| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.994962216624685, | |
| "eval_steps": 50, | |
| "global_step": 528, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "dpo_loss": 0.6931471824645996, | |
| "epoch": 0.003778337531486146, | |
| "grad_norm": 49.44846097246243, | |
| "learning_rate": 1.8867924528301883e-09, | |
| "logits": -1.1757584810256958, | |
| "logps": -94.8752670288086, | |
| "loss": 1.0913, | |
| "objective": 1.048144817352295, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.4375, | |
| "ranking_simple": 0.4375, | |
| "regularize": 0.3549971580505371, | |
| "step": 1 | |
| }, | |
| { | |
| "dpo_loss": 0.6931453347206116, | |
| "epoch": 0.018891687657430732, | |
| "grad_norm": 56.77495215138669, | |
| "learning_rate": 9.433962264150943e-09, | |
| "logits": -1.3608341217041016, | |
| "logps": -92.11934661865234, | |
| "loss": 1.1069, | |
| "objective": 1.06926691532135, | |
| "ranking_idealized": 0.5729166865348816, | |
| "ranking_idealized_expo": 0.4895833432674408, | |
| "ranking_simple": 0.4895833432674408, | |
| "regularize": 0.37612107396125793, | |
| "step": 5 | |
| }, | |
| { | |
| "dpo_loss": 0.693118155002594, | |
| "epoch": 0.037783375314861464, | |
| "grad_norm": 50.79305715348446, | |
| "learning_rate": 1.8867924528301887e-08, | |
| "logits": -1.3560354709625244, | |
| "logps": -90.93302917480469, | |
| "loss": 1.1118, | |
| "objective": 1.1139355897903442, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.4208172559738159, | |
| "step": 10 | |
| }, | |
| { | |
| "dpo_loss": 0.6927705407142639, | |
| "epoch": 0.05667506297229219, | |
| "grad_norm": 48.78735409468553, | |
| "learning_rate": 2.830188679245283e-08, | |
| "logits": -1.3467658758163452, | |
| "logps": -92.52275848388672, | |
| "loss": 1.1086, | |
| "objective": 1.1240794658660889, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.4313090443611145, | |
| "step": 15 | |
| }, | |
| { | |
| "dpo_loss": 0.6927476525306702, | |
| "epoch": 0.07556675062972293, | |
| "grad_norm": 48.89099638416194, | |
| "learning_rate": 3.7735849056603774e-08, | |
| "logits": -1.286866545677185, | |
| "logps": -91.98667907714844, | |
| "loss": 1.1147, | |
| "objective": 1.127905011177063, | |
| "ranking_idealized": 0.6458333134651184, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.43515732884407043, | |
| "step": 20 | |
| }, | |
| { | |
| "dpo_loss": 0.6929070353507996, | |
| "epoch": 0.09445843828715365, | |
| "grad_norm": 51.44180006710924, | |
| "learning_rate": 4.7169811320754715e-08, | |
| "logits": -1.3749586343765259, | |
| "logps": -92.29708862304688, | |
| "loss": 1.1069, | |
| "objective": 1.1106172800064087, | |
| "ranking_idealized": 0.6208333373069763, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.4177102744579315, | |
| "step": 25 | |
| }, | |
| { | |
| "dpo_loss": 0.6915452480316162, | |
| "epoch": 0.11335012594458438, | |
| "grad_norm": 52.52518235565533, | |
| "learning_rate": 5.660377358490566e-08, | |
| "logits": -1.4094866514205933, | |
| "logps": -92.03213500976562, | |
| "loss": 1.1138, | |
| "objective": 1.1041828393936157, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.4126375913619995, | |
| "step": 30 | |
| }, | |
| { | |
| "dpo_loss": 0.6903693675994873, | |
| "epoch": 0.13224181360201512, | |
| "grad_norm": 55.88065446639471, | |
| "learning_rate": 6.603773584905659e-08, | |
| "logits": -1.3921390771865845, | |
| "logps": -93.27505493164062, | |
| "loss": 1.0984, | |
| "objective": 1.1129175424575806, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.4225481450557709, | |
| "step": 35 | |
| }, | |
| { | |
| "dpo_loss": 0.6901296377182007, | |
| "epoch": 0.15113350125944586, | |
| "grad_norm": 48.11830818999703, | |
| "learning_rate": 7.547169811320755e-08, | |
