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -716.0,
"logps/rejected": -1064.0,
"loss": 0.2078,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.53125,
"rewards/margins": 3.59375,
"rewards/rejected": -9.125,
"step": 5350
},
{
"epoch": 0.4032804153186367,
"grad_norm": 8.596185905238903,
"learning_rate": 3.725468398788372e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.453125,
"logps/chosen": -716.0,
"logps/rejected": -1040.0,
"loss": 0.2174,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.53125,
"rewards/margins": 3.453125,
"rewards/rejected": -9.0,
"step": 5360
},
{
"epoch": 0.4040328041531864,
"grad_norm": 9.304031175627815,
"learning_rate": 3.719740857014284e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.46875,
"logps/chosen": -724.0,
"logps/rejected": -1072.0,
"loss": 0.2106,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.5625,
"rewards/margins": 3.625,
"rewards/rejected": -9.1875,
"step": 5370
},
{
"epoch": 0.4047851929877361,
"grad_norm": 9.684060809443176,
"learning_rate": 3.714004900667122e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.609375,
"logps/chosen": -660.0,
"logps/rejected": -1020.0,
"loss": 0.214,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.0,
"rewards/margins": 3.8125,
"rewards/rejected": -8.8125,
"step": 5380
},
{
"epoch": 0.4055375818222858,
"grad_norm": 9.201676557442841,
"learning_rate": 3.708260569317277e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -672.0,
"logps/rejected": -996.0,
"loss": 0.2026,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.15625,
"rewards/margins": 3.34375,
"rewards/rejected": -8.5,
"step": 5390
},
{
"epoch": 0.4062899706568355,
"grad_norm": 7.524754017547156,
"learning_rate": 3.702507902592917e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.359375,
"logps/chosen": -656.0,
"logps/rejected": -1032.0,
"loss": 0.1825,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.03125,
"rewards/margins": 3.796875,
"rewards/rejected": -8.8125,
"step": 5400
},
{
"epoch": 0.4070423594913852,
"grad_norm": 8.356584739363438,
"learning_rate": 3.696746940179713e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.5,
"logps/chosen": -688.0,
"logps/rejected": -1032.0,
"loss": 0.2052,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.21875,
"rewards/margins": 3.453125,
"rewards/rejected": -8.6875,
"step": 5410
},
{
"epoch": 0.4077947483259348,
"grad_norm": 9.40662099033202,
"learning_rate": 3.690977721820565e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.390625,
"logps/chosen": -648.0,
"logps/rejected": -980.0,
"loss": 0.2068,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.6875,
"rewards/margins": 3.546875,
"rewards/rejected": -8.25,
"step": 5420
},
{
"epoch": 0.4085471371604845,
"grad_norm": 48.40020616188885,
"learning_rate": 3.6852002873153293e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.46875,
"logps/chosen": -720.0,
"logps/rejected": -1064.0,
"loss": 0.2153,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.6875,
"rewards/margins": 3.453125,
"rewards/rejected": -9.125,
"step": 5430
},
{
"epoch": 0.4092995259950342,
"grad_norm": 12.662814287211088,
"learning_rate": 3.679414676520541e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.546875,
"logps/chosen": -724.0,
"logps/rejected": -1104.0,
"loss": 0.2081,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.5625,
"rewards/margins": 3.8125,
"rewards/rejected": -9.375,
"step": 5440
},
{
"epoch": 0.4100519148295839,
"grad_norm": 8.362044095992177,
"learning_rate": 3.673620929349141e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.375,
"logps/chosen": -680.0,
"logps/rejected": -1080.0,
"loss": 0.1942,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.3125,
"rewards/margins": 4.0625,
"rewards/rejected": -9.375,
"step": 5450
},
{
"epoch": 0.4108043036641336,
"grad_norm": 7.768648809813202,
"learning_rate": 3.6678190857701996e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.59375,
"logps/chosen": -684.0,
"logps/rejected": -1008.0,
"loss": 0.2033,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 3.375,
"rewards/rejected": -8.5625,
"step": 5460
},
{
"epoch": 0.4115566924986833,
"grad_norm": 9.699525227338295,
"learning_rate": 3.6620091858086426e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.609375,
"logps/chosen": -696.0,
"logps/rejected": -1088.0,
"loss": 0.2035,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.3125,
"rewards/margins": 3.8125,
"rewards/rejected": -9.125,
"step": 5470
},
{
"epoch": 0.412309081333233,
"grad_norm": 8.507906559213271,
"learning_rate": 3.6561912695449747e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.578125,
"logps/chosen": -692.0,
"logps/rejected": -1016.0,
"loss": 0.2238,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 3.28125,
"rewards/rejected": -8.5,
"step": 5480
},
{
"epoch": 0.4130614701677827,
"grad_norm": 7.2410857606643715,
"learning_rate": 3.6503653771150003e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -696.0,
"logps/rejected": -1008.0,
"loss": 0.2034,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.25,
"rewards/margins": 3.21875,
"rewards/rejected": -8.5,
"step": 5490
},
{
"epoch": 0.4138138590023324,
"grad_norm": 10.453403141917368,
"learning_rate": 3.6445315487095487e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.625,
"logps/chosen": -732.0,
"logps/rejected": -1020.0,
"loss": 0.196,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.6875,
"rewards/margins": 3.140625,
"rewards/rejected": -8.8125,
"step": 5500
},
{
"epoch": 0.4145662478368821,
"grad_norm": 9.877038756567623,
"learning_rate": 3.6386898245741987e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.390625,
"logps/chosen": -636.0,
"logps/rejected": -1056.0,
"loss": 0.2096,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.8125,
"rewards/margins": 4.125,
"rewards/rejected": -8.9375,
"step": 5510
},
{
"epoch": 0.4153186366714318,
"grad_norm": 10.571725785138156,
"learning_rate": 3.632840245008998e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -612.0,
"logps/rejected": -932.0,
"loss": 0.2066,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.5,
"rewards/margins": 3.25,
"rewards/rejected": -7.75,
"step": 5520
},
{
"epoch": 0.4160710255059815,
"grad_norm": 9.723399391845161,
"learning_rate": 3.626982850368185e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.4375,
"logps/chosen": -664.0,
"logps/rejected": -1020.0,
"loss": 0.2157,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.125,
"rewards/margins": 3.515625,
"rewards/rejected": -8.6875,
"step": 5530
},
{
"epoch": 0.41682341434053116,
"grad_norm": 10.430224163906624,
"learning_rate": 3.621117681059914e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.421875,
"logps/chosen": -664.0,
"logps/rejected": -1056.0,
"loss": 0.1964,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.9375,
"rewards/margins": 3.953125,
"rewards/rejected": -8.875,
"step": 5540
},
{
"epoch": 0.41757580317508086,
"grad_norm": 8.436788642136804,
"learning_rate": 3.6152447775459715e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.421875,
"logps/chosen": -632.0,
"logps/rejected": -968.0,
"loss": 0.2039,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.84375,
"rewards/margins": 3.53125,
"rewards/rejected": -8.375,
"step": 5550
},
{
"epoch": 0.41832819200963056,
"grad_norm": 9.121013611607667,
"learning_rate": 3.609364180341503e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.609375,
"logps/chosen": -676.0,
"logps/rejected": -1064.0,
"loss": 0.2005,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 4.0,
"rewards/rejected": -9.125,
"step": 5560
},
{
"epoch": 0.41908058084418026,
"grad_norm": 8.201889530984163,
"learning_rate": 3.603475930014726e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.328125,
"logps/chosen": -728.0,
"logps/rejected": -1072.0,
"loss": 0.1941,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.5625,
"rewards/margins": 3.6875,
"rewards/rejected": -9.25,
"step": 5570
},
{
"epoch": 0.41983296967872996,
"grad_norm": 9.310384454726352,
"learning_rate": 3.5975800671866575e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -740.0,
"logps/rejected": -1088.0,
"loss": 0.2214,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.71875,
"rewards/margins": 3.59375,
"rewards/rejected": -9.3125,
"step": 5580
},
{
"epoch": 0.42058535851327966,
"grad_norm": 10.912009015261056,
"learning_rate": 3.591676632530829e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.375,
"logps/chosen": -676.0,
"logps/rejected": -1040.0,
"loss": 0.1968,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 3.828125,
"rewards/rejected": -9.0,
"step": 5590
},
{
"epoch": 0.42133774734782936,
"grad_norm": 10.375001718266384,
"learning_rate": 3.5857656667730074e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.328125,
"logps/chosen": -700.0,
"logps/rejected": -1104.0,
"loss": 0.185,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.28125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.4375,
"step": 5600
},
{
"epoch": 0.42209013618237906,
"grad_norm": 9.32593833031528,
"learning_rate": 3.579847210690915e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.390625,
"logps/chosen": -736.0,
"logps/rejected": -1048.0,
"loss": 0.2146,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.5625,
"rewards/margins": 3.4375,
"rewards/rejected": -9.0,
"step": 5610
},
{
"epoch": 0.42284252501692876,
"grad_norm": 10.072222773238604,
"learning_rate": 3.573921305113947e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.359375,
"logps/chosen": -644.0,
"logps/rejected": -996.0,
"loss": 0.1781,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.9375,
"rewards/margins": 3.5625,
"rewards/rejected": -8.5,
"step": 5620
},
{
"epoch": 0.42359491385147846,
"grad_norm": 8.272122929093275,
"learning_rate": 3.5679879909228895e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.375,
"logps/chosen": -668.0,
"logps/rejected": -1024.0,
"loss": 0.222,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 3.6875,
"rewards/rejected": -8.75,
"step": 5630
},
{
"epoch": 0.42434730268602816,
"grad_norm": 9.659485715922361,
"learning_rate": 3.562047309049638e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.46875,
"logps/chosen": -672.0,
"logps/rejected": -1024.0,
"loss": 0.2115,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.1875,
"rewards/margins": 3.625,
"rewards/rejected": -8.8125,
"step": 5640
},
{
"epoch": 0.42509969152057786,
"grad_norm": 6.267941272449572,
"learning_rate": 3.5560993004769156e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -632.0,
"logps/rejected": -976.0,
"loss": 0.2074,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.75,
"rewards/margins": 3.5,
"rewards/rejected": -8.25,
"step": 5650
},
{
"epoch": 0.4258520803551275,
"grad_norm": 27.41279541635157,
"learning_rate": 3.550144006237991e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.28125,
"logps/chosen": -660.0,
"logps/rejected": -980.0,
"loss": 0.205,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.9375,
"rewards/margins": 3.328125,
"rewards/rejected": -8.25,
"step": 5660
},
{
"epoch": 0.4266044691896772,
"grad_norm": 13.511045743963125,
"learning_rate": 3.5441814674163914e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -672.0,
"logps/rejected": -1072.0,
"loss": 0.191,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.90625,
"rewards/margins": 4.21875,
"rewards/rejected": -9.125,
"step": 5670
},
{
"epoch": 0.4273568580242269,
"grad_norm": 7.938326652173086,
"learning_rate": 3.5382117251456245e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.4375,
"logps/chosen": -680.0,
"logps/rejected": -1048.0,
"loss": 0.194,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.125,
"rewards/margins": 3.8125,
"rewards/rejected": -8.9375,
"step": 5680
},
{
"epoch": 0.4281092468587766,
"grad_norm": 9.761973101389064,
"learning_rate": 3.5322348206088903e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.328125,
"logps/chosen": -664.0,
"logps/rejected": -1088.0,
"loss": 0.1848,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.0625,
"rewards/margins": 4.03125,
"rewards/rejected": -9.125,
"step": 5690
},
{
"epoch": 0.4288616356933263,
"grad_norm": 7.12764182792613,
"learning_rate": 3.526250795038801e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.3125,
"logps/chosen": -708.0,
"logps/rejected": -1064.0,
"loss": 0.1999,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.5625,
"rewards/margins": 3.625,
"rewards/rejected": -9.1875,
"step": 5700
},
{
"epoch": 0.429614024527876,
"grad_norm": 9.484679834050828,
"learning_rate": 3.520259689717091e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.4375,
"logps/chosen": -748.0,
"logps/rejected": -1080.0,
"loss": 0.2158,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.9375,
"rewards/margins": 3.296875,
"rewards/rejected": -9.25,
"step": 5710
},
{
"epoch": 0.4303664133624257,
"grad_norm": 12.28777473146823,
"learning_rate": 3.514261545974338e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -688.0,
"logps/rejected": -1064.0,
"loss": 0.2056,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 4.125,
"rewards/rejected": -9.125,
"step": 5720
},
{
"epoch": 0.4311188021969754,
"grad_norm": 8.808299496135412,
"learning_rate": 3.508256405189675e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.53125,
"logps/chosen": -696.0,
"logps/rejected": -996.0,
"loss": 0.2138,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 3.390625,
"rewards/rejected": -8.5625,
"step": 5730
},
{
"epoch": 0.4318711910315251,
"grad_norm": 8.00492738993031,
"learning_rate": 3.502244308790506e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -668.0,
"logps/rejected": -1032.0,
"loss": 0.188,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 3.765625,
"rewards/rejected": -8.75,
"step": 5740
},
{
"epoch": 0.4326235798660748,
"grad_norm": 13.4475750058397,
"learning_rate": 3.496225298252216e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.453125,
"logps/chosen": -704.0,
"logps/rejected": -1032.0,
"loss": 0.2117,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.4375,
"rewards/margins": 3.390625,
"rewards/rejected": -8.8125,
"step": 5750
},
{
"epoch": 0.4333759687006245,
"grad_norm": 9.233762716910919,
"learning_rate": 3.490199415097892e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.390625,
"logps/chosen": -676.0,
"logps/rejected": -1072.0,
"loss": 0.2074,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.34375,
"rewards/margins": 3.9375,
"rewards/rejected": -9.3125,
"step": 5760
},
{
"epoch": 0.4341283575351742,
"grad_norm": 10.459722572084084,
"learning_rate": 3.4841667008980316e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -672.0,
"logps/rejected": -1048.0,
"loss": 0.198,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.21875,
"rewards/margins": 3.859375,
"rewards/rejected": -9.125,
"step": 5770
},
{
"epoch": 0.43488074636972385,
"grad_norm": 5.794376913507954,
"learning_rate": 3.478127197270257e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -680.0,
"logps/rejected": -1080.0,
"loss": 0.1787,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.28125,
"rewards/margins": 3.984375,
"rewards/rejected": -9.25,
"step": 5780
},
{
"epoch": 0.43563313520427355,
"grad_norm": 7.350470215517664,
"learning_rate": 3.4720809458790277e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.421875,
"logps/chosen": -708.0,
"logps/rejected": -1176.0,
"loss": 0.2117,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.40625,
"rewards/margins": 4.71875,
"rewards/rejected": -10.125,
"step": 5790
},
{
"epoch": 0.43638552403882325,
"grad_norm": 7.922786635736011,
"learning_rate": 3.466027988435356e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.359375,
"logps/chosen": -684.0,
"logps/rejected": -1072.0,
"loss": 0.1993,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.21875,
"rewards/margins": 3.828125,
"rewards/rejected": -9.0625,
"step": 5800
},
{
"epoch": 0.43713791287337295,
"grad_norm": 8.764459781816843,
"learning_rate": 3.459968366696515e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.46875,
"logps/chosen": -648.0,
"logps/rejected": -1032.0,
"loss": 0.1983,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.84375,
"rewards/margins": 3.96875,
"rewards/rejected": -8.8125,
"step": 5810
},
{
"epoch": 0.43789030170792265,
"grad_norm": 9.693612208902172,
"learning_rate": 3.453902122465753e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.5,
"logps/chosen": -652.0,
"logps/rejected": -1008.0,
"loss": 0.2172,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 3.6875,
"rewards/rejected": -8.6875,
"step": 5820
},
{
"epoch": 0.43864269054247235,
"grad_norm": 10.013718419535769,
"learning_rate": 3.4478292975920043e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.53125,
"logps/chosen": -652.0,
"logps/rejected": -992.0,
"loss": 0.2155,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.609375,
"rewards/rejected": -8.5625,
"step": 5830
},
{
"epoch": 0.43939507937702205,
"grad_norm": 9.831133197345066,
"learning_rate": 3.441749933969601e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.375,
"logps/chosen": -668.0,
"logps/rejected": -960.0,
"loss": 0.2001,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.96875,
"rewards/margins": 3.25,
"rewards/rejected": -8.1875,
"step": 5840
},
{
"epoch": 0.44014746821157175,
"grad_norm": 8.309385028382076,
"learning_rate": 3.435664073537985e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.5,
"logps/chosen": -704.0,
"logps/rejected": -1020.0,
"loss": 0.2087,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.375,
"rewards/margins": 3.359375,
"rewards/rejected": -8.75,
"step": 5850
},
{
"epoch": 0.44089985704612145,
"grad_norm": 9.600889681406729,
"learning_rate": 3.4295717582814143e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.3125,
"logps/chosen": -712.0,
"logps/rejected": -1020.0,
"loss": 0.1912,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.46875,
"rewards/margins": 3.203125,
"rewards/rejected": -8.6875,
"step": 5860
},
{
"epoch": 0.44165224588067115,
"grad_norm": 9.084849358967578,
"learning_rate": 3.4234730302286814e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.40625,
"logps/chosen": -656.0,
"logps/rejected": -980.0,
"loss": 0.1878,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.96875,
"rewards/margins": 3.453125,
"rewards/rejected": -8.4375,
"step": 5870
},
{
"epoch": 0.44240463471522085,
"grad_norm": 11.796957707433691,
"learning_rate": 3.417367931452813e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.328125,
"logps/chosen": -700.0,
"logps/rejected": -1096.0,
"loss": 0.1898,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.21875,
"rewards/margins": 3.953125,
"rewards/rejected": -9.1875,
"step": 5880
},
{
"epoch": 0.44315702354977055,
"grad_norm": 8.20973312387044,
"learning_rate": 3.4112565040707893e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.421875,
"logps/chosen": -668.0,
"logps/rejected": -1012.0,
"loss": 0.2096,
"rewards/accuracies": 0.856249988079071,
"rewards/chosen": -5.1875,
"rewards/margins": 3.46875,
"rewards/rejected": -8.6875,
"step": 5890
},
{
"epoch": 0.4439094123843202,
"grad_norm": 7.879754564921931,
"learning_rate": 3.405138790243248e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.3125,
"logps/chosen": -660.0,
"logps/rejected": -988.0,
"loss": 0.2132,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.90625,
"rewards/margins": 3.40625,
"rewards/rejected": -8.3125,
"step": 5900
},
{
"epoch": 0.4446618012188699,
"grad_norm": 10.339986830006676,
"learning_rate": 3.399014832174194e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.28125,
"logps/chosen": -652.0,
"logps/rejected": -984.0,
"loss": 0.2051,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.96875,
"rewards/margins": 3.421875,
"rewards/rejected": -8.375,
"step": 5910
},
{
"epoch": 0.4454141900534196,
"grad_norm": 9.630730722355262,
"learning_rate": 3.3928846721107104e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.46875,
"logps/chosen": -660.0,
"logps/rejected": -1024.0,
"loss": 0.1756,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.734375,
"rewards/rejected": -8.6875,
"step": 5920
},
{
"epoch": 0.4461665788879693,
"grad_norm": 9.37045022556023,
"learning_rate": 3.386748352342664e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.609375,
"logps/chosen": -724.0,
"logps/rejected": -1080.0,
"loss": 0.2018,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.625,
"rewards/margins": 3.765625,
"rewards/rejected": -9.375,
"step": 5930
},
{
"epoch": 0.446918967722519,
"grad_norm": 12.326318807630711,
"learning_rate": 3.380605915202419e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.46875,
"logps/chosen": -696.0,
"logps/rejected": -1048.0,
"loss": 0.2138,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.34375,
"rewards/margins": 3.703125,
"rewards/rejected": -9.0625,
"step": 5940
},
{
"epoch": 0.4476713565570687,
"grad_norm": 7.9992465107079935,
"learning_rate": 3.3744574030645366e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.484375,
"logps/chosen": -688.0,
"logps/rejected": -984.0,
"loss": 0.2101,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.28125,
"rewards/margins": 3.078125,
"rewards/rejected": -8.375,
"step": 5950
},
{
"epoch": 0.4484237453916184,
"grad_norm": 10.267488592156788,
"learning_rate": 3.3683028583454896e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.53125,
"logps/chosen": -664.0,
"logps/rejected": -1024.0,
"loss": 0.2049,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.96875,
"rewards/margins": 3.65625,
"rewards/rejected": -8.625,
"step": 5960
},
{
"epoch": 0.4491761342261681,
"grad_norm": 9.587336120051534,
"learning_rate": 3.3621423235033687e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5625,
"logps/chosen": -680.0,
"logps/rejected": -1040.0,
"loss": 0.2091,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.15625,
"rewards/margins": 3.734375,
"rewards/rejected": -8.875,
"step": 5970
},
{
"epoch": 0.4499285230607178,
"grad_norm": 14.237958212512485,
"learning_rate": 3.355975841037585e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.390625,
"logps/chosen": -664.0,
"logps/rejected": -1016.0,
"loss": 0.2047,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 3.46875,
"rewards/rejected": -8.625,
"step": 5980
},
{
"epoch": 0.4506809118952675,
"grad_norm": 7.320491760776586,
"learning_rate": 3.3498034534885844e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.421875,
"logps/chosen": -684.0,
"logps/rejected": -1024.0,
"loss": 0.1951,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.3125,
"rewards/margins": 3.5625,
"rewards/rejected": -8.875,
"step": 5990
},
{
"epoch": 0.4514333007298172,
"grad_norm": 7.2013732968721085,
"learning_rate": 3.343625203437547e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -676.0,
"logps/rejected": -1048.0,
"loss": 0.1931,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.25,
"rewards/margins": 3.84375,
"rewards/rejected": -9.125,
"step": 6000
},
{
"epoch": 0.4521856895643669,
"grad_norm": 7.647738073826342,
"learning_rate": 3.337441133506096e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -680.0,
"logps/rejected": -1040.0,
"loss": 0.2241,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.25,
"rewards/margins": 3.640625,
"rewards/rejected": -8.875,
"step": 6010
},
{
"epoch": 0.45293807839891653,
"grad_norm": 12.023931746185132,
"learning_rate": 3.3312512863560053e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -648.0,
"logps/rejected": -1008.0,
"loss": 0.202,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.0,
"rewards/margins": 3.703125,
"rewards/rejected": -8.6875,
"step": 6020
},
{
"epoch": 0.45369046723346623,
"grad_norm": 6.179188085390876,
"learning_rate": 3.325055704688906e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.5,
"logps/chosen": -656.0,
"logps/rejected": -1000.0,
"loss": 0.1926,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.0,
"rewards/margins": 3.640625,
"rewards/rejected": -8.625,
"step": 6030
},
{
"epoch": 0.45444285606801593,
"grad_norm": 12.492144030308038,
"learning_rate": 3.318854431245984e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -668.0,
"logps/rejected": -1040.0,
"loss": 0.2073,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.09375,
"rewards/margins": 3.875,
"rewards/rejected": -9.0,
"step": 6040
},
{
"epoch": 0.45519524490256563,
"grad_norm": 8.812424602807239,
"learning_rate": 3.312647508807696e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5625,
"logps/chosen": -700.0,
"logps/rejected": -1048.0,
"loss": 0.2168,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.46875,
"rewards/margins": 3.421875,
"rewards/rejected": -8.875,
"step": 6050
},
{
"epoch": 0.45594763373711533,
"grad_norm": 13.63172423692031,
"learning_rate": 3.3064349801934664e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.328125,
"logps/chosen": -680.0,
"logps/rejected": -1072.0,
"loss": 0.1776,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.125,
"rewards/margins": 4.0625,
"rewards/rejected": -9.1875,
"step": 6060
},
{
"epoch": 0.45670002257166503,
"grad_norm": 8.121800577128553,
"learning_rate": 3.300216888261396e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.453125,
"logps/chosen": -752.0,
"logps/rejected": -1128.0,
"loss": 0.1985,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.78125,
"rewards/margins": 3.875,
"rewards/rejected": -9.625,
"step": 6070
},
{
"epoch": 0.45745241140621473,
"grad_norm": 8.190803307707759,
"learning_rate": 3.2939932759079655e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.40625,
"logps/chosen": -716.0,
"logps/rejected": -1064.0,
"loss": 0.1866,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.46875,
"rewards/margins": 3.640625,
"rewards/rejected": -9.0625,
"step": 6080
},
{
"epoch": 0.45820480024076443,
"grad_norm": 8.516271503772927,
"learning_rate": 3.2877641860677357e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.40625,
"logps/chosen": -696.0,
"logps/rejected": -1096.0,
"loss": 0.1821,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.65625,
"rewards/margins": 3.796875,
"rewards/rejected": -9.4375,
"step": 6090
},
{
"epoch": 0.45895718907531413,
"grad_norm": 9.771485569804804,
"learning_rate": 3.2815296617130585e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.28125,
"logps/chosen": -676.0,
"logps/rejected": -1032.0,
"loss": 0.2016,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.1875,
"rewards/margins": 3.484375,
"rewards/rejected": -8.6875,
"step": 6100
},
{
"epoch": 0.45970957790986383,
"grad_norm": 9.188812610635678,
"learning_rate": 3.275289745853775e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.421875,
"logps/chosen": -672.0,
"logps/rejected": -1024.0,
"loss": 0.1841,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 3.6875,
"rewards/rejected": -8.625,
"step": 6110
},
{
"epoch": 0.46046196674441353,
"grad_norm": 8.341594160581772,
"learning_rate": 3.269044481536921e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.609375,
"logps/chosen": -688.0,
"logps/rejected": -1056.0,
"loss": 0.2123,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.28125,
"rewards/margins": 3.859375,
"rewards/rejected": -9.125,
"step": 6120
},
{
"epoch": 0.46121435557896323,
"grad_norm": 10.532580186377736,
"learning_rate": 3.2627939118464296e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.515625,
"logps/chosen": -648.0,
"logps/rejected": -1032.0,
"loss": 0.1958,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.90625,
"rewards/margins": 3.984375,
"rewards/rejected": -8.875,
"step": 6130
},
{
"epoch": 0.46196674441351293,
"grad_norm": 7.320156752221225,
"learning_rate": 3.256538079902833e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.46875,
"logps/chosen": -676.0,
"logps/rejected": -1088.0,
"loss": 0.2068,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.15625,
"rewards/margins": 4.21875,
"rewards/rejected": -9.375,
"step": 6140
},
{
"epoch": 0.4627191332480626,
"grad_norm": 12.347346631437645,
"learning_rate": 3.2502770288629655e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5,
"logps/chosen": -708.0,
"logps/rejected": -1032.0,
"loss": 0.2143,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.375,
"rewards/margins": 3.421875,
"rewards/rejected": -8.8125,
"step": 6150
},
{
"epoch": 0.4634715220826123,
"grad_norm": 7.790431152555406,
"learning_rate": 3.244010801919668e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.484375,
"logps/chosen": -708.0,
"logps/rejected": -1072.0,
"loss": 0.2069,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.46875,
"rewards/margins": 3.84375,
"rewards/rejected": -9.3125,
"step": 6160
},
{
"epoch": 0.464223910917162,
"grad_norm": 9.668021425592368,
"learning_rate": 3.237739442301487e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1088.0,
"loss": 0.1832,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.09375,
"rewards/margins": 4.34375,
"rewards/rejected": -9.4375,
"step": 6170
},
{
"epoch": 0.4649762997517117,
"grad_norm": 7.34718078248086,
"learning_rate": 3.231462993272377e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.546875,
"logps/chosen": -660.0,
"logps/rejected": -1024.0,
"loss": 0.1953,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.9375,
"rewards/margins": 3.703125,
"rewards/rejected": -8.625,
"step": 6180
},
{
"epoch": 0.4657286885862614,
"grad_norm": 6.688761060186978,
"learning_rate": 3.225181498131404e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.484375,
"logps/chosen": -616.0,
"logps/rejected": -1032.0,
"loss": 0.1804,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.59375,
"rewards/margins": 4.09375,
"rewards/rejected": -8.6875,
"step": 6190
},
{
"epoch": 0.4664810774208111,
"grad_norm": 9.071499105270645,
"learning_rate": 3.218895000212445e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -628.0,
"logps/rejected": -1020.0,
"loss": 0.2185,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.6875,
"rewards/margins": 3.84375,
"rewards/rejected": -8.5,
"step": 6200
},
{
"epoch": 0.4672334662553608,
"grad_norm": 7.959567822906466,
"learning_rate": 3.2126035428838874e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.65625,
"logps/chosen": -676.0,
"logps/rejected": -1012.0,
"loss": 0.1909,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.90625,
"rewards/margins": 3.515625,
"rewards/rejected": -8.4375,
"step": 6210
},
{
"epoch": 0.4679858550899105,
"grad_norm": 10.87084790524432,
"learning_rate": 3.2063071695483354e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.546875,
"logps/chosen": -676.0,
"logps/rejected": -1016.0,
"loss": 0.2136,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -5.21875,
"rewards/margins": 3.265625,
"rewards/rejected": -8.5,
"step": 6220
},
{
"epoch": 0.4687382439244602,
"grad_norm": 8.853900543247734,
"learning_rate": 3.200005923642305e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.421875,
"logps/chosen": -672.0,
"logps/rejected": -1032.0,
"loss": 0.2018,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.1875,
"rewards/margins": 3.515625,
"rewards/rejected": -8.75,
"step": 6230
},
{
"epoch": 0.4694906327590099,
"grad_norm": 7.092157700338752,
"learning_rate": 3.193699848635925e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.296875,
"logps/chosen": -704.0,
"logps/rejected": -1056.0,
"loss": 0.1805,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.40625,
"rewards/margins": 3.640625,
"rewards/rejected": -9.0625,
"step": 6240
},
{
"epoch": 0.4702430215935596,
"grad_norm": 6.085087527736119,
"learning_rate": 3.1873889880326425e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.359375,
"logps/chosen": -644.0,
"logps/rejected": -968.0,
"loss": 0.2104,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.875,
"rewards/margins": 3.359375,
"rewards/rejected": -8.25,
"step": 6250
},
{
"epoch": 0.4709954104281093,
"grad_norm": 8.454444908810444,
"learning_rate": 3.1810733853689153e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.328125,
"logps/chosen": -656.0,
"logps/rejected": -1012.0,
"loss": 0.1951,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.59375,
"rewards/margins": 3.65625,
"rewards/rejected": -8.25,
"step": 6260
},
{
"epoch": 0.4717477992626589,
"grad_norm": 11.017228443394977,
"learning_rate": 3.1747530842139157e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.546875,
"logps/chosen": -620.0,
"logps/rejected": -972.0,
"loss": 0.2342,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.59375,
"rewards/margins": 3.5625,
"rewards/rejected": -8.1875,
"step": 6270
},
{
"epoch": 0.4725001880972086,
"grad_norm": 7.775119852317317,
"learning_rate": 3.1684281281692305e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.53125,
"logps/chosen": -616.0,
"logps/rejected": -924.0,
"loss": 0.2249,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -4.625,
"rewards/margins": 3.125,
"rewards/rejected": -7.71875,
"step": 6280
},
{
"epoch": 0.4732525769317583,
"grad_norm": 10.2283707956898,
"learning_rate": 3.162098560868558e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.40625,
"logps/chosen": -608.0,
"logps/rejected": -936.0,
"loss": 0.1865,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.4375,
"rewards/margins": 3.484375,
"rewards/rejected": -7.90625,
"step": 6290
},
{
"epoch": 0.474004965766308,
"grad_norm": 9.541214997477915,
"learning_rate": 3.155764425977408e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.515625,
"logps/chosen": -664.0,
"logps/rejected": -996.0,
"loss": 0.1723,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.9375,
"rewards/margins": 3.703125,
"rewards/rejected": -8.625,
"step": 6300
},
{
"epoch": 0.4747573546008577,
"grad_norm": 8.776310012064307,
"learning_rate": 3.1494257671928006e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.546875,
"logps/chosen": -708.0,
"logps/rejected": -1040.0,
"loss": 0.2202,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.625,
"rewards/margins": 3.390625,
"rewards/rejected": -9.0,
"step": 6310
},
{
"epoch": 0.4755097434354074,
"grad_norm": 10.364964175951227,
"learning_rate": 3.1430826282429653e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -692.0,
"logps/rejected": -1032.0,
"loss": 0.1931,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.34375,
"rewards/margins": 3.609375,
"rewards/rejected": -8.9375,
"step": 6320
},
{
"epoch": 0.4762621322699571,
"grad_norm": 9.44277278928709,
"learning_rate": 3.136735052887038e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.484375,
"logps/chosen": -644.0,
"logps/rejected": -956.0,
"loss": 0.208,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.625,
"rewards/margins": 3.5,
"rewards/rejected": -8.125,
"step": 6330
},
{
"epoch": 0.4770145211045068,
"grad_norm": 9.715746849550793,
"learning_rate": 3.1303830849147594e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.390625,
"logps/chosen": -676.0,
"logps/rejected": -972.0,
"loss": 0.2013,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 3.1875,
"rewards/rejected": -8.3125,
"step": 6340
},
{
"epoch": 0.4777669099390565,
"grad_norm": 10.84726188854446,
"learning_rate": 3.124026768146176e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.359375,
"logps/chosen": -660.0,
"logps/rejected": -1016.0,
"loss": 0.2007,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.0,
"rewards/margins": 3.59375,
"rewards/rejected": -8.5625,
"step": 6350
},
{
"epoch": 0.4785192987736062,
"grad_norm": 9.066014191689984,
"learning_rate": 3.117666146431331e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.296875,
"logps/chosen": -700.0,
"logps/rejected": -1020.0,
"loss": 0.2123,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.375,
"rewards/margins": 3.25,
"rewards/rejected": -8.625,
"step": 6360
},
{
"epoch": 0.4792716876081559,
"grad_norm": 6.842520890359096,
"learning_rate": 3.111301263649969e-07,
"logits/chosen": -2.4375,
"logits/rejected": -2.140625,
"logps/chosen": -680.0,
"logps/rejected": -1012.0,
"loss": 0.2037,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.28125,
"rewards/margins": 3.34375,
"rewards/rejected": -8.625,
"step": 6370
},
{
"epoch": 0.4800240764427056,
"grad_norm": 12.209356701860676,
"learning_rate": 3.1049321637112303e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.328125,
"logps/chosen": -696.0,
"logps/rejected": -1056.0,
"loss": 0.2112,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.3125,
"rewards/margins": 3.671875,
"rewards/rejected": -9.0,
"step": 6380
},
{
"epoch": 0.48077646527725526,
"grad_norm": 9.512552820232575,
"learning_rate": 3.0985588905533456e-07,
"logits/chosen": -2.53125,
"logits/rejected": -2.234375,
"logps/chosen": -652.0,
"logps/rejected": -1012.0,
"loss": 0.1866,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.84375,
"rewards/margins": 3.84375,
"rewards/rejected": -8.6875,
"step": 6390
},
{
"epoch": 0.48152885411180496,
"grad_norm": 10.059637531130333,
"learning_rate": 3.0921814881433373e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.265625,
"logps/chosen": -668.0,
"logps/rejected": -1008.0,
"loss": 0.2111,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0625,
"rewards/margins": 3.390625,
"rewards/rejected": -8.4375,
"step": 6400
},
{
"epoch": 0.48228124294635466,
"grad_norm": 8.644085022609678,
"learning_rate": 3.0858000004767125e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.234375,
"logps/chosen": -704.0,
"logps/rejected": -1032.0,
"loss": 0.1879,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.4375,
"rewards/margins": 3.40625,
"rewards/rejected": -8.875,
"step": 6410
},
{
"epoch": 0.48303363178090436,
"grad_norm": 12.066282329982663,
"learning_rate": 3.079414471577163e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.4375,
"logps/chosen": -636.0,
"logps/rejected": -992.0,
"loss": 0.2003,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.78125,
"rewards/margins": 3.765625,
"rewards/rejected": -8.5625,
"step": 6420
},
{
"epoch": 0.48378602061545406,
"grad_norm": 7.856803020586083,
"learning_rate": 3.073024945496258e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.453125,
"logps/chosen": -640.0,
"logps/rejected": -1008.0,
"loss": 0.1913,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.625,
"rewards/margins": 3.859375,
"rewards/rejected": -8.5,
"step": 6430
},
{
"epoch": 0.48453840945000376,
"grad_norm": 11.09236658150085,
"learning_rate": 3.0666314663131435e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.390625,
"logps/chosen": -688.0,
"logps/rejected": -1040.0,
"loss": 0.1982,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.3125,
"rewards/margins": 3.6875,
"rewards/rejected": -9.0,
"step": 6440
},
{
"epoch": 0.48529079828455346,
"grad_norm": 9.143651153138784,
"learning_rate": 3.0602340781342343e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -684.0,
"logps/rejected": -996.0,
"loss": 0.1978,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.0625,
"rewards/margins": 3.453125,
"rewards/rejected": -8.5,
"step": 6450
},
{
"epoch": 0.48604318711910316,
"grad_norm": 7.950015653785829,
"learning_rate": 3.0538328250929147e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.453125,
"logps/chosen": -668.0,
"logps/rejected": -1096.0,
"loss": 0.2173,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.21875,
"rewards/margins": 4.34375,
"rewards/rejected": -9.5625,
"step": 6460
},
{
"epoch": 0.48679557595365286,
"grad_norm": 6.443015459033557,
"learning_rate": 3.047427751349228e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -676.0,
"logps/rejected": -980.0,
"loss": 0.1794,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.90625,
"rewards/margins": 3.40625,
"rewards/rejected": -8.3125,
"step": 6470
},
{
"epoch": 0.48754796478820256,
"grad_norm": 10.78015035404081,
"learning_rate": 3.041018901089579e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.3125,
"logps/chosen": -640.0,
"logps/rejected": -1004.0,
"loss": 0.1987,
"rewards/accuracies": 0.875,
"rewards/chosen": -4.9375,
"rewards/margins": 3.5,
"rewards/rejected": -8.4375,
"step": 6480
},
{
"epoch": 0.48830035362275226,
"grad_norm": 10.682871563209549,
"learning_rate": 3.034606318526423e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -628.0,
"logps/rejected": -980.0,
"loss": 0.1876,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.6875,
"rewards/margins": 3.671875,
"rewards/rejected": -8.375,
"step": 6490
},
{
"epoch": 0.48905274245730196,
"grad_norm": 8.745725840708767,
"learning_rate": 3.028190047897964e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.53125,
"logps/chosen": -628.0,
"logps/rejected": -984.0,
"loss": 0.2016,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 3.640625,
"rewards/rejected": -8.4375,
"step": 6500
},
{
"epoch": 0.4898051312918516,
"grad_norm": 10.212810861352294,
"learning_rate": 3.021770133467848e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.359375,
"logps/chosen": -648.0,
"logps/rejected": -1000.0,
"loss": 0.2123,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.8125,
"rewards/margins": 3.390625,
"rewards/rejected": -8.25,
"step": 6510
},
{
"epoch": 0.4905575201264013,
"grad_norm": 10.969586831458997,