| "logits": -1.4104276895523071, | |
| "logps": -92.22798156738281, | |
| "loss": 1.1021, | |
| "objective": 1.083837628364563, | |
| "ranking_idealized": 0.6625000238418579, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 0.39370766282081604, | |
| "step": 40 | |
| }, | |
| { | |
| "dpo_loss": 0.6862795352935791, | |
| "epoch": 0.17002518891687657, | |
| "grad_norm": 45.66068468167867, | |
| "learning_rate": 8.490566037735849e-08, | |
| "logits": -1.3663307428359985, | |
| "logps": -91.8579330444336, | |
| "loss": 1.1032, | |
| "objective": 1.1153314113616943, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.4290517270565033, | |
| "step": 45 | |
| }, | |
| { | |
| "dpo_loss": 0.6879932284355164, | |
| "epoch": 0.1889168765743073, | |
| "grad_norm": 51.12091223625438, | |
| "learning_rate": 9.433962264150943e-08, | |
| "logits": -1.3217811584472656, | |
| "logps": -90.97481536865234, | |
| "loss": 1.0911, | |
| "objective": 1.0859743356704712, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.3979811370372772, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1889168765743073, | |
| "eval_dpo_loss": 0.690547525882721, | |
| "eval_logits": -1.3092193603515625, | |
| "eval_logps": -97.7940902709961, | |
| "eval_loss": 1.104289174079895, | |
| "eval_objective": 1.1041598320007324, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5165745615959167, | |
| "eval_regularize": 0.4136121869087219, | |
| "eval_runtime": 235.6732, | |
| "eval_samples_per_second": 24.568, | |
| "eval_steps_per_second": 1.536, | |
| "step": 50 | |
| }, | |
| { | |
| "dpo_loss": 0.6884905695915222, | |
| "epoch": 0.20780856423173805, | |
| "grad_norm": 54.01101403233421, | |
| "learning_rate": 9.99956257238817e-08, | |
| "logits": -1.4181103706359863, | |
| "logps": -89.9561538696289, | |
| "loss": 1.1009, | |
| "objective": 1.094197154045105, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.40570658445358276, | |
| "step": 55 | |
| }, | |
| { | |
| "dpo_loss": 0.6880350708961487, | |
| "epoch": 0.22670025188916876, | |
| "grad_norm": 54.996673931606395, | |
| "learning_rate": 9.994642390694308e-08, | |
| "logits": -1.3805103302001953, | |
| "logps": -90.7548828125, | |
| "loss": 1.1093, | |
| "objective": 1.114397406578064, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.4263623058795929, | |
| "step": 60 | |
| }, | |
| { | |
| "dpo_loss": 0.6825106143951416, | |
| "epoch": 0.2455919395465995, | |
| "grad_norm": 56.45685371856582, | |
| "learning_rate": 9.98426064087682e-08, | |
| "logits": -1.4092483520507812, | |
| "logps": -92.43305969238281, | |
| "loss": 1.1012, | |
| "objective": 1.104454755783081, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.4219440817832947, | |
| "step": 65 | |
| }, | |
| { | |
| "dpo_loss": 0.6836772561073303, | |
| "epoch": 0.26448362720403024, | |
| "grad_norm": 55.79791061619149, | |
| "learning_rate": 9.968428675226714e-08, | |
| "logits": -1.372382640838623, | |
| "logps": -92.17337036132812, | |
| "loss": 1.1123, | |
| "objective": 1.1599425077438354, | |
| "ranking_idealized": 0.6291666626930237, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.476265013217926, | |
| "step": 70 | |
| }, | |
| { | |
| "dpo_loss": 0.6837269067764282, | |
| "epoch": 0.28337531486146095, | |
| "grad_norm": 47.47237853468278, | |
| "learning_rate": 9.947163805765979e-08, | |
| "logits": -1.3373608589172363, | |
| "logps": -90.7457275390625, | |
| "loss": 1.0793, | |
| "objective": 1.0441726446151733, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.4375, | |
| "ranking_simple": 0.4333333373069763, | |
| "regularize": 0.3604455888271332, | |
| "step": 75 | |
| }, | |
| { | |
| "dpo_loss": 0.6796848773956299, | |
| "epoch": 0.3022670025188917, | |