"learning_rate": 3.015346619524859e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.375,
"logps/chosen": -660.0,
"logps/rejected": -1020.0,
"loss": 0.1798,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.765625,
"rewards/rejected": -8.6875,
"step": 6520
},
{
"epoch": 0.491309908960951,
"grad_norm": 10.171261226144825,
"learning_rate": 3.008919550382613e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.34375,
"logps/chosen": -640.0,
"logps/rejected": -1004.0,
"loss": 0.1922,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.71875,
"rewards/margins": 3.65625,
"rewards/rejected": -8.375,
"step": 6530
},
{
"epoch": 0.4920622977955007,
"grad_norm": 8.057760830640836,
"learning_rate": 3.0024889703792513e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.34375,
"logps/chosen": -652.0,
"logps/rejected": -996.0,
"loss": 0.1943,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 3.578125,
"rewards/rejected": -8.375,
"step": 6540
},
{
"epoch": 0.4928146866300504,
"grad_norm": 11.694363321978567,
"learning_rate": 2.996054923877136e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.328125,
"logps/chosen": -644.0,
"logps/rejected": -996.0,
"loss": 0.2085,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.0,
"rewards/margins": 3.546875,
"rewards/rejected": -8.5,
"step": 6550
},
{
"epoch": 0.4935670754646001,
"grad_norm": 7.79473303639143,
"learning_rate": 2.9896174552625427e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.4375,
"logps/chosen": -608.0,
"logps/rejected": -988.0,
"loss": 0.1914,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.34375,
"rewards/margins": 3.9375,
"rewards/rejected": -8.25,
"step": 6560
},
{
"epoch": 0.4943194642991498,
"grad_norm": 9.037599359618401,
"learning_rate": 2.983176608945356e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.34375,
"logps/chosen": -620.0,
"logps/rejected": -1020.0,
"loss": 0.209,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.65625,
"rewards/margins": 3.984375,
"rewards/rejected": -8.625,
"step": 6570
},
{
"epoch": 0.4950718531336995,
"grad_norm": 9.352384133988807,
"learning_rate": 2.9767324293587617e-07,
"logits/chosen": -2.546875,
"logits/rejected": -2.25,
"logps/chosen": -612.0,
"logps/rejected": -976.0,
"loss": 0.2026,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.40625,
"rewards/margins": 3.71875,
"rewards/rejected": -8.125,
"step": 6580
},
{
"epoch": 0.4958242419682492,
"grad_norm": 9.59161054448862,
"learning_rate": 2.9702849609589403e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.28125,
"logps/chosen": -588.0,
"logps/rejected": -952.0,
"loss": 0.2098,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.375,
"rewards/margins": 3.46875,
"rewards/rejected": -7.84375,
"step": 6590
},
{
"epoch": 0.4965766308027989,
"grad_norm": 8.118843911163207,
"learning_rate": 2.96383424822476e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.25,
"logps/chosen": -624.0,
"logps/rejected": -964.0,
"loss": 0.188,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 3.671875,
"rewards/rejected": -8.25,
"step": 6600
},
{
"epoch": 0.4973290196373486,
"grad_norm": 11.423896632062768,
"learning_rate": 2.957380335657473e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -624.0,
"logps/rejected": -1032.0,
"loss": 0.1888,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.6875,
"rewards/margins": 4.0625,
"rewards/rejected": -8.75,
"step": 6610
},
{
"epoch": 0.4980814084718983,
"grad_norm": 9.96296597496019,
"learning_rate": 2.950923267780405e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -668.0,
"logps/rejected": -1020.0,
"loss": 0.1912,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.125,
"rewards/margins": 3.578125,
"rewards/rejected": -8.6875,
"step": 6620
},
{
"epoch": 0.49883379730644795,
"grad_norm": 9.8686695233564,
"learning_rate": 2.944463089138648e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.375,
"logps/chosen": -620.0,
"logps/rejected": -976.0,
"loss": 0.2012,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.625,
"rewards/margins": 3.703125,
"rewards/rejected": -8.3125,
"step": 6630
},
{
"epoch": 0.49958618614099765,
"grad_norm": 6.179640034047122,
"learning_rate": 2.937999844298753e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.421875,
"logps/chosen": -640.0,
"logps/rejected": -1004.0,
"loss": 0.1938,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.9375,
"rewards/margins": 3.765625,
"rewards/rejected": -8.6875,
"step": 6640
},
{
"epoch": 0.5003385749755473,
"grad_norm": 7.675131257102138,
"learning_rate": 2.931533577848428e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -696.0,
"logps/rejected": -1056.0,
"loss": 0.1997,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.25,
"rewards/margins": 3.890625,
"rewards/rejected": -9.125,
"step": 6650
},
{
"epoch": 0.5010909638100971,
"grad_norm": 8.313739766862003,
"learning_rate": 2.9250643343962216e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -692.0,
"logps/rejected": -1072.0,
"loss": 0.2064,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.1875,
"rewards/margins": 3.9375,
"rewards/rejected": -9.125,
"step": 6660
},
{
"epoch": 0.5018433526446467,
"grad_norm": 11.372548847751178,
"learning_rate": 2.918592158571223e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -672.0,
"logps/rejected": -1032.0,
"loss": 0.1949,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.15625,
"rewards/margins": 3.59375,
"rewards/rejected": -8.75,
"step": 6670
},
{
"epoch": 0.5025957414791965,
"grad_norm": 11.266999098170334,
"learning_rate": 2.912117095022749e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.453125,
"logps/chosen": -676.0,
"logps/rejected": -1056.0,
"loss": 0.1877,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.03125,
"rewards/margins": 3.84375,
"rewards/rejected": -8.875,
"step": 6680
},
{
"epoch": 0.5033481303137461,
"grad_norm": 9.136922702076072,
"learning_rate": 2.9056391884200374e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.375,
"logps/chosen": -652.0,
"logps/rejected": -1012.0,
"loss": 0.1963,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.90625,
"rewards/margins": 3.734375,
"rewards/rejected": -8.625,
"step": 6690
},
{
"epoch": 0.5041005191482958,
"grad_norm": 9.283390523324467,
"learning_rate": 2.8991584834519405e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.46875,
"logps/chosen": -664.0,
"logps/rejected": -1000.0,
"loss": 0.2166,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.90625,
"rewards/margins": 3.40625,
"rewards/rejected": -8.3125,
"step": 6700
},
{
"epoch": 0.5048529079828455,
"grad_norm": 9.816366740181282,
"learning_rate": 2.8926750248266165e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.375,
"logps/chosen": -696.0,
"logps/rejected": -1040.0,
"loss": 0.1932,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.125,
"rewards/margins": 3.65625,
"rewards/rejected": -8.8125,
"step": 6710
},
{
"epoch": 0.5056052968173952,
"grad_norm": 22.193797590045083,
"learning_rate": 2.886188857271217e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.375,
"logps/chosen": -680.0,
"logps/rejected": -1056.0,
"loss": 0.1897,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.15625,
"rewards/margins": 3.984375,
"rewards/rejected": -9.125,
"step": 6720
},
{
"epoch": 0.5063576856519449,
"grad_norm": 10.433022889811856,
"learning_rate": 2.879700025531583e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.34375,
"logps/chosen": -644.0,
"logps/rejected": -980.0,
"loss": 0.1868,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 3.609375,
"rewards/rejected": -8.5625,
"step": 6730
},
{
"epoch": 0.5071100744864946,
"grad_norm": 9.7116225698569,
"learning_rate": 2.873208574371937e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -692.0,
"logps/rejected": -1032.0,
"loss": 0.2132,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.28125,
"rewards/margins": 3.515625,
"rewards/rejected": -8.8125,
"step": 6740
},
{
"epoch": 0.5078624633210443,
"grad_norm": 9.385981192467634,
"learning_rate": 2.8667145485745684e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.40625,
"logps/chosen": -660.0,
"logps/rejected": -1016.0,
"loss": 0.1993,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.75,
"rewards/margins": 3.8125,
"rewards/rejected": -8.5625,
"step": 6750
},
{
"epoch": 0.508614852155594,
"grad_norm": 8.52486755295249,
"learning_rate": 2.860217992939532e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.3125,
"logps/chosen": -624.0,
"logps/rejected": -980.0,
"loss": 0.1908,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.625,
"rewards/margins": 3.59375,
"rewards/rejected": -8.25,
"step": 6760
},
{
"epoch": 0.5093672409901437,
"grad_norm": 16.725319922468845,
"learning_rate": 2.853718952284331e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.53125,
"logps/chosen": -636.0,
"logps/rejected": -992.0,
"loss": 0.1945,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.625,
"rewards/margins": 3.84375,
"rewards/rejected": -8.4375,
"step": 6770
},
{
"epoch": 0.5101196298246934,
"grad_norm": 10.766654597956364,
"learning_rate": 2.8472174714436137e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.359375,
"logps/chosen": -632.0,
"logps/rejected": -1008.0,
"loss": 0.2055,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 3.859375,
"rewards/rejected": -8.625,
"step": 6780
},
{
"epoch": 0.5108720186592431,
"grad_norm": 10.372433884469864,
"learning_rate": 2.8407135952688634e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -648.0,
"logps/rejected": -968.0,
"loss": 0.1685,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.90625,
"rewards/margins": 3.34375,
"rewards/rejected": -8.25,
"step": 6790
},
{
"epoch": 0.5116244074937928,
"grad_norm": 7.884038617129319,
"learning_rate": 2.834207368628088e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.453125,
"logps/chosen": -648.0,
"logps/rejected": -1040.0,
"loss": 0.2004,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.90625,
"rewards/margins": 3.9375,
"rewards/rejected": -8.875,
"step": 6800
},
{
"epoch": 0.5123767963283424,
"grad_norm": 8.052127827588135,
"learning_rate": 2.8276988364055076e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5625,
"logps/chosen": -652.0,
"logps/rejected": -1112.0,
"loss": 0.1753,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.71875,
"rewards/margins": 4.84375,
"rewards/rejected": -9.5625,
"step": 6810
},
{
"epoch": 0.5131291851628922,
"grad_norm": 9.558785990781068,
"learning_rate": 2.821188043501251e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.671875,
"logps/chosen": -664.0,
"logps/rejected": -1032.0,
"loss": 0.1988,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.15625,
"rewards/margins": 3.78125,
"rewards/rejected": -8.9375,
"step": 6820
},
{
"epoch": 0.5138815739974418,
"grad_norm": 10.191515513058384,
"learning_rate": 2.8146750348310406e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.640625,
"logps/chosen": -652.0,
"logps/rejected": -1000.0,
"loss": 0.1967,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 3.65625,
"rewards/rejected": -8.4375,
"step": 6830
},
{
"epoch": 0.5146339628319916,
"grad_norm": 9.206459675501884,
"learning_rate": 2.8081598553258863e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.59375,
"logps/chosen": -696.0,
"logps/rejected": -1088.0,
"loss": 0.1693,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 4.0625,
"rewards/rejected": -9.25,
"step": 6840
},
{
"epoch": 0.5153863516665412,
"grad_norm": 8.438051844247585,
"learning_rate": 2.801642549931773e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.5625,
"logps/chosen": -696.0,
"logps/rejected": -1056.0,
"loss": 0.166,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.21875,
"rewards/margins": 3.765625,
"rewards/rejected": -9.0,
"step": 6850
},
{
"epoch": 0.516138740501091,
"grad_norm": 8.411109863368345,
"learning_rate": 2.7951231636093496e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.625,
"logps/chosen": -648.0,
"logps/rejected": -1040.0,
"loss": 0.1671,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.75,
"rewards/margins": 4.1875,
"rewards/rejected": -8.9375,
"step": 6860
},
{
"epoch": 0.5168911293356406,
"grad_norm": 9.113752670152984,
"learning_rate": 2.788601741333623e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.46875,
"logps/chosen": -652.0,
"logps/rejected": -1056.0,
"loss": 0.1667,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 4.15625,
"rewards/rejected": -9.0,
"step": 6870
},
{
"epoch": 0.5176435181701904,
"grad_norm": 9.555789599756581,
"learning_rate": 2.782078328093646e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.46875,
"logps/chosen": -620.0,
"logps/rejected": -1056.0,
"loss": 0.1981,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.65625,
"rewards/margins": 4.25,
"rewards/rejected": -8.875,
"step": 6880
},
{
"epoch": 0.51839590700474,
"grad_norm": 9.329766699762661,
"learning_rate": 2.775552968892205e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -616.0,
"logps/rejected": -972.0,
"loss": 0.1863,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.59375,
"rewards/margins": 3.703125,
"rewards/rejected": -8.3125,
"step": 6890
},
{
"epoch": 0.5191482958392898,
"grad_norm": 10.161837533103533,
"learning_rate": 2.76902570874551e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5,
"logps/chosen": -632.0,
"logps/rejected": -1024.0,
"loss": 0.1849,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.71875,
"rewards/margins": 4.03125,
"rewards/rejected": -8.75,
"step": 6900
},
{
"epoch": 0.5199006846738394,
"grad_norm": 10.983457686576111,
"learning_rate": 2.7624965926828856e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -604.0,
"logps/rejected": -984.0,
"loss": 0.2057,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.5625,
"rewards/margins": 3.828125,
"rewards/rejected": -8.375,
"step": 6910
},
{
"epoch": 0.5206530735083892,
"grad_norm": 9.70579239659648,
"learning_rate": 2.7559656657464615e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -652.0,
"logps/rejected": -968.0,
"loss": 0.2082,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.8125,
"rewards/margins": 3.3125,
"rewards/rejected": -8.125,
"step": 6920
},
{
"epoch": 0.5214054623429388,
"grad_norm": 11.541497496246196,
"learning_rate": 2.7494329729908585e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.4375,
"logps/chosen": -652.0,
"logps/rejected": -1072.0,
"loss": 0.1915,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.0625,
"rewards/margins": 4.03125,
"rewards/rejected": -9.125,
"step": 6930
},
{
"epoch": 0.5221578511774885,
"grad_norm": 7.92010618143839,
"learning_rate": 2.7428985594828785e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.65625,
"logps/chosen": -664.0,
"logps/rejected": -1064.0,
"loss": 0.2087,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.90625,
"rewards/margins": 4.03125,
"rewards/rejected": -8.9375,
"step": 6940
},
{
"epoch": 0.5229102400120382,
"grad_norm": 9.537578165888158,
"learning_rate": 2.736362470301195e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -660.0,
"logps/rejected": -1008.0,
"loss": 0.1796,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.15625,
"rewards/margins": 3.59375,
"rewards/rejected": -8.75,
"step": 6950
},
{
"epoch": 0.5236626288465879,
"grad_norm": 10.626977421177818,
"learning_rate": 2.7298247505360414e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.578125,
"logps/chosen": -672.0,
"logps/rejected": -1080.0,
"loss": 0.1903,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.0,
"rewards/margins": 4.21875,
"rewards/rejected": -9.25,
"step": 6960
},
{
"epoch": 0.5244150176811376,
"grad_norm": 8.431235259296844,
"learning_rate": 2.723285445288902e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.609375,
"logps/chosen": -640.0,
"logps/rejected": -1004.0,
"loss": 0.1986,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.765625,
"rewards/rejected": -8.5625,
"step": 6970
},
{
"epoch": 0.5251674065156873,
"grad_norm": 10.03229995476381,
"learning_rate": 2.7167445996721954e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.625,
"logps/chosen": -660.0,
"logps/rejected": -976.0,
"loss": 0.1903,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.265625,
"rewards/rejected": -8.1875,
"step": 6980
},
{
"epoch": 0.525919795350237,
"grad_norm": 7.476779653388887,
"learning_rate": 2.710202258808967e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5625,
"logps/chosen": -616.0,
"logps/rejected": -976.0,
"loss": 0.186,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.46875,
"rewards/margins": 3.765625,
"rewards/rejected": -8.25,
"step": 6990
},
{
"epoch": 0.5266721841847867,
"grad_norm": 9.682232421431149,
"learning_rate": 2.70365846783258e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.46875,
"logps/chosen": -660.0,
"logps/rejected": -1032.0,
"loss": 0.1915,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.0,
"rewards/margins": 3.703125,
"rewards/rejected": -8.6875,
"step": 7000
},
{
"epoch": 0.5274245730193364,
"grad_norm": 6.954785345590251,
"learning_rate": 2.6971132718864005e-07,
"logits/chosen": -2.609375,
"logits/rejected": -2.375,
"logps/chosen": -648.0,
"logps/rejected": -1012.0,
"loss": 0.1862,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.75,
"rewards/margins": 3.828125,
"rewards/rejected": -8.5625,
"step": 7010
},
{
"epoch": 0.5281769618538861,
"grad_norm": 10.205161787990173,
"learning_rate": 2.6905667161234844e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -640.0,
"logps/rejected": -996.0,
"loss": 0.2083,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.84375,
"rewards/margins": 3.609375,
"rewards/rejected": -8.4375,
"step": 7020
},
{
"epoch": 0.5289293506884358,
"grad_norm": 6.3527477235968215,
"learning_rate": 2.6840188457062725e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.515625,
"logps/chosen": -672.0,
"logps/rejected": -1024.0,
"loss": 0.1782,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.78125,
"rewards/margins": 3.96875,
"rewards/rejected": -8.75,
"step": 7030
},
{
"epoch": 0.5296817395229855,
"grad_norm": 12.108587765802628,
"learning_rate": 2.6774697058062713e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.421875,
"logps/chosen": -640.0,
"logps/rejected": -1024.0,
"loss": 0.1955,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.8125,
"rewards/margins": 3.96875,
"rewards/rejected": -8.8125,
"step": 7040
},
{
"epoch": 0.5304341283575352,
"grad_norm": 8.516026422713471,
"learning_rate": 2.6709193416037475e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.4375,
"logps/chosen": -600.0,
"logps/rejected": -992.0,
"loss": 0.1814,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.4375,
"rewards/margins": 3.90625,
"rewards/rejected": -8.3125,
"step": 7050
},
{
"epoch": 0.5311865171920849,
"grad_norm": 9.720629632343826,
"learning_rate": 2.6643677982874133e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.546875,
"logps/chosen": -588.0,
"logps/rejected": -952.0,
"loss": 0.2008,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.46875,
"rewards/margins": 3.75,
"rewards/rejected": -8.1875,
"step": 7060
},
{
"epoch": 0.5319389060266345,
"grad_norm": 9.509223667047454,
"learning_rate": 2.657815121054115e-07,
"logits/chosen": -3.03125,
"logits/rejected": -2.703125,
"logps/chosen": -616.0,
"logps/rejected": -984.0,
"loss": 0.1894,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -4.4375,
"rewards/margins": 3.734375,
"rewards/rejected": -8.1875,
"step": 7070
},
{
"epoch": 0.5326912948611843,
"grad_norm": 9.245136404559346,
"learning_rate": 2.6512613551085214e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.609375,
"logps/chosen": -624.0,
"logps/rejected": -960.0,
"loss": 0.1823,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.4375,
"rewards/margins": 3.640625,
"rewards/rejected": -8.0625,
"step": 7080
},
{
"epoch": 0.5334436836957339,
"grad_norm": 9.894052286966772,
"learning_rate": 2.6447065456628105e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.484375,
"logps/chosen": -624.0,
"logps/rejected": -984.0,
"loss": 0.1845,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -4.5625,
"rewards/margins": 3.75,
"rewards/rejected": -8.3125,
"step": 7090
},
{
"epoch": 0.5341960725302837,
"grad_norm": 9.553544752988854,
"learning_rate": 2.6381507379363627e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.53125,
"logps/chosen": -584.0,
"logps/rejected": -976.0,
"loss": 0.1861,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.09375,
"rewards/margins": 4.125,
"rewards/rejected": -8.25,
"step": 7100
},
{
"epoch": 0.5349484613648333,
"grad_norm": 10.193963283951748,
"learning_rate": 2.6315939771554407e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5,
"logps/chosen": -604.0,
"logps/rejected": -972.0,
"loss": 0.1899,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.375,
"rewards/margins": 3.84375,
"rewards/rejected": -8.25,
"step": 7110
},
{
"epoch": 0.5357008501993831,
"grad_norm": 9.637160856140365,
"learning_rate": 2.6250363085528867e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.375,
"logps/chosen": -632.0,
"logps/rejected": -988.0,
"loss": 0.2127,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.8125,
"rewards/margins": 3.578125,
"rewards/rejected": -8.375,
"step": 7120
},
{
"epoch": 0.5364532390339327,
"grad_norm": 8.700264918397732,
"learning_rate": 2.618477777367801e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.25,
"logps/chosen": -640.0,
"logps/rejected": -1012.0,
"loss": 0.2108,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.6875,
"rewards/margins": 3.921875,
"rewards/rejected": -8.625,
"step": 7130
},
{
"epoch": 0.5372056278684825,
"grad_norm": 8.134833894710736,
"learning_rate": 2.6119184288452377e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.28125,
"logps/chosen": -656.0,
"logps/rejected": -1032.0,
"loss": 0.1793,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.96875,
"rewards/margins": 3.75,
"rewards/rejected": -8.6875,
"step": 7140
},
{
"epoch": 0.5379580167030321,
"grad_norm": 6.603760811837454,
"learning_rate": 2.6053583082358887e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.484375,
"logps/chosen": -616.0,
"logps/rejected": -984.0,
"loss": 0.1813,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.625,
"rewards/margins": 3.78125,
"rewards/rejected": -8.375,
"step": 7150
},
{
"epoch": 0.5387104055375819,
"grad_norm": 12.925965387407317,
"learning_rate": 2.598797460795772e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.546875,
"logps/chosen": -628.0,
"logps/rejected": -960.0,
"loss": 0.1801,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.5625,
"rewards/margins": 3.6875,
"rewards/rejected": -8.25,
"step": 7160
},
{
"epoch": 0.5394627943721315,
"grad_norm": 7.161982538239592,
"learning_rate": 2.5922359317859195e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.53125,
"logps/chosen": -616.0,
"logps/rejected": -1008.0,
"loss": 0.1821,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.5,
"rewards/margins": 4.03125,
"rewards/rejected": -8.5625,
"step": 7170
},
{
"epoch": 0.5402151832066812,
"grad_norm": 9.670500724682935,
"learning_rate": 2.585673766472065e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -652.0,
"logps/rejected": -1048.0,
"loss": 0.1662,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.8125,
"rewards/margins": 4.125,
"rewards/rejected": -8.9375,
"step": 7180
},
{
"epoch": 0.5409675720412309,
"grad_norm": 10.653981044551228,
"learning_rate": 2.5791110101243337e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.46875,
"logps/chosen": -720.0,
"logps/rejected": -1088.0,
"loss": 0.2079,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.65625,
"rewards/margins": 3.828125,
"rewards/rejected": -9.4375,
"step": 7190
},
{
"epoch": 0.5417199608757806,
"grad_norm": 8.566022489278872,
"learning_rate": 2.5725477080169243e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.421875,
"logps/chosen": -728.0,
"logps/rejected": -1080.0,
"loss": 0.209,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.59375,
"rewards/margins": 3.59375,
"rewards/rejected": -9.1875,
"step": 7200
},
{
"epoch": 0.5424723497103303,
"grad_norm": 9.199996930661813,
"learning_rate": 2.565983905427806e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.453125,
"logps/chosen": -676.0,
"logps/rejected": -1004.0,
"loss": 0.1991,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.28125,
"rewards/margins": 3.28125,
"rewards/rejected": -8.5625,
"step": 7210
},
{
"epoch": 0.54322473854488,
"grad_norm": 10.610837436282099,
"learning_rate": 2.559419647638395e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -660.0,
"logps/rejected": -1000.0,
"loss": 0.2025,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.09375,
"rewards/margins": 3.5625,
"rewards/rejected": -8.625,
"step": 7220
},
{
"epoch": 0.5439771273794297,
"grad_norm": 7.6443150182433035,
"learning_rate": 2.5528549799332515e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.375,
"logps/chosen": -688.0,
"logps/rejected": -1048.0,
"loss": 0.202,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.40625,
"rewards/margins": 3.6875,
"rewards/rejected": -9.0625,
"step": 7230
},
{
"epoch": 0.5447295162139794,
"grad_norm": 10.089524522681916,
"learning_rate": 2.5462899475997617e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.421875,
"logps/chosen": -684.0,
"logps/rejected": -1056.0,
"loss": 0.2053,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.21875,
"rewards/margins": 3.921875,
"rewards/rejected": -9.125,
"step": 7240
},
{
"epoch": 0.5454819050485291,
"grad_norm": 11.14010778305217,
"learning_rate": 2.5397245959278284e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.453125,
"logps/chosen": -680.0,
"logps/rejected": -1040.0,
"loss": 0.2039,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.1875,
"rewards/margins": 3.640625,
"rewards/rejected": -8.8125,
"step": 7250
},
{
"epoch": 0.5462342938830788,
"grad_norm": 9.959907622564678,
"learning_rate": 2.533158970209558e-07,
"logits/chosen": -2.515625,
"logits/rejected": -2.25,
"logps/chosen": -696.0,
"logps/rejected": -1048.0,
"loss": 0.1998,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.1875,
"rewards/margins": 3.640625,
"rewards/rejected": -8.8125,
"step": 7260
},
{
"epoch": 0.5469866827176285,
"grad_norm": 9.70323307418387,
"learning_rate": 2.526593115738945e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.3125,
"logps/chosen": -716.0,
"logps/rejected": -1056.0,
"loss": 0.1889,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.53125,
"rewards/margins": 3.59375,
"rewards/rejected": -9.125,
"step": 7270
},
{
"epoch": 0.5477390715521782,
"grad_norm": 12.429429648330792,
"learning_rate": 2.5200270778115634e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.390625,
"logps/chosen": -692.0,
"logps/rejected": -1064.0,
"loss": 0.1802,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 4.0625,
"rewards/rejected": -9.1875,
"step": 7280
},
{
"epoch": 0.5484914603867279,
"grad_norm": 12.425165028439498,
"learning_rate": 2.513460901724253e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.53125,
"logps/chosen": -692.0,
"logps/rejected": -1064.0,
"loss": 0.1761,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.3125,
"rewards/margins": 3.8125,
"rewards/rejected": -9.125,
"step": 7290
},
{
"epoch": 0.5492438492212776,
"grad_norm": 6.42017367577843,
"learning_rate": 2.5068946327748073e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.34375,
"logps/chosen": -652.0,
"logps/rejected": -1128.0,
"loss": 0.182,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 4.65625,
"rewards/rejected": -9.625,
"step": 7300
},
{
"epoch": 0.5499962380558272,
"grad_norm": 10.146319490014568,
"learning_rate": 2.5003283162616585e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -696.0,
"logps/rejected": -1080.0,
"loss": 0.1667,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.21875,
"rewards/margins": 4.03125,
"rewards/rejected": -9.25,
"step": 7310
},
{
"epoch": 0.550748626890377,
"grad_norm": 9.133315990487066,
"learning_rate": 2.493761997483569e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -700.0,
"logps/rejected": -1096.0,
"loss": 0.1965,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.5,
"rewards/margins": 3.9375,
"rewards/rejected": -9.4375,
"step": 7320
},
{
"epoch": 0.5515010157249266,
"grad_norm": 9.96126188722454,
"learning_rate": 2.4871957217393155e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -720.0,
"logps/rejected": -1080.0,
"loss": 0.2008,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.46875,
"rewards/margins": 3.6875,
"rewards/rejected": -9.125,
"step": 7330
},
{
"epoch": 0.5522534045594764,
"grad_norm": 7.703266984369524,
"learning_rate": 2.480629534327378e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.34375,
"logps/chosen": -668.0,
"logps/rejected": -1056.0,
"loss": 0.1767,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.03125,
"rewards/margins": 3.96875,
"rewards/rejected": -9.0,
"step": 7340
},
{
"epoch": 0.553005793394026,
"grad_norm": 11.036520110749903,
"learning_rate": 2.474063480545628e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.5625,
"logps/chosen": -636.0,
"logps/rejected": -1012.0,
"loss": 0.1908,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.84375,
"rewards/margins": 3.8125,
"rewards/rejected": -8.625,
"step": 7350
},
{
"epoch": 0.5537581822285758,
"grad_norm": 10.372930262089573,
"learning_rate": 2.4674976056910136e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -628.0,
"logps/rejected": -992.0,
"loss": 0.1867,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.65625,
"rewards/margins": 3.765625,
"rewards/rejected": -8.4375,
"step": 7360
},
{
"epoch": 0.5545105710631254,
"grad_norm": 7.469778055442896,
"learning_rate": 2.460931955059251e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.5625,
"logps/chosen": -628.0,
"logps/rejected": -1032.0,
"loss": 0.1806,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.875,
"rewards/margins": 4.03125,
"rewards/rejected": -8.875,
"step": 7370
},
{
"epoch": 0.5552629598976752,
"grad_norm": 8.649079014691724,
"learning_rate": 2.4543665739445054e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.53125,
"logps/chosen": -624.0,
"logps/rejected": -1032.0,
"loss": 0.1668,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.625,
"rewards/margins": 4.1875,
"rewards/rejected": -8.8125,
"step": 7380
},
{
"epoch": 0.5560153487322248,
"grad_norm": 8.931574849238933,
"learning_rate": 2.447801507639087e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.515625,
"logps/chosen": -664.0,
"logps/rejected": -1040.0,
"loss": 0.2007,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.09375,
"rewards/margins": 3.75,
"rewards/rejected": -8.875,
"step": 7390
},
{
"epoch": 0.5567677375667746,
"grad_norm": 13.202148704345227,
"learning_rate": 2.4412368014331326e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.359375,
"logps/chosen": -644.0,
"logps/rejected": -1048.0,
"loss": 0.214,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.03125,
"rewards/margins": 4.03125,
"rewards/rejected": -9.0625,
"step": 7400
},
{
"epoch": 0.5575201264013242,
"grad_norm": 5.9662166661161775,
"learning_rate": 2.434672500614294e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.421875,
"logps/chosen": -632.0,
"logps/rejected": -988.0,
"loss": 0.1808,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.71875,
"rewards/margins": 3.65625,
"rewards/rejected": -8.375,
"step": 7410
},
{
"epoch": 0.5582725152358738,
"grad_norm": 8.568530720173657,
"learning_rate": 2.428108650467427e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.46875,
"logps/chosen": -700.0,
"logps/rejected": -1032.0,
"loss": 0.1973,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.28125,
"rewards/margins": 3.40625,
"rewards/rejected": -8.6875,
"step": 7420
},
{
"epoch": 0.5590249040704236,
"grad_norm": 9.518151204978215,
"learning_rate": 2.4215452962742795e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -676.0,
"logps/rejected": -1056.0,
"loss": 0.176,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.25,
"rewards/margins": 3.890625,
"rewards/rejected": -9.125,
"step": 7430
},
{
"epoch": 0.5597772929049732,
"grad_norm": 9.381073859805955,
"learning_rate": 2.4149824833131766e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.546875,
"logps/chosen": -648.0,
"logps/rejected": -1064.0,
"loss": 0.1927,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 4.34375,
"rewards/rejected": -9.3125,
"step": 7440
},
{
"epoch": 0.560529681739523,
"grad_norm": 8.53592549544634,
"learning_rate": 2.408420256858709e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.578125,
"logps/chosen": -632.0,
"logps/rejected": -996.0,
"loss": 0.1949,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.84375,
"rewards/margins": 3.609375,
"rewards/rejected": -8.4375,
"step": 7450
},
{
"epoch": 0.5612820705740726,
"grad_norm": 7.3386201760751915,
"learning_rate": 2.4018586621814245e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.5,
"logps/chosen": -660.0,
"logps/rejected": -1032.0,
"loss": 0.174,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.9375,
"rewards/margins": 3.78125,
"rewards/rejected": -8.6875,
"step": 7460
},
{
"epoch": 0.5620344594086224,
"grad_norm": 9.509046000241845,
"learning_rate": 2.395297744547507e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.4375,
"logps/chosen": -660.0,
"logps/rejected": -996.0,
"loss": 0.2118,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.03125,
"rewards/margins": 3.4375,
"rewards/rejected": -8.5,
"step": 7470
},
{
"epoch": 0.562786848243172,
"grad_norm": 6.611905514599548,
"learning_rate": 2.3887375492184754e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.53125,
"logps/chosen": -672.0,
"logps/rejected": -1040.0,
"loss": 0.1714,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.125,
"rewards/margins": 3.75,
"rewards/rejected": -8.875,
"step": 7480
},
{
"epoch": 0.5635392370777218,
"grad_norm": 10.287896987883586,
"learning_rate": 2.3821781214508624e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.5625,
"logps/chosen": -692.0,
"logps/rejected": -1024.0,
"loss": 0.1993,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.15625,
"rewards/margins": 3.46875,
"rewards/rejected": -8.625,
"step": 7490
},
{
"epoch": 0.5642916259122714,
"grad_norm": 7.132483843089514,
"learning_rate": 2.3756195064959062e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.453125,
"logps/chosen": -660.0,
"logps/rejected": -1056.0,
"loss": 0.1616,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.9375,
"rewards/margins": 4.1875,
"rewards/rejected": -9.125,
"step": 7500
},
{
"epoch": 0.5650440147468212,
"grad_norm": 9.38473146664766,
"learning_rate": 2.369061749599238e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.5625,
"logps/chosen": -688.0,
"logps/rejected": -1072.0,
"loss": 0.1822,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 4.0625,
"rewards/rejected": -9.125,
"step": 7510
},
{
"epoch": 0.5657964035813708,
"grad_norm": 11.06870568028871,
"learning_rate": 2.36250489600057e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -656.0,
"logps/rejected": -1020.0,
"loss": 0.1714,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.875,
"rewards/margins": 3.84375,
"rewards/rejected": -8.75,
"step": 7520
},
{
"epoch": 0.5665487924159206,
"grad_norm": 13.638691339075915,
"learning_rate": 2.3559489909333812e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.453125,
"logps/chosen": -676.0,
"logps/rejected": -1048.0,
"loss": 0.1795,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.1875,
"rewards/margins": 3.765625,
"rewards/rejected": -8.9375,
"step": 7530
},
{
"epoch": 0.5673011812504702,
"grad_norm": 12.446245321810611,
"learning_rate": 2.3493940796246088e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.5625,
"logps/chosen": -676.0,
"logps/rejected": -1096.0,
"loss": 0.1821,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.34375,
"rewards/margins": 4.09375,
"rewards/rejected": -9.4375,
"step": 7540
},
{
"epoch": 0.5680535700850199,
"grad_norm": 10.008249249632032,
"learning_rate": 2.342840207294335e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -692.0,
"logps/rejected": -1088.0,
"loss": 0.1748,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.3125,
"rewards/margins": 4.03125,
"rewards/rejected": -9.375,
"step": 7550
},
{
"epoch": 0.5688059589195696,
"grad_norm": 7.044727387892051,
"learning_rate": 2.33628741915547e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.34375,
"logps/chosen": -672.0,
"logps/rejected": -1072.0,
"loss": 0.1548,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.34375,
"rewards/margins": 3.953125,
"rewards/rejected": -9.25,
"step": 7560
},
{
"epoch": 0.5695583477541193,
"grad_norm": 8.545277753714306,
"learning_rate": 2.32973576041345e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.40625,
"logps/chosen": -696.0,
"logps/rejected": -1120.0,
"loss": 0.1877,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.3125,
"rewards/margins": 4.28125,
"rewards/rejected": -9.625,
"step": 7570
},
{
"epoch": 0.570310736588669,
"grad_norm": 8.15048607027103,
"learning_rate": 2.3231852762659165e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.515625,
"logps/chosen": -672.0,
"logps/rejected": -1072.0,
"loss": 0.1849,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.125,
"rewards/margins": 4.125,
"rewards/rejected": -9.25,
"step": 7580
},
{
"epoch": 0.5710631254232187,
"grad_norm": 8.425289105784666,
"learning_rate": 2.3166360119024094e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.328125,
"logps/chosen": -752.0,
"logps/rejected": -1112.0,
"loss": 0.1735,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.65625,
"rewards/margins": 3.875,
"rewards/rejected": -9.5,
"step": 7590
},
{
"epoch": 0.5718155142577684,
"grad_norm": 7.974257289934345,
"learning_rate": 2.310088012504053e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.3125,
"logps/chosen": -720.0,
"logps/rejected": -1104.0,
"loss": 0.1985,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.6875,
"rewards/margins": 3.78125,
"rewards/rejected": -9.5,
"step": 7600
},
{
"epoch": 0.5725679030923181,
"grad_norm": 7.55519028519371,
"learning_rate": 2.303541323243246e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -732.0,
"logps/rejected": -1088.0,
"loss": 0.1793,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.75,
"rewards/margins": 3.671875,
"rewards/rejected": -9.4375,
"step": 7610
},
{
"epoch": 0.5733202919268678,
"grad_norm": 8.69426857579776,
"learning_rate": 2.2969959892833474e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.296875,
"logps/chosen": -720.0,
"logps/rejected": -1152.0,
"loss": 0.1844,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.71875,
"rewards/margins": 4.15625,
"rewards/rejected": -9.875,
"step": 7620
},
{
"epoch": 0.5740726807614175,
"grad_norm": 7.995168153719178,
"learning_rate": 2.290452055778367e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.4375,
"logps/chosen": -728.0,
"logps/rejected": -1104.0,
"loss": 0.1952,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.65625,
"rewards/margins": 3.96875,
"rewards/rejected": -9.625,
"step": 7630
},
{
"epoch": 0.5748250695959672,
"grad_norm": 9.47978147859696,
"learning_rate": 2.2839095678726553e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.46875,
"logps/chosen": -680.0,
"logps/rejected": -1080.0,
"loss": 0.1751,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 3.890625,
"rewards/rejected": -9.1875,
"step": 7640
},
{
"epoch": 0.5755774584305169,
"grad_norm": 8.068045603290606,
"learning_rate": 2.2773685707005863e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -744.0,
"logps/rejected": -1112.0,
"loss": 0.1766,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.6875,
"rewards/margins": 3.84375,
"rewards/rejected": -9.5,
"step": 7650
},
{
"epoch": 0.5763298472650666,
"grad_norm": 10.132016914801373,
"learning_rate": 2.270829109386253e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -712.0,
"logps/rejected": -1080.0,
"loss": 0.1861,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.40625,
"rewards/margins": 3.8125,
"rewards/rejected": -9.25,
"step": 7660
},
{
"epoch": 0.5770822360996163,
"grad_norm": 7.187891351598791,
"learning_rate": 2.2642912290431518e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.453125,
"logps/chosen": -688.0,
"logps/rejected": -1120.0,
"loss": 0.184,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.3125,
"rewards/margins": 4.375,
"rewards/rejected": -9.6875,
"step": 7670
},
{
"epoch": 0.5778346249341659,
"grad_norm": 7.084610122669814,
"learning_rate": 2.257754974773873e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.359375,
"logps/chosen": -744.0,
"logps/rejected": -1160.0,
"loss": 0.2082,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.78125,
"rewards/margins": 4.25,
"rewards/rejected": -10.0,
"step": 7680
},
{
"epoch": 0.5785870137687157,
"grad_norm": 7.898558900449446,
"learning_rate": 2.2512203916697892e-07,
"logits/chosen": -2.578125,
"logits/rejected": -2.359375,
"logps/chosen": -712.0,
"logps/rejected": -1080.0,