| "grad_norm": 47.94506148186036, | |
| "learning_rate": 9.920489285317169e-08, | |
| "logits": -1.4042983055114746, | |
| "logps": -91.38846588134766, | |
| "loss": 1.0778, | |
| "objective": 1.0585055351257324, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.3788203001022339, | |
| "step": 80 | |
| }, | |
| { | |
| "dpo_loss": 0.6791465282440186, | |
| "epoch": 0.3211586901763224, | |
| "grad_norm": 47.95858131470083, | |
| "learning_rate": 9.888434282076758e-08, | |
| "logits": -1.3911387920379639, | |
| "logps": -92.44884490966797, | |
| "loss": 1.0925, | |
| "objective": 1.0910677909851074, | |
| "ranking_idealized": 0.5708333253860474, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.4119212031364441, | |
| "step": 85 | |
| }, | |
| { | |
| "dpo_loss": 0.6837694644927979, | |
| "epoch": 0.34005037783375314, | |
| "grad_norm": 55.1730704885422, | |
| "learning_rate": 9.851033847720166e-08, | |
| "logits": -1.3752093315124512, | |
| "logps": -91.67774200439453, | |
| "loss": 1.0847, | |
| "objective": 1.085326075553894, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.4015560448169708, | |
| "step": 90 | |
| }, | |
| { | |
| "dpo_loss": 0.6765946745872498, | |
| "epoch": 0.3589420654911839, | |
| "grad_norm": 48.33215476425452, | |
| "learning_rate": 9.808328879073251e-08, | |
| "logits": -1.3342726230621338, | |
| "logps": -92.56273651123047, | |
| "loss": 1.0612, | |
| "objective": 1.060593605041504, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.3839987814426422, | |
| "step": 95 | |
| }, | |
| { | |
| "dpo_loss": 0.6816121935844421, | |
| "epoch": 0.3778337531486146, | |
| "grad_norm": 49.179351832428715, | |
| "learning_rate": 9.760366073392245e-08, | |
| "logits": -1.3950403928756714, | |
| "logps": -92.40387725830078, | |
| "loss": 1.0629, | |
| "objective": 1.0402370691299438, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.35862475633621216, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3778337531486146, | |
| "eval_dpo_loss": 0.6871094703674316, | |
| "eval_logits": -1.3221344947814941, | |
| "eval_logps": -98.14198303222656, | |
| "eval_loss": 1.093752384185791, | |
| "eval_objective": 1.0948374271392822, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5179557800292969, | |
| "eval_regularize": 0.4077278971672058, | |
| "eval_runtime": 235.7453, | |
| "eval_samples_per_second": 24.56, | |
| "eval_steps_per_second": 1.536, | |
| "step": 100 | |
| }, | |
| { | |
| "dpo_loss": 0.6785728931427002, | |
| "epoch": 0.3967254408060453, | |
| "grad_norm": 51.47899227122413, | |
| "learning_rate": 9.707197877300973e-08, | |
| "logits": -1.4077008962631226, | |
| "logps": -91.70052337646484, | |
| "loss": 1.0696, | |
| "objective": 1.086281180381775, | |
| "ranking_idealized": 0.612500011920929, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.407708078622818, | |
| "step": 105 | |
| }, | |
| { | |
| "dpo_loss": 0.6762667298316956, | |
| "epoch": 0.4156171284634761, | |
| "grad_norm": 50.28143364432592, | |
| "learning_rate": 9.648882429441257e-08, | |
| "logits": -1.3699510097503662, | |
| "logps": -93.08635711669922, | |
| "loss": 1.0782, | |
| "objective": 1.1022285223007202, | |
| "ranking_idealized": 0.6291666626930237, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 0.4259616434574127, | |
| "step": 110 | |
| }, | |
| { | |
| "dpo_loss": 0.6772736310958862, | |
| "epoch": 0.4345088161209068, | |
| "grad_norm": 49.35211256173949, | |
| "learning_rate": 9.585483496899149e-08, | |
| "logits": -1.4434921741485596, | |
| "logps": -91.60735321044922, | |
| "loss": 1.049, | |
| "objective": 1.0325186252593994, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 0.3552449345588684, | |
| "step": 115 | |
| }, | |
| { | |
| "dpo_loss": 0.67376708984375, | |