"loss": 0.1747,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.46875,
"rewards/margins": 3.75,
"rewards/rejected": -9.1875,
"step": 7690
},
{
"epoch": 0.5793394026032653,
"grad_norm": 7.349864652389393,
"learning_rate": 2.2446875248107432e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.296875,
"logps/chosen": -704.0,
"logps/rejected": -1112.0,
"loss": 0.188,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.46875,
"rewards/margins": 4.125,
"rewards/rejected": -9.5625,
"step": 7700
},
{
"epoch": 0.5800917914378151,
"grad_norm": 7.426858571337103,
"learning_rate": 2.2381564192647396e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.328125,
"logps/chosen": -676.0,
"logps/rejected": -1048.0,
"loss": 0.1832,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.125,
"rewards/margins": 3.78125,
"rewards/rejected": -8.9375,
"step": 7710
},
{
"epoch": 0.5808441802723647,
"grad_norm": 9.780007794128954,
"learning_rate": 2.231627120087631e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.328125,
"logps/chosen": -664.0,
"logps/rejected": -1088.0,
"loss": 0.1757,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.09375,
"rewards/margins": 4.25,
"rewards/rejected": -9.3125,
"step": 7720
},
{
"epoch": 0.5815965691069145,
"grad_norm": 9.862081709378993,
"learning_rate": 2.2250996723228104e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -700.0,
"logps/rejected": -1104.0,
"loss": 0.1918,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.4375,
"rewards/margins": 4.03125,
"rewards/rejected": -9.5,
"step": 7730
},
{
"epoch": 0.5823489579414641,
"grad_norm": 9.796114568966233,
"learning_rate": 2.2185741210008947e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.4375,
"logps/chosen": -684.0,
"logps/rejected": -1072.0,
"loss": 0.1859,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.0625,
"rewards/margins": 3.90625,
"rewards/rejected": -9.0,
"step": 7740
},
{
"epoch": 0.5831013467760139,
"grad_norm": 9.088441129163732,
"learning_rate": 2.2120505111394224e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.5,
"logps/chosen": -692.0,
"logps/rejected": -1048.0,
"loss": 0.1815,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.21875,
"rewards/margins": 3.609375,
"rewards/rejected": -8.8125,
"step": 7750
},
{
"epoch": 0.5838537356105635,
"grad_norm": 11.190049192399824,
"learning_rate": 2.205528887742535e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.328125,
"logps/chosen": -720.0,
"logps/rejected": -1112.0,
"loss": 0.1963,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.375,
"rewards/margins": 4.0,
"rewards/rejected": -9.375,
"step": 7760
},
{
"epoch": 0.5846061244451133,
"grad_norm": 7.882822688571665,
"learning_rate": 2.199009295800672e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.34375,
"logps/chosen": -712.0,
"logps/rejected": -1096.0,
"loss": 0.1811,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.5625,
"rewards/margins": 3.9375,
"rewards/rejected": -9.5,
"step": 7770
},
{
"epoch": 0.5853585132796629,
"grad_norm": 12.163265635174126,
"learning_rate": 2.192491780290259e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.40625,
"logps/chosen": -740.0,
"logps/rejected": -1144.0,
"loss": 0.18,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.65625,
"rewards/margins": 4.25,
"rewards/rejected": -9.875,
"step": 7780
},
{
"epoch": 0.5861109021142126,
"grad_norm": 11.023139609307764,
"learning_rate": 2.1859763861733947e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.375,
"logps/chosen": -688.0,
"logps/rejected": -1040.0,
"loss": 0.1954,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -5.34375,
"rewards/margins": 3.6875,
"rewards/rejected": -9.0,
"step": 7790
},
{
"epoch": 0.5868632909487623,
"grad_norm": 8.846784626177584,
"learning_rate": 2.179463158397545e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.4375,
"logps/chosen": -712.0,
"logps/rejected": -1072.0,
"loss": 0.1868,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.625,
"rewards/margins": 3.703125,
"rewards/rejected": -9.3125,
"step": 7800
},
{
"epoch": 0.587615679783312,
"grad_norm": 9.983411475943397,
"learning_rate": 2.1729521418952304e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.53125,
"logps/chosen": -668.0,
"logps/rejected": -1104.0,
"loss": 0.1717,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.0625,
"rewards/margins": 4.4375,
"rewards/rejected": -9.5,
"step": 7810
},
{
"epoch": 0.5883680686178617,
"grad_norm": 7.96272848222726,
"learning_rate": 2.166443381583718e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.453125,
"logps/chosen": -716.0,
"logps/rejected": -1112.0,
"loss": 0.1632,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.6875,
"rewards/margins": 3.8125,
"rewards/rejected": -9.5,
"step": 7820
},
{
"epoch": 0.5891204574524114,
"grad_norm": 8.079780562583833,
"learning_rate": 2.1599369223647068e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.546875,
"logps/chosen": -716.0,
"logps/rejected": -1072.0,
"loss": 0.2225,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.53125,
"rewards/margins": 3.78125,
"rewards/rejected": -9.3125,
"step": 7830
},
{
"epoch": 0.5898728462869611,
"grad_norm": 12.685745100141066,
"learning_rate": 2.1534328091240238e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.546875,
"logps/chosen": -720.0,
"logps/rejected": -1096.0,
"loss": 0.198,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.5625,
"rewards/margins": 3.765625,
"rewards/rejected": -9.3125,
"step": 7840
},
{
"epoch": 0.5906252351215108,
"grad_norm": 9.694926490966695,
"learning_rate": 2.1469310867313118e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.53125,
"logps/chosen": -720.0,
"logps/rejected": -1088.0,
"loss": 0.1962,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 4.0,
"rewards/rejected": -9.3125,
"step": 7850
},
{
"epoch": 0.5913776239560605,
"grad_norm": 7.470161379455044,
"learning_rate": 2.1404318000397192e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.359375,
"logps/chosen": -644.0,
"logps/rejected": -1008.0,
"loss": 0.1719,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.75,
"rewards/margins": 3.875,
"rewards/rejected": -8.625,
"step": 7860
},
{
"epoch": 0.5921300127906102,
"grad_norm": 15.635656624979678,
"learning_rate": 2.1339349938855929e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.46875,
"logps/chosen": -624.0,
"logps/rejected": -1000.0,
"loss": 0.189,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.84375,
"rewards/rejected": -8.6875,
"step": 7870
},
{
"epoch": 0.5928824016251599,
"grad_norm": 8.84669114325579,
"learning_rate": 2.127440713088165e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.515625,
"logps/chosen": -676.0,
"logps/rejected": -1080.0,
"loss": 0.1959,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -5.125,
"rewards/margins": 4.125,
"rewards/rejected": -9.25,
"step": 7880
},
{
"epoch": 0.5936347904597096,
"grad_norm": 10.608839437296322,
"learning_rate": 2.1209490024492477e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1040.0,
"loss": 0.19,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.125,
"rewards/margins": 3.828125,
"rewards/rejected": -8.9375,
"step": 7890
},
{
"epoch": 0.5943871792942593,
"grad_norm": 11.261167695549236,
"learning_rate": 2.114459906752923e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -696.0,
"logps/rejected": -1056.0,
"loss": 0.1984,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.28125,
"rewards/margins": 3.8125,
"rewards/rejected": -9.125,
"step": 7900
},
{
"epoch": 0.595139568128809,
"grad_norm": 7.217595152184596,
"learning_rate": 2.1079734707652315e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -704.0,
"logps/rejected": -1040.0,
"loss": 0.206,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.3125,
"rewards/margins": 3.578125,
"rewards/rejected": -8.875,
"step": 7910
},
{
"epoch": 0.5958919569633586,
"grad_norm": 6.7734942212408855,
"learning_rate": 2.101489739233867e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.3125,
"logps/chosen": -656.0,
"logps/rejected": -1072.0,
"loss": 0.1851,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.96875,
"rewards/margins": 4.25,
"rewards/rejected": -9.1875,
"step": 7920
},
{
"epoch": 0.5966443457979084,
"grad_norm": 8.189998016070621,
"learning_rate": 2.0950087568878643e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.375,
"logps/chosen": -696.0,
"logps/rejected": -1080.0,
"loss": 0.1908,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.28125,
"rewards/margins": 4.0625,
"rewards/rejected": -9.3125,
"step": 7930
},
{
"epoch": 0.597396734632458,
"grad_norm": 8.188292839949094,
"learning_rate": 2.088530568437295e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.40625,
"logps/chosen": -712.0,
"logps/rejected": -1088.0,
"loss": 0.1911,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.5625,
"rewards/margins": 3.84375,
"rewards/rejected": -9.4375,
"step": 7940
},
{
"epoch": 0.5981491234670078,
"grad_norm": 9.397579034044462,
"learning_rate": 2.0820552185729553e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.484375,
"logps/chosen": -696.0,
"logps/rejected": -1064.0,
"loss": 0.1733,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.25,
"rewards/margins": 3.890625,
"rewards/rejected": -9.125,
"step": 7950
},
{
"epoch": 0.5989015123015574,
"grad_norm": 9.047964513784462,
"learning_rate": 2.0755827519660585e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.453125,
"logps/chosen": -668.0,
"logps/rejected": -1040.0,
"loss": 0.1839,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 4.0,
"rewards/rejected": -8.9375,
"step": 7960
},
{
"epoch": 0.5996539011361072,
"grad_norm": 11.445154601848932,
"learning_rate": 2.069113213267928e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.484375,
"logps/chosen": -672.0,
"logps/rejected": -1032.0,
"loss": 0.1736,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 3.71875,
"rewards/rejected": -8.8125,
"step": 7970
},
{
"epoch": 0.6004062899706568,
"grad_norm": 10.501805165582523,
"learning_rate": 2.062646647109688e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.546875,
"logps/chosen": -632.0,
"logps/rejected": -1040.0,
"loss": 0.1783,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.78125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.0,
"step": 7980
},
{
"epoch": 0.6011586788052066,
"grad_norm": 10.339753662857872,
"learning_rate": 2.0561830981019582e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.421875,
"logps/chosen": -704.0,
"logps/rejected": -1128.0,
"loss": 0.185,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.40625,
"rewards/margins": 4.3125,
"rewards/rejected": -9.75,
"step": 7990
},
{
"epoch": 0.6019110676397562,
"grad_norm": 5.477932125415138,
"learning_rate": 2.04972261083454e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.484375,
"logps/chosen": -704.0,
"logps/rejected": -1088.0,
"loss": 0.1581,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.3125,
"rewards/margins": 3.890625,
"rewards/rejected": -9.1875,
"step": 8000
},
{
"epoch": 0.602663456474306,
"grad_norm": 10.860208268593594,
"learning_rate": 2.0432652298761176e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.484375,
"logps/chosen": -684.0,
"logps/rejected": -1056.0,
"loss": 0.2022,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.25,
"rewards/margins": 3.71875,
"rewards/rejected": -9.0,
"step": 8010
},
{
"epoch": 0.6034158453088556,
"grad_norm": 8.134360329596852,
"learning_rate": 2.0368109997739415e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -664.0,
"logps/rejected": -1104.0,
"loss": 0.1554,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.0625,
"rewards/margins": 4.34375,
"rewards/rejected": -9.4375,
"step": 8020
},
{
"epoch": 0.6041682341434053,
"grad_norm": 8.877210411595883,
"learning_rate": 2.0303599650535283e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.34375,
"logps/chosen": -724.0,
"logps/rejected": -1096.0,
"loss": 0.1929,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.53125,
"rewards/margins": 3.875,
"rewards/rejected": -9.375,
"step": 8030
},
{
"epoch": 0.604920622977955,
"grad_norm": 8.533779164861182,
"learning_rate": 2.0239121702183505e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.40625,
"logps/chosen": -652.0,
"logps/rejected": -1072.0,
"loss": 0.1867,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 4.1875,
"rewards/rejected": -9.25,
"step": 8040
},
{
"epoch": 0.6056730118125047,
"grad_norm": 7.678079728481108,
"learning_rate": 2.017467659749528e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.46875,
"logps/chosen": -740.0,
"logps/rejected": -1112.0,
"loss": 0.1775,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.4375,
"rewards/margins": 4.1875,
"rewards/rejected": -9.625,
"step": 8050
},
{
"epoch": 0.6064254006470544,
"grad_norm": 7.362214903304492,
"learning_rate": 2.011026478105525e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.578125,
"logps/chosen": -664.0,
"logps/rejected": -1064.0,
"loss": 0.1865,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.0,
"rewards/margins": 4.03125,
"rewards/rejected": -9.0,
"step": 8060
},
{
"epoch": 0.6071777894816041,
"grad_norm": 6.554803722933717,
"learning_rate": 2.004588669721841e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -684.0,
"logps/rejected": -1096.0,
"loss": 0.1729,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.28125,
"rewards/margins": 4.09375,
"rewards/rejected": -9.375,
"step": 8070
},
{
"epoch": 0.6079301783161538,
"grad_norm": 7.301637306446734,
"learning_rate": 1.9981542790107032e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.515625,
"logps/chosen": -676.0,
"logps/rejected": -1080.0,
"loss": 0.1685,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.9375,
"rewards/margins": 4.15625,
"rewards/rejected": -9.0625,
"step": 8080
},
{
"epoch": 0.6086825671507035,
"grad_norm": 9.634613626300386,
"learning_rate": 1.9917233503607625e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.453125,
"logps/chosen": -660.0,
"logps/rejected": -1088.0,
"loss": 0.1636,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.0625,
"rewards/margins": 4.40625,
"rewards/rejected": -9.4375,
"step": 8090
},
{
"epoch": 0.6094349559852532,
"grad_norm": 9.650023332883672,
"learning_rate": 1.9852959281367872e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.46875,
"logps/chosen": -720.0,
"logps/rejected": -1088.0,
"loss": 0.1797,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.65625,
"rewards/margins": 3.78125,
"rewards/rejected": -9.4375,
"step": 8100
},
{
"epoch": 0.6101873448198029,
"grad_norm": 8.50128482310566,
"learning_rate": 1.9788720566793527e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.421875,
"logps/chosen": -704.0,
"logps/rejected": -1056.0,
"loss": 0.1888,
"rewards/accuracies": 0.862500011920929,
"rewards/chosen": -5.5,
"rewards/margins": 3.59375,
"rewards/rejected": -9.125,
"step": 8110
},
{
"epoch": 0.6109397336543526,
"grad_norm": 7.556379718476982,
"learning_rate": 1.9724517803045417e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.453125,
"logps/chosen": -712.0,
"logps/rejected": -1080.0,
"loss": 0.1978,
"rewards/accuracies": 0.8687499761581421,
"rewards/chosen": -5.5625,
"rewards/margins": 3.5625,
"rewards/rejected": -9.125,
"step": 8120
},
{
"epoch": 0.6116921224889023,
"grad_norm": 9.095065944956065,
"learning_rate": 1.966035143303636e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.421875,
"logps/chosen": -700.0,
"logps/rejected": -1024.0,
"loss": 0.1675,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.1875,
"rewards/margins": 3.53125,
"rewards/rejected": -8.75,
"step": 8130
},
{
"epoch": 0.612444511323452,
"grad_norm": 8.963937404932498,
"learning_rate": 1.959622189942808e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -684.0,
"logps/rejected": -1072.0,
"loss": 0.17,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.21875,
"rewards/margins": 3.828125,
"rewards/rejected": -9.0625,
"step": 8140
},
{
"epoch": 0.6131969001580017,
"grad_norm": 8.450587169477771,
"learning_rate": 1.9532129644628204e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.453125,
"logps/chosen": -680.0,
"logps/rejected": -1020.0,
"loss": 0.1888,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.25,
"rewards/margins": 3.46875,
"rewards/rejected": -8.6875,
"step": 8150
},
{
"epoch": 0.6139492889925513,
"grad_norm": 10.13442209524969,
"learning_rate": 1.946807511078718e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.46875,
"logps/chosen": -712.0,
"logps/rejected": -1088.0,
"loss": 0.1807,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.34375,
"rewards/margins": 4.0625,
"rewards/rejected": -9.375,
"step": 8160
},
{
"epoch": 0.6147016778271011,
"grad_norm": 13.447019109532024,
"learning_rate": 1.9404058739795217e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.5,
"logps/chosen": -708.0,
"logps/rejected": -1088.0,
"loss": 0.1775,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.40625,
"rewards/margins": 4.0,
"rewards/rejected": -9.4375,
"step": 8170
},
{
"epoch": 0.6154540666616507,
"grad_norm": 8.774865738407245,
"learning_rate": 1.9340080973279268e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.4375,
"logps/chosen": -696.0,
"logps/rejected": -1120.0,
"loss": 0.164,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.3125,
"rewards/margins": 4.3125,
"rewards/rejected": -9.625,
"step": 8180
},
{
"epoch": 0.6162064554962005,
"grad_norm": 11.89253519639188,
"learning_rate": 1.9276142252599971e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -704.0,
"logps/rejected": -1080.0,
"loss": 0.1704,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.4375,
"rewards/margins": 3.953125,
"rewards/rejected": -9.375,
"step": 8190
},
{
"epoch": 0.6169588443307501,
"grad_norm": 11.177241943586392,
"learning_rate": 1.9212243018848572e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.546875,
"logps/chosen": -692.0,
"logps/rejected": -1128.0,
"loss": 0.1966,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.3125,
"rewards/margins": 4.375,
"rewards/rejected": -9.6875,
"step": 8200
},
{
"epoch": 0.6177112331652999,
"grad_norm": 7.67297793851783,
"learning_rate": 1.9148383712843946e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.578125,
"logps/chosen": -724.0,
"logps/rejected": -1096.0,
"loss": 0.1746,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.5,
"rewards/margins": 4.0,
"rewards/rejected": -9.5,
"step": 8210
},
{
"epoch": 0.6184636219998495,
"grad_norm": 6.957493941018748,
"learning_rate": 1.908456477512949e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.421875,
"logps/chosen": -708.0,
"logps/rejected": -1120.0,
"loss": 0.1758,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.40625,
"rewards/margins": 4.25,
"rewards/rejected": -9.6875,
"step": 8220
},
{
"epoch": 0.6192160108343993,
"grad_norm": 8.931938647060058,
"learning_rate": 1.9020786645970132e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.359375,
"logps/chosen": -716.0,
"logps/rejected": -1096.0,
"loss": 0.2052,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.46875,
"rewards/margins": 3.90625,
"rewards/rejected": -9.375,
"step": 8230
},
{
"epoch": 0.6199683996689489,
"grad_norm": 11.634130049543522,
"learning_rate": 1.8957049765349275e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.34375,
"logps/chosen": -684.0,
"logps/rejected": -1056.0,
"loss": 0.1936,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.3125,
"rewards/margins": 3.75,
"rewards/rejected": -9.0625,
"step": 8240
},
{
"epoch": 0.6207207885034987,
"grad_norm": 7.675316481738456,
"learning_rate": 1.8893354572965764e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.515625,
"logps/chosen": -684.0,
"logps/rejected": -1072.0,
"loss": 0.1721,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.46875,
"rewards/margins": 3.78125,
"rewards/rejected": -9.25,
"step": 8250
},
{
"epoch": 0.6214731773380483,
"grad_norm": 9.973255437064827,
"learning_rate": 1.882970150823083e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5625,
"logps/chosen": -700.0,
"logps/rejected": -1088.0,
"loss": 0.1814,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.3125,
"rewards/margins": 3.953125,
"rewards/rejected": -9.25,
"step": 8260
},
{
"epoch": 0.6222255661725979,
"grad_norm": 5.705753808858491,
"learning_rate": 1.8766091010265107e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -688.0,
"logps/rejected": -1088.0,
"loss": 0.1667,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 4.25,
"rewards/rejected": -9.4375,
"step": 8270
},
{
"epoch": 0.6229779550071477,
"grad_norm": 9.144340823375144,
"learning_rate": 1.870252351789557e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.59375,
"logps/chosen": -676.0,
"logps/rejected": -1072.0,
"loss": 0.1827,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -5.15625,
"rewards/margins": 4.09375,
"rewards/rejected": -9.25,
"step": 8280
},
{
"epoch": 0.6237303438416973,
"grad_norm": 8.597966874281301,
"learning_rate": 1.8638999469652485e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.46875,
"logps/chosen": -680.0,
"logps/rejected": -1072.0,
"loss": 0.1901,
"rewards/accuracies": 0.875,
"rewards/chosen": -5.3125,
"rewards/margins": 3.875,
"rewards/rejected": -9.1875,
"step": 8290
},
{
"epoch": 0.6244827326762471,
"grad_norm": 8.356923305837611,
"learning_rate": 1.8575519303766456e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.484375,
"logps/chosen": -732.0,
"logps/rejected": -1144.0,
"loss": 0.2041,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.59375,
"rewards/margins": 4.34375,
"rewards/rejected": -9.9375,
"step": 8300
},
{
"epoch": 0.6252351215107967,
"grad_norm": 8.330276529550803,
"learning_rate": 1.8512083458165322e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.5,
"logps/chosen": -696.0,
"logps/rejected": -1080.0,
"loss": 0.1908,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.375,
"rewards/margins": 3.9375,
"rewards/rejected": -9.3125,
"step": 8310
},
{
"epoch": 0.6259875103453465,
"grad_norm": 9.745317202075682,
"learning_rate": 1.8448692370471197e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -680.0,
"logps/rejected": -1064.0,
"loss": 0.1772,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.21875,
"rewards/margins": 3.875,
"rewards/rejected": -9.0625,
"step": 8320
},
{
"epoch": 0.6267398991798961,
"grad_norm": 10.412831393628213,
"learning_rate": 1.838534647799741e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.421875,
"logps/chosen": -708.0,
"logps/rejected": -1032.0,
"loss": 0.1868,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.34375,
"rewards/margins": 3.40625,
"rewards/rejected": -8.75,
"step": 8330
},
{
"epoch": 0.6274922880144459,
"grad_norm": 11.039358030397528,
"learning_rate": 1.83220462177455e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.375,
"logps/chosen": -704.0,
"logps/rejected": -1104.0,
"loss": 0.1793,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.46875,
"rewards/margins": 4.125,
"rewards/rejected": -9.5625,
"step": 8340
},
{
"epoch": 0.6282446768489955,
"grad_norm": 8.167227600473227,
"learning_rate": 1.825879202640222e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.421875,
"logps/chosen": -704.0,
"logps/rejected": -1088.0,
"loss": 0.1827,
"rewards/accuracies": 0.887499988079071,
"rewards/chosen": -5.40625,
"rewards/margins": 3.765625,
"rewards/rejected": -9.1875,
"step": 8350
},
{
"epoch": 0.6289970656835453,
"grad_norm": 9.909887900151999,
"learning_rate": 1.81955843403365e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.40625,
"logps/chosen": -656.0,
"logps/rejected": -1064.0,
"loss": 0.1742,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.125,
"rewards/margins": 4.15625,
"rewards/rejected": -9.25,
"step": 8360
},
{
"epoch": 0.6297494545180949,
"grad_norm": 6.664940260255985,
"learning_rate": 1.8132423595596464e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -700.0,
"logps/rejected": -1080.0,
"loss": 0.1782,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.34375,
"rewards/margins": 3.875,
"rewards/rejected": -9.25,
"step": 8370
},
{
"epoch": 0.6305018433526447,
"grad_norm": 8.040497731217704,
"learning_rate": 1.8069310227906365e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -728.0,
"logps/rejected": -1088.0,
"loss": 0.1883,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.40625,
"rewards/margins": 3.921875,
"rewards/rejected": -9.3125,
"step": 8380
},
{
"epoch": 0.6312542321871943,
"grad_norm": 7.855479759367712,
"learning_rate": 1.8006244672663644e-07,
"logits/chosen": -2.59375,
"logits/rejected": -2.40625,
"logps/chosen": -724.0,
"logps/rejected": -1096.0,
"loss": 0.1708,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.65625,
"rewards/margins": 3.78125,
"rewards/rejected": -9.4375,
"step": 8390
},
{
"epoch": 0.632006621021744,
"grad_norm": 10.782491204548583,
"learning_rate": 1.794322736493591e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -732.0,
"logps/rejected": -1112.0,
"loss": 0.1897,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.53125,
"rewards/margins": 4.03125,
"rewards/rejected": -9.5625,
"step": 8400
},
{
"epoch": 0.6327590098562937,
"grad_norm": 6.395845132416539,
"learning_rate": 1.7880258739457905e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.625,
"logps/chosen": -708.0,
"logps/rejected": -1088.0,
"loss": 0.1809,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.4375,
"rewards/margins": 3.859375,
"rewards/rejected": -9.3125,
"step": 8410
},
{
"epoch": 0.6335113986908434,
"grad_norm": 11.500493086122349,
"learning_rate": 1.7817339230628553e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.53125,
"logps/chosen": -684.0,
"logps/rejected": -1048.0,
"loss": 0.1948,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.09375,
"rewards/margins": 3.921875,
"rewards/rejected": -9.0,
"step": 8420
},
{
"epoch": 0.6342637875253931,
"grad_norm": 9.480030645590286,
"learning_rate": 1.7754469272507914e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.265625,
"logps/chosen": -700.0,
"logps/rejected": -1152.0,
"loss": 0.1784,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.4375,
"rewards/margins": 4.4375,
"rewards/rejected": -9.875,
"step": 8430
},
{
"epoch": 0.6350161763599428,
"grad_norm": 9.324651618478956,
"learning_rate": 1.7691649298814232e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.59375,
"logps/chosen": -660.0,
"logps/rejected": -1056.0,
"loss": 0.168,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.125,
"rewards/margins": 3.921875,
"rewards/rejected": -9.0625,
"step": 8440
},
{
"epoch": 0.6357685651944925,
"grad_norm": 8.774893284452308,
"learning_rate": 1.7628879742920928e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.5625,
"logps/chosen": -656.0,
"logps/rejected": -1136.0,
"loss": 0.1803,
"rewards/accuracies": 0.987500011920929,
"rewards/chosen": -5.0,
"rewards/margins": 4.71875,
"rewards/rejected": -9.75,
"step": 8450
},
{
"epoch": 0.6365209540290422,
"grad_norm": 9.340652243790641,
"learning_rate": 1.7566161037853594e-07,
"logits/chosen": -2.921875,
"logits/rejected": -2.5625,
"logps/chosen": -656.0,
"logps/rejected": -1056.0,
"loss": 0.1712,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.84375,
"rewards/margins": 4.125,
"rewards/rejected": -9.0,
"step": 8460
},
{
"epoch": 0.6372733428635919,
"grad_norm": 8.096050960575042,
"learning_rate": 1.7503493616287023e-07,
"logits/chosen": -2.984375,
"logits/rejected": -2.703125,
"logps/chosen": -660.0,
"logps/rejected": -1088.0,
"loss": 0.202,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.09375,
"rewards/margins": 4.25,
"rewards/rejected": -9.375,
"step": 8470
},
{
"epoch": 0.6380257316981416,
"grad_norm": 7.937269964263276,
"learning_rate": 1.7440877910542225e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.453125,
"logps/chosen": -684.0,
"logps/rejected": -1072.0,
"loss": 0.1962,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.9375,
"rewards/margins": 4.21875,
"rewards/rejected": -9.1875,
"step": 8480
},
{
"epoch": 0.6387781205326913,
"grad_norm": 9.754686407049281,
"learning_rate": 1.7378314352583446e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1080.0,
"loss": 0.1633,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.90625,
"rewards/margins": 4.0,
"rewards/rejected": -8.9375,
"step": 8490
},
{
"epoch": 0.639530509367241,
"grad_norm": 9.467168252886747,
"learning_rate": 1.731580337401517e-07,
"logits/chosen": -2.96875,
"logits/rejected": -2.609375,
"logps/chosen": -648.0,
"logps/rejected": -1048.0,
"loss": 0.1861,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.78125,
"rewards/margins": 4.0,
"rewards/rejected": -8.8125,
"step": 8500
},
{
"epoch": 0.6402828982017907,
"grad_norm": 11.663198748608533,
"learning_rate": 1.7253345406079161e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.671875,
"logps/chosen": -712.0,
"logps/rejected": -1072.0,
"loss": 0.1913,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.375,
"rewards/margins": 3.765625,
"rewards/rejected": -9.125,
"step": 8510
},
{
"epoch": 0.6410352870363404,
"grad_norm": 9.445575901580403,
"learning_rate": 1.719094087965148e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.515625,
"logps/chosen": -688.0,
"logps/rejected": -1104.0,
"loss": 0.1721,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -5.21875,
"rewards/margins": 4.125,
"rewards/rejected": -9.375,
"step": 8520
},
{
"epoch": 0.64178767587089,
"grad_norm": 7.603228701217873,
"learning_rate": 1.7128590225239515e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.6875,
"logps/chosen": -656.0,
"logps/rejected": -1048.0,
"loss": 0.1947,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 4.03125,
"rewards/rejected": -9.0,
"step": 8530
},
{
"epoch": 0.6425400647054398,
"grad_norm": 11.514964298081834,
"learning_rate": 1.7066293872979005e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5,
"logps/chosen": -700.0,
"logps/rejected": -1080.0,
"loss": 0.2049,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 3.90625,
"rewards/rejected": -9.0625,
"step": 8540
},
{
"epoch": 0.6432924535399894,
"grad_norm": 8.976859425848769,
"learning_rate": 1.7004052252631085e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.53125,
"logps/chosen": -676.0,
"logps/rejected": -1064.0,
"loss": 0.1992,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 4.0,
"rewards/rejected": -9.0625,
"step": 8550
},
{
"epoch": 0.6440448423745392,
"grad_norm": 8.397828268335257,
"learning_rate": 1.6941865793579302e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.40625,
"logps/chosen": -620.0,
"logps/rejected": -1048.0,
"loss": 0.177,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.625,
"rewards/margins": 4.25,
"rewards/rejected": -8.8125,
"step": 8560
},
{
"epoch": 0.6447972312090888,
"grad_norm": 7.577018016800494,
"learning_rate": 1.687973492482666e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -644.0,
"logps/rejected": -1024.0,
"loss": 0.1597,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.71875,
"rewards/margins": 3.8125,
"rewards/rejected": -8.5625,
"step": 8570
},
{
"epoch": 0.6455496200436386,
"grad_norm": 11.235843949131914,
"learning_rate": 1.6817660074992694e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.65625,
"logps/chosen": -644.0,
"logps/rejected": -1016.0,
"loss": 0.1608,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.875,
"rewards/margins": 3.984375,
"rewards/rejected": -8.875,
"step": 8580
},
{
"epoch": 0.6463020088781882,
"grad_norm": 9.991635501784412,
"learning_rate": 1.6755641672310456e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.53125,
"logps/chosen": -664.0,
"logps/rejected": -1080.0,
"loss": 0.17,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.3125,
"rewards/margins": 4.125,
"rewards/rejected": -9.4375,
"step": 8590
},
{
"epoch": 0.647054397712738,
"grad_norm": 10.353997306005274,
"learning_rate": 1.6693680144623594e-07,
"logits/chosen": -2.9375,
"logits/rejected": -2.6875,
"logps/chosen": -692.0,
"logps/rejected": -1088.0,
"loss": 0.1757,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.3125,
"rewards/margins": 4.0625,
"rewards/rejected": -9.375,
"step": 8600
},
{
"epoch": 0.6478067865472876,
"grad_norm": 10.78521221480445,
"learning_rate": 1.6631775919383398e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.46875,
"logps/chosen": -688.0,
"logps/rejected": -1112.0,
"loss": 0.1858,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.15625,
"rewards/margins": 4.4375,
"rewards/rejected": -9.625,
"step": 8610
},
{
"epoch": 0.6485591753818374,
"grad_norm": 9.525179269161058,
"learning_rate": 1.656992942364585e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.515625,
"logps/chosen": -688.0,
"logps/rejected": -1096.0,
"loss": 0.1698,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.25,
"rewards/margins": 3.90625,
"rewards/rejected": -9.1875,
"step": 8620
},
{
"epoch": 0.649311564216387,
"grad_norm": 8.303050969135796,
"learning_rate": 1.6508141084068682e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.484375,
"logps/chosen": -664.0,
"logps/rejected": -1056.0,
"loss": 0.1718,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 4.0,
"rewards/rejected": -9.125,
"step": 8630
},
{
"epoch": 0.6500639530509367,
"grad_norm": 10.120955544801877,
"learning_rate": 1.6446411326908415e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.5625,
"logps/chosen": -704.0,
"logps/rejected": -1120.0,
"loss": 0.1684,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.4375,
"rewards/margins": 4.375,
"rewards/rejected": -9.8125,
"step": 8640
},
{
"epoch": 0.6508163418854864,
"grad_norm": 7.469156784372811,
"learning_rate": 1.6384740578017419e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.53125,
"logps/chosen": -708.0,
"logps/rejected": -1128.0,
"loss": 0.1837,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.5,
"rewards/margins": 4.28125,
"rewards/rejected": -9.75,
"step": 8650
},
{
"epoch": 0.6515687307200361,
"grad_norm": 10.847781757212124,
"learning_rate": 1.6323129262841016e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.609375,
"logps/chosen": -672.0,
"logps/rejected": -1080.0,
"loss": 0.1601,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.21875,
"rewards/margins": 4.21875,
"rewards/rejected": -9.4375,
"step": 8660
},
{
"epoch": 0.6523211195545858,
"grad_norm": 11.914460280940185,
"learning_rate": 1.626157780641449e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.546875,
"logps/chosen": -676.0,
"logps/rejected": -1048.0,
"loss": 0.1815,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.09375,
"rewards/margins": 3.953125,
"rewards/rejected": -9.0625,
"step": 8670
},
{
"epoch": 0.6530735083891355,
"grad_norm": 6.998242989519436,
"learning_rate": 1.6200086633360203e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.453125,
"logps/chosen": -664.0,
"logps/rejected": -1072.0,
"loss": 0.1642,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.03125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.25,
"step": 8680
},
{
"epoch": 0.6538258972236852,
"grad_norm": 11.464774461391972,
"learning_rate": 1.6138656167884615e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.4375,
"logps/chosen": -688.0,
"logps/rejected": -1088.0,
"loss": 0.2046,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.46875,
"rewards/margins": 4.03125,
"rewards/rejected": -9.5,
"step": 8690
},
{
"epoch": 0.6545782860582349,
"grad_norm": 8.25921588629842,
"learning_rate": 1.6077286833775407e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5625,
"logps/chosen": -668.0,
"logps/rejected": -1080.0,
"loss": 0.1626,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 4.09375,
"rewards/rejected": -9.25,
"step": 8700
},
{
"epoch": 0.6553306748927846,
"grad_norm": 8.075272226411743,
"learning_rate": 1.6015979054398537e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.46875,
"logps/chosen": -656.0,
"logps/rejected": -1024.0,
"loss": 0.1733,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.0,
"rewards/margins": 3.890625,
"rewards/rejected": -8.875,
"step": 8710
},
{
"epoch": 0.6560830637273343,
"grad_norm": 6.549709126315134,
"learning_rate": 1.5954733252695297e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.484375,
"logps/chosen": -628.0,
"logps/rejected": -1000.0,
"loss": 0.1884,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 3.734375,
"rewards/rejected": -8.5625,
"step": 8720
},
{
"epoch": 0.656835452561884,
"grad_norm": 6.948138061116944,
"learning_rate": 1.5893549851179443e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.4375,
"logps/chosen": -656.0,
"logps/rejected": -1072.0,
"loss": 0.1691,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.03125,
"rewards/margins": 4.125,
"rewards/rejected": -9.1875,
"step": 8730
},
{
"epoch": 0.6575878413964337,
"grad_norm": 10.007531481704225,
"learning_rate": 1.5832429271934216e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.453125,
"logps/chosen": -688.0,
"logps/rejected": -1048.0,
"loss": 0.1776,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.28125,
"rewards/margins": 3.71875,
"rewards/rejected": -9.0,
"step": 8740
},
{
"epoch": 0.6583402302309834,
"grad_norm": 11.37778642015286,
"learning_rate": 1.5771371936609512e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.484375,
"logps/chosen": -692.0,
"logps/rejected": -1064.0,
"loss": 0.1982,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 3.875,
"rewards/rejected": -9.1875,
"step": 8750
},
{
"epoch": 0.6590926190655331,
"grad_norm": 9.387826228804798,
"learning_rate": 1.5710378266418909e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.453125,
"logps/chosen": -684.0,
"logps/rejected": -1120.0,
"loss": 0.1817,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.28125,
"rewards/margins": 4.40625,
"rewards/rejected": -9.6875,
"step": 8760
},
{
"epoch": 0.6598450079000827,
"grad_norm": 9.516184140445384,
"learning_rate": 1.5649448682136768e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -648.0,
"logps/rejected": -1032.0,
"loss": 0.1815,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 3.953125,
"rewards/rejected": -8.9375,
"step": 8770
},
{
"epoch": 0.6605973967346325,
"grad_norm": 11.06259854813874,
"learning_rate": 1.558858360409536e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.4375,
"logps/chosen": -660.0,
"logps/rejected": -1072.0,
"loss": 0.1983,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.15625,
"rewards/margins": 3.96875,
"rewards/rejected": -9.125,
"step": 8780
},
{
"epoch": 0.6613497855691821,
"grad_norm": 8.901877165227225,
"learning_rate": 1.5527783452181947e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.578125,
"logps/chosen": -700.0,
"logps/rejected": -1080.0,
"loss": 0.1985,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 4.0,
"rewards/rejected": -9.3125,
"step": 8790
},
{
"epoch": 0.6621021744037319,
"grad_norm": 9.179292363251772,
"learning_rate": 1.5467048645835895e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -684.0,
"logps/rejected": -1048.0,
"loss": 0.1684,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.15625,
"rewards/margins": 3.828125,
"rewards/rejected": -8.9375,
"step": 8800
},
{
"epoch": 0.6628545632382815,
"grad_norm": 9.274499894853587,
"learning_rate": 1.540637960404575e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.484375,
"logps/chosen": -652.0,
"logps/rejected": -992.0,
"loss": 0.183,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.96875,
"rewards/margins": 3.4375,
"rewards/rejected": -8.375,
"step": 8810
},
{
"epoch": 0.6636069520728313,
"grad_norm": 9.77106572903292,
"learning_rate": 1.53457767453464e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.390625,
"logps/chosen": -652.0,
"logps/rejected": -1012.0,
"loss": 0.1981,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.875,
"rewards/margins": 3.796875,
"rewards/rejected": -8.6875,
"step": 8820
},
{
"epoch": 0.6643593409073809,
"grad_norm": 9.949300022505936,
"learning_rate": 1.528524048781613e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -616.0,
"logps/rejected": -1032.0,
"loss": 0.1754,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.6875,
"rewards/margins": 4.1875,
"rewards/rejected": -8.875,
"step": 8830
},
{
"epoch": 0.6651117297419307,
"grad_norm": 8.253969377620539,
"learning_rate": 1.5224771249073787e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.515625,
"logps/chosen": -676.0,
"logps/rejected": -1072.0,
"loss": 0.183,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.1875,
"rewards/margins": 3.953125,
"rewards/rejected": -9.125,
"step": 8840
},
{
"epoch": 0.6658641185764803,
"grad_norm": 12.424443319980968,
"learning_rate": 1.5164369446275878e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.609375,
"logps/chosen": -676.0,
"logps/rejected": -1020.0,
"loss": 0.2074,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 3.71875,
"rewards/rejected": -8.8125,
"step": 8850
},
{
"epoch": 0.6666165074110301,
"grad_norm": 8.456340145386433,
"learning_rate": 1.5104035496113672e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.59375,
"logps/chosen": -656.0,
"logps/rejected": -1024.0,
"loss": 0.1718,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.84375,