| "epoch": 0.4534005037783375, | |
| "grad_norm": 49.769392715448596, | |
| "learning_rate": 9.517070405476573e-08, | |
| "logits": -1.3977102041244507, | |
| "logps": -91.86192321777344, | |
| "loss": 1.0722, | |
| "objective": 1.0750477313995361, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.40128055214881897, | |
| "step": 120 | |
| }, | |
| { | |
| "dpo_loss": 0.6729015111923218, | |
| "epoch": 0.4722921914357683, | |
| "grad_norm": 48.20763298222792, | |
| "learning_rate": 9.443717963884567e-08, | |
| "logits": -1.4215410947799683, | |
| "logps": -91.61782836914062, | |
| "loss": 1.0433, | |
| "objective": 1.0424840450286865, | |
| "ranking_idealized": 0.6291666626930237, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.36958202719688416, | |
| "step": 125 | |
| }, | |
| { | |
| "dpo_loss": 0.6743431687355042, | |
| "epoch": 0.491183879093199, | |
| "grad_norm": 48.45479288893814, | |
| "learning_rate": 9.365506381941065e-08, | |
| "logits": -1.3215110301971436, | |
| "logps": -92.12106323242188, | |
| "loss": 1.0515, | |
| "objective": 1.097717046737671, | |
| "ranking_idealized": 0.625, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.4233737289905548, | |
| "step": 130 | |
| }, | |
| { | |
| "dpo_loss": 0.6695905327796936, | |
| "epoch": 0.5100755667506297, | |
| "grad_norm": 43.646349799244106, | |
| "learning_rate": 9.282521182862628e-08, | |
| "logits": -1.3272545337677002, | |
| "logps": -92.60189056396484, | |
| "loss": 1.057, | |
| "objective": 1.0827622413635254, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.41317158937454224, | |
| "step": 135 | |
| }, | |
| { | |
| "dpo_loss": 0.6709803342819214, | |
| "epoch": 0.5289672544080605, | |
| "grad_norm": 47.3960543467576, | |
| "learning_rate": 9.194853109746073e-08, | |
| "logits": -1.4292913675308228, | |
| "logps": -92.06829071044922, | |
| "loss": 1.0462, | |
| "objective": 1.046062707901001, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.3750823736190796, | |
| "step": 140 | |
| }, | |
| { | |
| "dpo_loss": 0.6690364480018616, | |
| "epoch": 0.5478589420654912, | |
| "grad_norm": 51.65656404072737, | |
| "learning_rate": 9.102598026342222e-08, | |
| "logits": -1.4306373596191406, | |
| "logps": -91.35098266601562, | |
| "loss": 1.0533, | |
| "objective": 1.0885671377182007, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5583333373069763, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.41953060030937195, | |
| "step": 145 | |
| }, | |
| { | |
| "dpo_loss": 0.6675416231155396, | |
| "epoch": 0.5667506297229219, | |
| "grad_norm": 47.60377632937882, | |
| "learning_rate": 9.005856812230304e-08, | |
| "logits": -1.4119346141815186, | |
| "logps": -90.73808288574219, | |
| "loss": 1.0597, | |
| "objective": 1.0471808910369873, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.379639208316803, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5667506297229219, | |
| "eval_dpo_loss": 0.6842523813247681, | |
| "eval_logits": -1.3387939929962158, | |
| "eval_logps": -97.57598114013672, | |
| "eval_loss": 1.0886038541793823, | |
| "eval_objective": 1.089935064315796, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.519336998462677, | |
| "eval_regularize": 0.4056825637817383, | |
| "eval_runtime": 236.1326, | |
| "eval_samples_per_second": 24.52, | |
| "eval_steps_per_second": 1.533, | |
| "step": 150 | |
| }, | |
| { | |
| "dpo_loss": 0.6675500273704529, | |
| "epoch": 0.5856423173803527, | |
| "grad_norm": 48.130868373709134, | |
| "learning_rate": 8.904735252507609e-08, | |
| "logits": -1.3894909620285034, | |
| "logps": -91.1141128540039, | |
| "loss": 1.0454, | |
| "objective": 1.0189038515090942, | |
| "ranking_idealized": 0.6333333253860474, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.3513537645339966, | |
| "step": 155 | |
| }, | |
| { | |
| "dpo_loss": 0.6642627716064453, | |
| "epoch": 0.6045340050377834, | |
| "grad_norm": 48.76010002249353, | |
| "learning_rate": 8.799343922115043e-08, | |
| "logits": -1.3890469074249268, | |
| "logps": -91.98741149902344, | |
| "loss": 1.0636, | |
| "objective": 1.0629956722259521, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.39873284101486206, | |
| "step": 160 | |
| }, | |
| { | |
| "dpo_loss": 0.6647918224334717, | |
| "epoch": 0.6234256926952141, | |
| "grad_norm": 51.709247719469836, | |
| "learning_rate": 8.689798064925048e-08, | |
| "logits": -1.430177927017212, | |
| "logps": -92.05915832519531, | |
| "loss": 1.0499, | |
| "objective": 1.05256187915802, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 0.3877698481082916, | |
| "step": 165 | |
| }, | |
| { | |
| "dpo_loss": 0.6606873273849487, | |
| "epoch": 0.6423173803526449, | |
| "grad_norm": 46.459328825125596, | |
| "learning_rate": 8.576217467724127e-08, | |
| "logits": -1.3984259366989136, | |
| "logps": -91.71315002441406, | |
| "loss": 1.029, | |
| "objective": 1.0234616994857788, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.42500001192092896, | |
| "ranking_simple": 0.44583332538604736, | |
| "regularize": 0.36277416348457336, | |
| "step": 170 | |
| }, | |
| { | |
| "dpo_loss": 0.6623333096504211, | |
| "epoch": 0.6612090680100756, | |
| "grad_norm": 51.1586426932559, | |
| "learning_rate": 8.458726329227747e-08, | |
| "logits": -1.4344308376312256, | |
| "logps": -92.77267456054688, | |
| "loss": 1.0316, | |
| "objective": 1.0282338857650757, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.36590057611465454, | |
| "step": 175 | |
| }, | |
| { | |
| "dpo_loss": 0.6674898862838745, | |
| "epoch": 0.6801007556675063, | |
| "grad_norm": 48.61836865824933, | |
| "learning_rate": 8.337453124270862e-08, | |
| "logits": -1.3953392505645752, | |
| "logps": -90.3520736694336, | |
| "loss": 1.0387, | |
| "objective": 1.0503102540969849, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.3828202784061432, | |
| "step": 180 | |
| }, | |
| { | |
| "dpo_loss": 0.6639277935028076, | |
| "epoch": 0.698992443324937, | |
| "grad_norm": 56.3868937423093, | |
| "learning_rate": 8.212530463322582e-08, | |
| "logits": -1.376102089881897, | |
| "logps": -90.30011749267578, | |
| "loss": 1.0192, | |
| "objective": 1.0141814947128296, | |
| "ranking_idealized": 0.637499988079071, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.3502536714076996, | |
| "step": 185 | |
| }, | |
| { | |
| "dpo_loss": 0.6700864434242249, | |
| "epoch": 0.7178841309823678, | |
| "grad_norm": 51.6918769313309, | |
| "learning_rate": 8.084094947478554e-08, | |
| "logits": -1.4650901556015015, | |
| "logps": -92.19945526123047, | |
| "loss": 1.0536, | |
| "objective": 1.0666062831878662, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.3965199291706085, | |
| "step": 190 | |
| }, | |
| { | |
| "dpo_loss": 0.6677178740501404, | |
| "epoch": 0.7367758186397985, | |
| "grad_norm": 51.07369943909954, | |
| "learning_rate": 7.952287019089686e-08, | |
| "logits": -1.384783387184143, | |
| "logps": -92.98999786376953, | |
| "loss": 1.0332, | |
| "objective": 1.0433076620101929, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3755895793437958, | |
| "step": 195 | |
| }, | |
| { | |
| "dpo_loss": 0.6689361333847046, | |
| "epoch": 0.7556675062972292, | |
| "grad_norm": 50.59449995506128, | |
| "learning_rate": 7.817250808190483e-08, | |
| "logits": -1.4448987245559692, | |
| "logps": -92.8608169555664, | |
| "loss": 1.0201, | |
| "objective": 0.9872502684593201, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.31831392645835876, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7556675062972292, | |