"rewards/margins": 3.921875,
"rewards/rejected": -8.75,
"step": 8860
},
{
"epoch": 0.6673688962455797,
"grad_norm": 6.214576997509315,
"learning_rate": 1.5043769814810374e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.28125,
"logps/chosen": -636.0,
"logps/rejected": -1048.0,
"loss": 0.1833,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.71875,
"rewards/margins": 4.125,
"rewards/rejected": -8.875,
"step": 8870
},
{
"epoch": 0.6681212850801294,
"grad_norm": 10.767908060045015,
"learning_rate": 1.4983572818118213e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.28125,
"logps/chosen": -656.0,
"logps/rejected": -1048.0,
"loss": 0.1946,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.9375,
"rewards/margins": 3.8125,
"rewards/rejected": -8.75,
"step": 8880
},
{
"epoch": 0.6688736739146791,
"grad_norm": 10.280012486077073,
"learning_rate": 1.4923444921315575e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.40625,
"logps/chosen": -636.0,
"logps/rejected": -1040.0,
"loss": 0.1904,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.75,
"rewards/margins": 4.0,
"rewards/rejected": -8.75,
"step": 8890
},
{
"epoch": 0.6696260627492288,
"grad_norm": 6.968371247196364,
"learning_rate": 1.4863386539204167e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.484375,
"logps/chosen": -656.0,
"logps/rejected": -1012.0,
"loss": 0.1847,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.84375,
"rewards/margins": 3.71875,
"rewards/rejected": -8.5625,
"step": 8900
},
{
"epoch": 0.6703784515837785,
"grad_norm": 8.052669523242836,
"learning_rate": 1.480339808610614e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.515625,
"logps/chosen": -624.0,
"logps/rejected": -1040.0,
"loss": 0.1789,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.6875,
"rewards/margins": 4.125,
"rewards/rejected": -8.8125,
"step": 8910
},
{
"epoch": 0.6711308404183282,
"grad_norm": 11.694396467204639,
"learning_rate": 1.47434799758612e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.40625,
"logps/chosen": -668.0,
"logps/rejected": -1012.0,
"loss": 0.1953,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0,
"rewards/margins": 3.71875,
"rewards/rejected": -8.75,
"step": 8920
},
{
"epoch": 0.6718832292528779,
"grad_norm": 10.816484640860933,
"learning_rate": 1.4683632621823822e-07,
"logits/chosen": -2.5625,
"logits/rejected": -2.375,
"logps/chosen": -668.0,
"logps/rejected": -1020.0,
"loss": 0.1864,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.0,
"rewards/margins": 3.796875,
"rewards/rejected": -8.8125,
"step": 8930
},
{
"epoch": 0.6726356180874276,
"grad_norm": 10.54852159257338,
"learning_rate": 1.4623856436860322e-07,
"logits/chosen": -2.625,
"logits/rejected": -2.390625,
"logps/chosen": -640.0,
"logps/rejected": -1056.0,
"loss": 0.2048,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.875,
"rewards/margins": 3.96875,
"rewards/rejected": -8.875,
"step": 8940
},
{
"epoch": 0.6733880069219773,
"grad_norm": 8.950735778962162,
"learning_rate": 1.4564151833346072e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.421875,
"logps/chosen": -696.0,
"logps/rejected": -1064.0,
"loss": 0.2021,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.21875,
"rewards/margins": 3.84375,
"rewards/rejected": -9.0625,
"step": 8950
},
{
"epoch": 0.674140395756527,
"grad_norm": 8.68936003944014,
"learning_rate": 1.4504519223162602e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.40625,
"logps/chosen": -680.0,
"logps/rejected": -1032.0,
"loss": 0.1832,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 3.578125,
"rewards/rejected": -8.75,
"step": 8960
},
{
"epoch": 0.6748927845910767,
"grad_norm": 9.489397991424491,
"learning_rate": 1.4444959017694827e-07,
"logits/chosen": -2.640625,
"logits/rejected": -2.484375,
"logps/chosen": -668.0,
"logps/rejected": -1032.0,
"loss": 0.1752,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.03125,
"rewards/margins": 3.8125,
"rewards/rejected": -8.8125,
"step": 8970
},
{
"epoch": 0.6756451734256264,
"grad_norm": 9.236883649300923,
"learning_rate": 1.4385471627828112e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.390625,
"logps/chosen": -700.0,
"logps/rejected": -1072.0,
"loss": 0.1826,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.3125,
"rewards/margins": 3.78125,
"rewards/rejected": -9.0625,
"step": 8980
},
{
"epoch": 0.6763975622601761,
"grad_norm": 9.774332294200663,
"learning_rate": 1.4326057463945544e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.53125,
"logps/chosen": -660.0,
"logps/rejected": -1088.0,
"loss": 0.1658,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 4.125,
"rewards/rejected": -9.1875,
"step": 8990
},
{
"epoch": 0.6771499510947258,
"grad_norm": 8.167283611622954,
"learning_rate": 1.4266716935925025e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1024.0,
"loss": 0.1437,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.125,
"rewards/margins": 3.90625,
"rewards/rejected": -9.0,
"step": 9000
},
{
"epoch": 0.6779023399292754,
"grad_norm": 8.628674441098303,
"learning_rate": 1.4207450453136434e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.59375,
"logps/chosen": -704.0,
"logps/rejected": -1128.0,
"loss": 0.186,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.375,
"rewards/margins": 4.375,
"rewards/rejected": -9.75,
"step": 9010
},
{
"epoch": 0.6786547287638252,
"grad_norm": 10.650440588006587,
"learning_rate": 1.4148258424438912e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.578125,
"logps/chosen": -736.0,
"logps/rejected": -1104.0,
"loss": 0.1771,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.65625,
"rewards/margins": 3.828125,
"rewards/rejected": -9.5,
"step": 9020
},
{
"epoch": 0.6794071175983748,
"grad_norm": 8.996896689300682,
"learning_rate": 1.4089141258177894e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.5625,
"logps/chosen": -672.0,
"logps/rejected": -1048.0,
"loss": 0.1569,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.125,
"rewards/margins": 3.9375,
"rewards/rejected": -9.0625,
"step": 9030
},
{
"epoch": 0.6801595064329246,
"grad_norm": 13.995169040536673,
"learning_rate": 1.4030099362182429e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.546875,
"logps/chosen": -716.0,
"logps/rejected": -1160.0,
"loss": 0.1598,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.5625,
"rewards/margins": 4.5625,
"rewards/rejected": -10.125,
"step": 9040
},
{
"epoch": 0.6809118952674742,
"grad_norm": 10.307881060915332,
"learning_rate": 1.3971133143762255e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -696.0,
"logps/rejected": -1152.0,
"loss": 0.1824,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.375,
"rewards/margins": 4.53125,
"rewards/rejected": -9.875,
"step": 9050
},
{
"epoch": 0.681664284102024,
"grad_norm": 10.583713776190244,
"learning_rate": 1.3912243009705043e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.546875,
"logps/chosen": -752.0,
"logps/rejected": -1128.0,
"loss": 0.1965,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.875,
"rewards/margins": 3.875,
"rewards/rejected": -9.75,
"step": 9060
},
{
"epoch": 0.6824166729365736,
"grad_norm": 8.823124004557704,
"learning_rate": 1.3853429366273617e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.453125,
"logps/chosen": -700.0,
"logps/rejected": -1096.0,
"loss": 0.192,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.375,
"rewards/margins": 4.0625,
"rewards/rejected": -9.4375,
"step": 9070
},
{
"epoch": 0.6831690617711234,
"grad_norm": 11.27009671510071,
"learning_rate": 1.379469261920308e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.421875,
"logps/chosen": -704.0,
"logps/rejected": -1088.0,
"loss": 0.1684,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.3125,
"rewards/margins": 4.03125,
"rewards/rejected": -9.375,
"step": 9080
},
{
"epoch": 0.683921450605673,
"grad_norm": 7.825518749938118,
"learning_rate": 1.373603317369807e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -708.0,
"logps/rejected": -1136.0,
"loss": 0.1885,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.4375,
"rewards/margins": 4.375,
"rewards/rejected": -9.8125,
"step": 9090
},
{
"epoch": 0.6846738394402228,
"grad_norm": 8.501497320045095,
"learning_rate": 1.3677451434429945e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5625,
"logps/chosen": -676.0,
"logps/rejected": -1088.0,
"loss": 0.173,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.28125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.5,
"step": 9100
},
{
"epoch": 0.6854262282747724,
"grad_norm": 7.399791243037484,
"learning_rate": 1.3618947805533993e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.484375,
"logps/chosen": -700.0,
"logps/rejected": -1080.0,
"loss": 0.1628,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.40625,
"rewards/margins": 3.90625,
"rewards/rejected": -9.3125,
"step": 9110
},
{
"epoch": 0.6861786171093222,
"grad_norm": 7.0064300761151745,
"learning_rate": 1.3560522690606657e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.484375,
"logps/chosen": -708.0,
"logps/rejected": -1152.0,
"loss": 0.1753,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.625,
"rewards/margins": 4.4375,
"rewards/rejected": -10.0625,
"step": 9120
},
{
"epoch": 0.6869310059438718,
"grad_norm": 9.795949518702054,
"learning_rate": 1.3502176492702732e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -700.0,
"logps/rejected": -1128.0,
"loss": 0.1783,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.28125,
"rewards/margins": 4.46875,
"rewards/rejected": -9.75,
"step": 9130
},
{
"epoch": 0.6876833947784214,
"grad_norm": 11.25158699559909,
"learning_rate": 1.344390961433257e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -676.0,
"logps/rejected": -1056.0,
"loss": 0.1681,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 3.96875,
"rewards/rejected": -9.0625,
"step": 9140
},
{
"epoch": 0.6884357836129712,
"grad_norm": 6.844617722011671,
"learning_rate": 1.338572245745937e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -660.0,
"logps/rejected": -1072.0,
"loss": 0.171,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.3125,
"step": 9150
},
{
"epoch": 0.6891881724475208,
"grad_norm": 11.525771622235945,
"learning_rate": 1.3327615423496324e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.5,
"logps/chosen": -688.0,
"logps/rejected": -1080.0,
"loss": 0.1894,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.125,
"rewards/margins": 4.1875,
"rewards/rejected": -9.3125,
"step": 9160
},
{
"epoch": 0.6899405612820706,
"grad_norm": 7.667288119796913,
"learning_rate": 1.326958891330388e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.453125,
"logps/chosen": -712.0,
"logps/rejected": -1112.0,
"loss": 0.1762,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.5,
"rewards/margins": 3.890625,
"rewards/rejected": -9.375,
"step": 9170
},
{
"epoch": 0.6906929501166202,
"grad_norm": 8.78840029740877,
"learning_rate": 1.3211643327187028e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.53125,
"logps/chosen": -704.0,
"logps/rejected": -1104.0,
"loss": 0.1767,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.5,
"rewards/margins": 4.03125,
"rewards/rejected": -9.5625,
"step": 9180
},
{
"epoch": 0.69144533895117,
"grad_norm": 10.296498754282055,
"learning_rate": 1.3153779064892417e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.609375,
"logps/chosen": -692.0,
"logps/rejected": -1104.0,
"loss": 0.187,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.21875,
"rewards/margins": 4.3125,
"rewards/rejected": -9.5625,
"step": 9190
},
{
"epoch": 0.6921977277857196,
"grad_norm": 13.779489718183672,
"learning_rate": 1.309599652560574e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.578125,
"logps/chosen": -652.0,
"logps/rejected": -1088.0,
"loss": 0.1735,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.0625,
"rewards/margins": 4.3125,
"rewards/rejected": -9.375,
"step": 9200
},
{
"epoch": 0.6929501166202694,
"grad_norm": 9.16509339113319,
"learning_rate": 1.3038296107948872e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.515625,
"logps/chosen": -700.0,
"logps/rejected": -1104.0,
"loss": 0.1817,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.25,
"rewards/margins": 4.09375,
"rewards/rejected": -9.3125,
"step": 9210
},
{
"epoch": 0.693702505454819,
"grad_norm": 10.21618698882234,
"learning_rate": 1.2980678209977158e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.578125,
"logps/chosen": -676.0,
"logps/rejected": -1096.0,
"loss": 0.1713,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.1875,
"rewards/margins": 4.15625,
"rewards/rejected": -9.375,
"step": 9220
},
{
"epoch": 0.6944548942893688,
"grad_norm": 11.642858173571353,
"learning_rate": 1.2923143229176696e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.4375,
"logps/chosen": -716.0,
"logps/rejected": -1080.0,
"loss": 0.2089,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.4375,
"rewards/margins": 3.890625,
"rewards/rejected": -9.3125,
"step": 9230
},
{
"epoch": 0.6952072831239184,
"grad_norm": 7.082808348259589,
"learning_rate": 1.286569156246154e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.515625,
"logps/chosen": -716.0,
"logps/rejected": -1128.0,
"loss": 0.15,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.625,
"rewards/margins": 4.15625,
"rewards/rejected": -9.8125,
"step": 9240
},
{
"epoch": 0.6959596719584681,
"grad_norm": 8.349773826508297,
"learning_rate": 1.2808323606171006e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.515625,
"logps/chosen": -708.0,
"logps/rejected": -1080.0,
"loss": 0.1748,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -5.46875,
"rewards/margins": 4.03125,
"rewards/rejected": -9.5,
"step": 9250
},
{
"epoch": 0.6967120607930178,
"grad_norm": 9.112706785252943,
"learning_rate": 1.2751039756066907e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -692.0,
"logps/rejected": -1088.0,
"loss": 0.1645,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.1875,
"rewards/margins": 4.0625,
"rewards/rejected": -9.25,
"step": 9260
},
{
"epoch": 0.6974644496275675,
"grad_norm": 10.470465403023757,
"learning_rate": 1.2693840407330838e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.53125,
"logps/chosen": -672.0,
"logps/rejected": -1056.0,
"loss": 0.1935,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.21875,
"rewards/margins": 3.9375,
"rewards/rejected": -9.125,
"step": 9270
},
{
"epoch": 0.6982168384621172,
"grad_norm": 9.337580317905838,
"learning_rate": 1.2636725954561479e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.46875,
"logps/chosen": -680.0,
"logps/rejected": -1120.0,
"loss": 0.1863,
"rewards/accuracies": 0.9375,
"rewards/chosen": -5.03125,
"rewards/margins": 4.59375,
"rewards/rejected": -9.625,
"step": 9280
},
{
"epoch": 0.6989692272966669,
"grad_norm": 9.891066885435336,
"learning_rate": 1.257969679177177e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.421875,
"logps/chosen": -648.0,
"logps/rejected": -1096.0,
"loss": 0.1777,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -4.90625,
"rewards/margins": 4.375,
"rewards/rejected": -9.25,
"step": 9290
},
{
"epoch": 0.6997216161312166,
"grad_norm": 8.508062070892356,
"learning_rate": 1.2522753312386347e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.46875,
"logps/chosen": -664.0,
"logps/rejected": -1056.0,
"loss": 0.1799,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -4.96875,
"rewards/margins": 4.0625,
"rewards/rejected": -9.0,
"step": 9300
},
{
"epoch": 0.7004740049657663,
"grad_norm": 9.102667875750308,
"learning_rate": 1.2465895909238698e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.546875,
"logps/chosen": -680.0,
"logps/rejected": -1032.0,
"loss": 0.1791,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.0,
"rewards/margins": 3.796875,
"rewards/rejected": -8.8125,
"step": 9310
},
{
"epoch": 0.701226393800316,
"grad_norm": 11.374327168541502,
"learning_rate": 1.24091249745685e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.46875,
"logps/chosen": -656.0,
"logps/rejected": -1056.0,
"loss": 0.1778,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.84375,
"rewards/margins": 4.21875,
"rewards/rejected": -9.0625,
"step": 9320
},
{
"epoch": 0.7019787826348657,
"grad_norm": 8.211157418676647,
"learning_rate": 1.2352440900018943e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.40625,
"logps/chosen": -668.0,
"logps/rejected": -1080.0,
"loss": 0.1684,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 4.0625,
"rewards/rejected": -9.1875,
"step": 9330
},
{
"epoch": 0.7027311714694154,
"grad_norm": 10.469371653774871,
"learning_rate": 1.2295844076633973e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -660.0,
"logps/rejected": -1072.0,
"loss": 0.191,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 4.25,
"rewards/rejected": -9.3125,
"step": 9340
},
{
"epoch": 0.7034835603039651,
"grad_norm": 9.42145146901656,
"learning_rate": 1.2239334894855622e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.4375,
"logps/chosen": -636.0,
"logps/rejected": -1040.0,
"loss": 0.1921,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.84375,
"rewards/margins": 4.1875,
"rewards/rejected": -9.0,
"step": 9350
},
{
"epoch": 0.7042359491385148,
"grad_norm": 8.009475055610869,
"learning_rate": 1.2182913744521332e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.46875,
"logps/chosen": -652.0,
"logps/rejected": -992.0,
"loss": 0.1896,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -4.9375,
"rewards/margins": 3.5625,
"rewards/rejected": -8.5,
"step": 9360
},
{
"epoch": 0.7049883379730645,
"grad_norm": 10.882108464272054,
"learning_rate": 1.2126581014861227e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.53125,
"logps/chosen": -676.0,
"logps/rejected": -1024.0,
"loss": 0.1932,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 3.75,
"rewards/rejected": -8.6875,
"step": 9370
},
{
"epoch": 0.7057407268076141,
"grad_norm": 6.842288897849046,
"learning_rate": 1.207033709449545e-07,
"logits/chosen": -2.65625,
"logits/rejected": -2.484375,
"logps/chosen": -652.0,
"logps/rejected": -1016.0,
"loss": 0.1937,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.0,
"rewards/margins": 3.8125,
"rewards/rejected": -8.8125,
"step": 9380
},
{
"epoch": 0.7064931156421639,
"grad_norm": 9.393673021501025,
"learning_rate": 1.2014182371431486e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.40625,
"logps/chosen": -640.0,
"logps/rejected": -1024.0,
"loss": 0.1577,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.78125,
"rewards/margins": 4.125,
"rewards/rejected": -8.875,
"step": 9390
},
{
"epoch": 0.7072455044767135,
"grad_norm": 10.770423337068612,
"learning_rate": 1.1958117233061465e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.390625,
"logps/chosen": -664.0,
"logps/rejected": -1056.0,
"loss": 0.1626,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.9375,
"rewards/margins": 4.09375,
"rewards/rejected": -9.0625,
"step": 9400
},
{
"epoch": 0.7079978933112633,
"grad_norm": 21.808391597326604,
"learning_rate": 1.1902142066159529e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.5625,
"logps/chosen": -668.0,
"logps/rejected": -1104.0,
"loss": 0.1889,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0,
"rewards/margins": 4.3125,
"rewards/rejected": -9.3125,
"step": 9410
},
{
"epoch": 0.7087502821458129,
"grad_norm": 10.702203908885501,
"learning_rate": 1.1846257256879116e-07,
"logits/chosen": -2.796875,
"logits/rejected": -2.4375,
"logps/chosen": -680.0,
"logps/rejected": -1072.0,
"loss": 0.1845,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.15625,
"rewards/margins": 4.0625,
"rewards/rejected": -9.25,
"step": 9420
},
{
"epoch": 0.7095026709803627,
"grad_norm": 11.364598529357458,
"learning_rate": 1.1790463190750313e-07,
"logits/chosen": -2.90625,
"logits/rejected": -2.578125,
"logps/chosen": -660.0,
"logps/rejected": -1056.0,
"loss": 0.1853,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.96875,
"rewards/margins": 4.28125,
"rewards/rejected": -9.25,
"step": 9430
},
{
"epoch": 0.7102550598149123,
"grad_norm": 8.753631634329706,
"learning_rate": 1.1734760252677228e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.484375,
"logps/chosen": -652.0,
"logps/rejected": -1032.0,
"loss": 0.1669,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.84375,
"rewards/margins": 3.953125,
"rewards/rejected": -8.8125,
"step": 9440
},
{
"epoch": 0.7110074486494621,
"grad_norm": 9.94382735225419,
"learning_rate": 1.1679148826935284e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.515625,
"logps/chosen": -656.0,
"logps/rejected": -1032.0,
"loss": 0.1847,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.03125,
"rewards/margins": 3.796875,
"rewards/rejected": -8.8125,
"step": 9450
},
{
"epoch": 0.7117598374840117,
"grad_norm": 9.838493336923607,
"learning_rate": 1.1623629297168599e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.4375,
"logps/chosen": -616.0,
"logps/rejected": -1024.0,
"loss": 0.1896,
"rewards/accuracies": 0.90625,
"rewards/chosen": -4.625,
"rewards/margins": 4.09375,
"rewards/rejected": -8.6875,
"step": 9460
},
{
"epoch": 0.7125122263185615,
"grad_norm": 9.585492473960032,
"learning_rate": 1.1568202046387334e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.484375,
"logps/chosen": -664.0,
"logps/rejected": -1016.0,
"loss": 0.1807,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -5.03125,
"rewards/margins": 3.546875,
"rewards/rejected": -8.5625,
"step": 9470
},
{
"epoch": 0.7132646151531111,
"grad_norm": 9.265983416511775,
"learning_rate": 1.1512867456965036e-07,
"logits/chosen": -2.890625,
"logits/rejected": -2.59375,
"logps/chosen": -644.0,
"logps/rejected": -1056.0,
"loss": 0.1676,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.71875,
"rewards/margins": 4.34375,
"rewards/rejected": -9.0625,
"step": 9480
},
{
"epoch": 0.7140170039876608,
"grad_norm": 9.23155140930734,
"learning_rate": 1.1457625910636042e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.359375,
"logps/chosen": -668.0,
"logps/rejected": -1080.0,
"loss": 0.1809,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.125,
"rewards/margins": 4.21875,
"rewards/rejected": -9.3125,
"step": 9490
},
{
"epoch": 0.7147693928222105,
"grad_norm": 10.055006739194052,
"learning_rate": 1.1402477788492796e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.546875,
"logps/chosen": -664.0,
"logps/rejected": -1072.0,
"loss": 0.2063,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.125,
"rewards/margins": 4.0625,
"rewards/rejected": -9.1875,
"step": 9500
},
{
"epoch": 0.7155217816567602,
"grad_norm": 9.051779429722119,
"learning_rate": 1.134742347098323e-07,
"logits/chosen": -2.765625,
"logits/rejected": -2.40625,
"logps/chosen": -644.0,
"logps/rejected": -1040.0,
"loss": 0.1737,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.90625,
"rewards/margins": 4.15625,
"rewards/rejected": -9.0625,
"step": 9510
},
{
"epoch": 0.7162741704913099,
"grad_norm": 9.738489694790811,
"learning_rate": 1.1292463337908185e-07,
"logits/chosen": -2.71875,
"logits/rejected": -2.46875,
"logps/chosen": -708.0,
"logps/rejected": -1056.0,
"loss": 0.1818,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.46875,
"rewards/margins": 3.671875,
"rewards/rejected": -9.125,
"step": 9520
},
{
"epoch": 0.7170265593258596,
"grad_norm": 7.3835263007077785,
"learning_rate": 1.1237597768418714e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.390625,
"logps/chosen": -684.0,
"logps/rejected": -1040.0,
"loss": 0.1705,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 3.796875,
"rewards/rejected": -8.875,
"step": 9530
},
{
"epoch": 0.7177789481604093,
"grad_norm": 9.635982346463374,
"learning_rate": 1.1182827141013549e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.484375,
"logps/chosen": -664.0,
"logps/rejected": -1048.0,
"loss": 0.1891,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.0625,
"rewards/margins": 4.0625,
"rewards/rejected": -9.125,
"step": 9540
},
{
"epoch": 0.718531336994959,
"grad_norm": 9.078747354114716,
"learning_rate": 1.112815183353642e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.4375,
"logps/chosen": -684.0,
"logps/rejected": -1080.0,
"loss": 0.1714,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -5.15625,
"rewards/margins": 4.1875,
"rewards/rejected": -9.3125,
"step": 9550
},
{
"epoch": 0.7192837258295087,
"grad_norm": 7.906910158972716,
"learning_rate": 1.1073572223173489e-07,
"logits/chosen": -2.6875,
"logits/rejected": -2.40625,
"logps/chosen": -672.0,
"logps/rejected": -1048.0,
"loss": 0.1691,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0625,
"rewards/margins": 3.875,
"rewards/rejected": -8.9375,
"step": 9560
},
{
"epoch": 0.7200361146640584,
"grad_norm": 9.668892587581858,
"learning_rate": 1.1019088686450731e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.5,
"logps/chosen": -672.0,
"logps/rejected": -1004.0,
"loss": 0.1994,
"rewards/accuracies": 0.8812500238418579,
"rewards/chosen": -5.0625,
"rewards/margins": 3.4375,
"rewards/rejected": -8.5,
"step": 9570
},
{
"epoch": 0.7207885034986081,
"grad_norm": 10.8770164787015,
"learning_rate": 1.0964701599231341e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.546875,
"logps/chosen": -644.0,
"logps/rejected": -1056.0,
"loss": 0.1706,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.84375,
"rewards/margins": 4.25,
"rewards/rejected": -9.0625,
"step": 9580
},
{
"epoch": 0.7215408923331578,
"grad_norm": 8.388635696346707,
"learning_rate": 1.091041133671316e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.65625,
"logps/chosen": -668.0,
"logps/rejected": -1040.0,
"loss": 0.1709,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.15625,
"rewards/margins": 3.828125,
"rewards/rejected": -9.0,
"step": 9590
},
{
"epoch": 0.7222932811677075,
"grad_norm": 7.955366283806861,
"learning_rate": 1.0856218273426049e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.546875,
"logps/chosen": -628.0,
"logps/rejected": -1072.0,
"loss": 0.1787,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -4.75,
"rewards/margins": 4.6875,
"rewards/rejected": -9.4375,
"step": 9600
},
{
"epoch": 0.7230456700022572,
"grad_norm": 10.223005539532137,
"learning_rate": 1.0802122783229323e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.5,
"logps/chosen": -652.0,
"logps/rejected": -1048.0,
"loss": 0.1659,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.96875,
"rewards/margins": 4.0625,
"rewards/rejected": -9.0625,
"step": 9610
},
{
"epoch": 0.7237980588368068,
"grad_norm": 10.96860806650249,
"learning_rate": 1.0748125239309197e-07,
"logits/chosen": -2.78125,
"logits/rejected": -2.53125,
"logps/chosen": -692.0,
"logps/rejected": -1104.0,
"loss": 0.1728,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.21875,
"rewards/margins": 4.3125,
"rewards/rejected": -9.5625,
"step": 9620
},
{
"epoch": 0.7245504476713566,
"grad_norm": 8.692095566318107,
"learning_rate": 1.0694226014176167e-07,
"logits/chosen": -2.828125,
"logits/rejected": -2.4375,
"logps/chosen": -640.0,
"logps/rejected": -1056.0,
"loss": 0.1753,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.875,
"rewards/margins": 4.21875,
"rewards/rejected": -9.0625,
"step": 9630
},
{
"epoch": 0.7253028365059062,
"grad_norm": 9.805529480204843,
"learning_rate": 1.0640425479662465e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.484375,
"logps/chosen": -644.0,
"logps/rejected": -1088.0,
"loss": 0.1719,
"rewards/accuracies": 0.9624999761581421,
"rewards/chosen": -4.875,
"rewards/margins": 4.40625,
"rewards/rejected": -9.25,
"step": 9640
},
{
"epoch": 0.726055225340456,
"grad_norm": 8.46465203189475,
"learning_rate": 1.0586724006919496e-07,
"logits/chosen": -2.859375,
"logits/rejected": -2.53125,
"logps/chosen": -636.0,
"logps/rejected": -1024.0,
"loss": 0.162,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.84375,
"rewards/margins": 3.828125,
"rewards/rejected": -8.6875,
"step": 9650
},
{
"epoch": 0.7268076141750056,
"grad_norm": 8.467009438269589,
"learning_rate": 1.0533121966415257e-07,
"logits/chosen": -2.84375,
"logits/rejected": -2.578125,
"logps/chosen": -648.0,
"logps/rejected": -1032.0,
"loss": 0.1656,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.90625,
"rewards/margins": 3.984375,
"rewards/rejected": -8.875,
"step": 9660
},
{
"epoch": 0.7275600030095554,
"grad_norm": 9.939899833719256,
"learning_rate": 1.0479619727931827e-07,
"logits/chosen": -2.875,
"logits/rejected": -2.5625,
"logps/chosen": -664.0,
"logps/rejected": -1080.0,
"loss": 0.1605,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.03125,
"rewards/margins": 4.28125,
"rewards/rejected": -9.3125,
"step": 9670
},
{
"epoch": 0.728312391844105,
"grad_norm": 9.86220382562208,
"learning_rate": 1.0426217660562758e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.34375,
"logps/chosen": -676.0,
"logps/rejected": -1040.0,
"loss": 0.1803,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.09375,
"rewards/margins": 3.796875,
"rewards/rejected": -8.875,
"step": 9680
},
{
"epoch": 0.7290647806786548,
"grad_norm": 7.5288526501323005,
"learning_rate": 1.0372916132710555e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.484375,
"logps/chosen": -696.0,
"logps/rejected": -1016.0,
"loss": 0.1591,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.0625,
"rewards/margins": 3.65625,
"rewards/rejected": -8.6875,
"step": 9690
},
{
"epoch": 0.7298171695132044,
"grad_norm": 12.651261448629873,
"learning_rate": 1.031971551208416e-07,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -620.0,
"logps/rejected": -1032.0,
"loss": 0.1852,
"rewards/accuracies": 0.956250011920929,
"rewards/chosen": -4.65625,
"rewards/margins": 4.1875,
"rewards/rejected": -8.8125,
"step": 9700
},
{
"epoch": 0.7305695583477542,
"grad_norm": 6.553804927250272,
"learning_rate": 1.026661616569637e-07,
"logits/chosen": -2.671875,
"logits/rejected": -2.375,
"logps/chosen": -680.0,
"logps/rejected": -1040.0,
"loss": 0.1711,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.25,
"rewards/margins": 3.609375,
"rewards/rejected": -8.875,
"step": 9710
},
{
"epoch": 0.7313219471823038,
"grad_norm": 6.778448048748569,
"learning_rate": 1.0213618459861321e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.40625,
"logps/chosen": -652.0,
"logps/rejected": -1032.0,
"loss": 0.1722,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.0625,
"rewards/margins": 3.78125,
"rewards/rejected": -8.8125,
"step": 9720
},
{
"epoch": 0.7320743360168535,
"grad_norm": 10.188751496163695,
"learning_rate": 1.0160722760192e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.46875,
"logps/chosen": -660.0,
"logps/rejected": -1056.0,
"loss": 0.1754,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.96875,
"rewards/margins": 4.1875,
"rewards/rejected": -9.1875,
"step": 9730
},
{
"epoch": 0.7328267248514032,
"grad_norm": 9.468220885030195,
"learning_rate": 1.010792943159763e-07,
"logits/chosen": -2.8125,
"logits/rejected": -2.546875,
"logps/chosen": -700.0,
"logps/rejected": -1072.0,
"loss": 0.1825,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -5.1875,
"rewards/margins": 3.78125,
"rewards/rejected": -9.0,
"step": 9740
},
{
"epoch": 0.7335791136859529,
"grad_norm": 10.86395812337371,
"learning_rate": 1.0055238838281275e-07,
"logits/chosen": -2.75,
"logits/rejected": -2.421875,
"logps/chosen": -668.0,
"logps/rejected": -1080.0,
"loss": 0.1707,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.15625,
"rewards/margins": 3.90625,
"rewards/rejected": -9.0625,
"step": 9750
},
{
"epoch": 0.7343315025205026,
"grad_norm": 8.879896833215565,
"learning_rate": 1.000265134373722e-07,
"logits/chosen": -2.734375,
"logits/rejected": -2.453125,
"logps/chosen": -692.0,
"logps/rejected": -1032.0,
"loss": 0.1886,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.28125,
"rewards/margins": 3.546875,
"rewards/rejected": -8.8125,
"step": 9760
},
{
"epoch": 0.7350838913550523,
"grad_norm": 11.673191469930098,
"learning_rate": 9.950167310748516e-08,
"logits/chosen": -2.8125,
"logits/rejected": -2.5625,
"logps/chosen": -680.0,
"logps/rejected": -1024.0,
"loss": 0.2021,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.0,
"rewards/margins": 3.71875,
"rewards/rejected": -8.75,
"step": 9770
},
{
"epoch": 0.735836280189602,
"grad_norm": 8.480948647247656,
"learning_rate": 9.897787101384492e-08,
"logits/chosen": -2.859375,
"logits/rejected": -2.5,
"logps/chosen": -676.0,
"logps/rejected": -1072.0,
"loss": 0.1879,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 4.0,
"rewards/rejected": -9.1875,
"step": 9780
},
{
"epoch": 0.7365886690241517,
"grad_norm": 11.316343261425263,
"learning_rate": 9.845511076998195e-08,
"logits/chosen": -2.734375,
"logits/rejected": -2.5,
"logps/chosen": -652.0,
"logps/rejected": -1024.0,
"loss": 0.1941,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.9375,
"rewards/margins": 3.84375,
"rewards/rejected": -8.75,
"step": 9790
},
{
"epoch": 0.7373410578587014,
"grad_norm": 7.946997283617458,
"learning_rate": 9.79333959822397e-08,
"logits/chosen": -2.796875,
"logits/rejected": -2.421875,
"logps/chosen": -652.0,
"logps/rejected": -1088.0,
"loss": 0.1749,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 4.34375,
"rewards/rejected": -9.25,
"step": 9800
},
{
"epoch": 0.738093446693251,
"grad_norm": 11.014704537409067,
"learning_rate": 9.741273024974919e-08,
"logits/chosen": -2.78125,
"logits/rejected": -2.4375,
"logps/chosen": -700.0,
"logps/rejected": -1096.0,
"loss": 0.1855,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.4375,
"rewards/margins": 3.921875,
"rewards/rejected": -9.3125,
"step": 9810
},
{
"epoch": 0.7388458355278008,
"grad_norm": 7.359825088225063,
"learning_rate": 9.68931171644044e-08,
"logits/chosen": -2.703125,
"logits/rejected": -2.390625,
"logps/chosen": -692.0,
"logps/rejected": -1048.0,
"loss": 0.1879,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.375,
"rewards/margins": 3.609375,
"rewards/rejected": -9.0,
"step": 9820
},
{
"epoch": 0.7395982243623505,
"grad_norm": 13.258771156337893,
"learning_rate": 9.637456031083746e-08,
"logits/chosen": -2.640625,
"logits/rejected": -2.34375,
"logps/chosen": -660.0,
"logps/rejected": -1024.0,
"loss": 0.1734,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.8125,
"rewards/margins": 3.8125,
"rewards/rejected": -8.625,
"step": 9830
},
{
"epoch": 0.7403506131969002,
"grad_norm": 8.149307767061735,
"learning_rate": 9.585706326639383e-08,
"logits/chosen": -2.734375,
"logits/rejected": -2.515625,
"logps/chosen": -656.0,
"logps/rejected": -996.0,
"loss": 0.1707,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.875,
"rewards/margins": 3.578125,
"rewards/rejected": -8.4375,
"step": 9840
},
{
"epoch": 0.7411030020314499,
"grad_norm": 7.294894487594937,
"learning_rate": 9.534062960110803e-08,
"logits/chosen": -2.765625,
"logits/rejected": -2.4375,
"logps/chosen": -668.0,
"logps/rejected": -1048.0,
"loss": 0.1913,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -5.09375,
"rewards/margins": 3.828125,
"rewards/rejected": -8.9375,
"step": 9850
},
{
"epoch": 0.7418553908659995,
"grad_norm": 6.963836197185632,
"learning_rate": 9.482526287767836e-08,
"logits/chosen": -2.71875,
"logits/rejected": -2.4375,
"logps/chosen": -676.0,
"logps/rejected": -1032.0,
"loss": 0.1938,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": -5.1875,
"rewards/margins": 3.640625,
"rewards/rejected": -8.8125,
"step": 9860
},
{
"epoch": 0.7426077797005493,
"grad_norm": 10.019178889430904,
"learning_rate": 9.431096665144268e-08,
"logits/chosen": -2.703125,
"logits/rejected": -2.46875,
"logps/chosen": -712.0,
"logps/rejected": -1040.0,
"loss": 0.18,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.40625,
"rewards/margins": 3.609375,
"rewards/rejected": -9.0,
"step": 9870
},
{
"epoch": 0.7433601685350989,
"grad_norm": 11.771486703890151,
"learning_rate": 9.379774447035408e-08,
"logits/chosen": -2.9375,
"logits/rejected": -2.40625,
"logps/chosen": -656.0,
"logps/rejected": -1080.0,
"loss": 0.1924,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 4.28125,
"rewards/rejected": -9.25,
"step": 9880
},
{
"epoch": 0.7441125573696487,
"grad_norm": 8.416236345319431,
"learning_rate": 9.328559987495602e-08,
"logits/chosen": -2.8125,
"logits/rejected": -2.4375,
"logps/chosen": -648.0,
"logps/rejected": -1064.0,
"loss": 0.2017,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -4.9375,
"rewards/margins": 4.03125,
"rewards/rejected": -8.9375,
"step": 9890
},
{
"epoch": 0.7448649462041983,
"grad_norm": 9.134436384409844,
"learning_rate": 9.277453639835795e-08,
"logits/chosen": -2.796875,
"logits/rejected": -2.609375,
"logps/chosen": -700.0,
"logps/rejected": -1024.0,
"loss": 0.1766,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -5.28125,
"rewards/margins": 3.390625,
"rewards/rejected": -8.6875,
"step": 9900
},
{
"epoch": 0.745617335038748,
"grad_norm": 9.151728884375531,
"learning_rate": 9.226455756621152e-08,
"logits/chosen": -2.65625,
"logits/rejected": -2.515625,
"logps/chosen": -684.0,
"logps/rejected": -1008.0,
"loss": 0.1791,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.03125,
"rewards/margins": 3.640625,
"rewards/rejected": -8.6875,
"step": 9910
},
{
"epoch": 0.7463697238732977,
"grad_norm": 10.840220850853532,
"learning_rate": 9.175566689668506e-08,
"logits/chosen": -2.71875,
"logits/rejected": -2.40625,
"logps/chosen": -636.0,
"logps/rejected": -1024.0,
"loss": 0.2038,
"rewards/accuracies": 0.9312499761581421,
"rewards/chosen": -4.6875,
"rewards/margins": 4.09375,
"rewards/rejected": -8.75,
"step": 9920
},
{
"epoch": 0.7471221127078475,
"grad_norm": 9.358295766377482,
"learning_rate": 9.124786790044076e-08,
"logits/chosen": -2.78125,
"logits/rejected": -2.5,
"logps/chosen": -636.0,
"logps/rejected": -1072.0,
"loss": 0.1776,
"rewards/accuracies": 0.9437500238418579,
"rewards/chosen": -4.78125,
"rewards/margins": 4.40625,
"rewards/rejected": -9.1875,
"step": 9930
},
{
"epoch": 0.7478745015423971,
"grad_norm": 7.152463775059365,
"learning_rate": 9.074116408060931e-08,
"logits/chosen": -2.75,
"logits/rejected": -2.359375,
"logps/chosen": -688.0,
"logps/rejected": -1064.0,
"loss": 0.1562,
"rewards/accuracies": 0.90625,
"rewards/chosen": -5.21875,
"rewards/margins": 3.890625,
"rewards/rejected": -9.125,
"step": 9940
},
{
"epoch": 0.7486268903769469,
"grad_norm": 12.766664981308125,
"learning_rate": 9.023555893276613e-08,
"logits/chosen": -2.671875,
"logits/rejected": -2.515625,
"logps/chosen": -668.0,
"logps/rejected": -1080.0,
"loss": 0.1911,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.875,
"rewards/margins": 4.0,
"rewards/rejected": -8.875,
"step": 9950
},
{
"epoch": 0.7493792792114965,
"grad_norm": 10.858578672050218,
"learning_rate": 8.973105594490766e-08,
"logits/chosen": -2.71875,
"logits/rejected": -2.453125,
"logps/chosen": -640.0,
"logps/rejected": -1032.0,
"loss": 0.1734,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.8125,
"rewards/margins": 4.0625,
"rewards/rejected": -8.875,
"step": 9960
},
{
"epoch": 0.7501316680460463,
"grad_norm": 10.227524754185085,
"learning_rate": 8.922765859742654e-08,
"logits/chosen": -2.78125,
"logits/rejected": -2.40625,
"logps/chosen": -632.0,
"logps/rejected": -1024.0,
"loss": 0.1805,
"rewards/accuracies": 0.9375,
"rewards/chosen": -4.75,
"rewards/margins": 4.03125,
"rewards/rejected": -8.8125,
"step": 9970
},
{
"epoch": 0.7508840568805959,
"grad_norm": 7.220708688993811,
"learning_rate": 8.872537036308802e-08,
"logits/chosen": -2.71875,
"logits/rejected": -2.546875,
"logps/chosen": -680.0,
"logps/rejected": -1072.0,
"loss": 0.1596,
"rewards/accuracies": 0.893750011920929,
"rewards/chosen": -4.96875,
"rewards/margins": 4.03125,
"rewards/rejected": -9.0,
"step": 9980
},
{
"epoch": 0.7516364457151455,
"grad_norm": 8.982727497144737,
"learning_rate": 8.822419470700624e-08,
"logits/chosen": -2.84375,
"logits/rejected": -2.53125,
"logps/chosen": -644.0,
"logps/rejected": -1072.0,
"loss": 0.1833,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -4.875,
"rewards/margins": 4.34375,
"rewards/rejected": -9.1875,
"step": 9990
},
{
"epoch": 0.7523888345496953,
"grad_norm": 7.428998419527347,
"learning_rate": 8.772413508661972e-08,
"logits/chosen": -2.6875,
"logits/rejected": -2.421875,
"logps/chosen": -668.0,
"logps/rejected": -1048.0,
"loss": 0.1665,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": -5.03125,
"rewards/margins": 3.796875,
"rewards/rejected": -8.8125,
"step": 10000
},
{
"epoch": 0.7523888345496953,
"eval_logits/chosen": -2.75,
"eval_logits/rejected": -2.46875,
"eval_logps/chosen": -680.0,
"eval_logps/rejected": -1048.0,
"eval_loss": 0.23313412070274353,
"eval_rewards/accuracies": 0.8960731029510498,
"eval_rewards/chosen": -5.125,
"eval_rewards/margins": 3.78125,
"eval_rewards/rejected": -8.9375,
"eval_runtime": 3416.6122,
"eval_samples_per_second": 27.657,
"eval_steps_per_second": 0.432,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 13291,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}
Reference in New Issue View Git Blame Copy Permalink