| "eval_dpo_loss": 0.6845039129257202, | |
| "eval_logits": -1.3518892526626587, | |
| "eval_logps": -98.55247497558594, | |
| "eval_loss": 1.0877532958984375, | |
| "eval_objective": 1.0928053855895996, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.519336998462677, | |
| "eval_regularize": 0.40830138325691223, | |
| "eval_runtime": 236.1466, | |
| "eval_samples_per_second": 24.519, | |
| "eval_steps_per_second": 1.533, | |
| "step": 200 | |
| }, | |
| { | |
| "dpo_loss": 0.6665075421333313, | |
| "epoch": 0.77455919395466, | |
| "grad_norm": 44.9309987154925, | |
| "learning_rate": 7.679133974894983e-08, | |
| "logits": -1.377119779586792, | |
| "logps": -91.88671112060547, | |
| "loss": 1.0103, | |
| "objective": 1.0105198621749878, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.49166667461395264, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.34401220083236694, | |
| "step": 205 | |
| }, | |
| { | |
| "dpo_loss": 0.6583543419837952, | |
| "epoch": 0.7934508816120907, | |
| "grad_norm": 51.30313684013195, | |
| "learning_rate": 7.538087547932585e-08, | |
| "logits": -1.3690478801727295, | |
| "logps": -93.26469421386719, | |
| "loss": 1.0206, | |
| "objective": 1.019605040550232, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.3612505793571472, | |
| "step": 210 | |
| }, | |
| { | |
| "dpo_loss": 0.6641567349433899, | |
| "epoch": 0.8123425692695214, | |
| "grad_norm": 50.968568724158025, | |
| "learning_rate": 7.394265759500347e-08, | |
| "logits": -1.3541967868804932, | |
| "logps": -92.59414672851562, | |
| "loss": 1.0263, | |
| "objective": 1.0428659915924072, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.3787091076374054, | |
| "step": 215 | |
| }, | |
| { | |
| "dpo_loss": 0.6569988131523132, | |
| "epoch": 0.8312342569269522, | |
| "grad_norm": 53.082634877219036, | |
| "learning_rate": 7.247825876612352e-08, | |
| "logits": -1.4405725002288818, | |
| "logps": -93.02523803710938, | |
| "loss": 1.0183, | |
| "objective": 1.0499581098556519, | |
| "ranking_idealized": 0.6458333134651184, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.3929591476917267, | |
| "step": 220 | |
| }, | |
| { | |
| "dpo_loss": 0.6647858023643494, | |
| "epoch": 0.8501259445843828, | |
| "grad_norm": 51.335148967482006, | |
| "learning_rate": 7.098928029130528e-08, | |
| "logits": -1.4285895824432373, | |
| "logps": -92.1469955444336, | |
| "loss": 1.0123, | |
| "objective": 1.034002423286438, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3692165017127991, | |
| "step": 225 | |
| }, | |
| { | |
| "dpo_loss": 0.6575040221214294, | |
| "epoch": 0.8690176322418136, | |
| "grad_norm": 48.651529158644536, | |
| "learning_rate": 6.947735034665001e-08, | |
| "logits": -1.4194341897964478, | |
| "logps": -92.3643798828125, | |
| "loss": 1.0129, | |
| "objective": 1.019044041633606, | |
| "ranking_idealized": 0.6583333611488342, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.3615398705005646, | |
| "step": 230 | |
| }, | |
| { | |
| "dpo_loss": 0.6606828570365906, | |
| "epoch": 0.8879093198992444, | |
| "grad_norm": 48.73605551669952, | |
| "learning_rate": 6.794412220535425e-08, | |
| "logits": -1.4252243041992188, | |
| "logps": -91.66600799560547, | |
| "loss": 1.0227, | |
| "objective": 1.0363646745681763, | |
| "ranking_idealized": 0.6791666746139526, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5958333611488342, | |
| "regularize": 0.375681608915329, | |
| "step": 235 | |
| }, | |
| { | |
| "dpo_loss": 0.661056399345398, | |
| "epoch": 0.906801007556675, | |
| "grad_norm": 48.248478909206035, | |
| "learning_rate": 6.639127242987987e-08, | |
| "logits": -1.4544317722320557, | |
| "logps": -92.43217468261719, | |
| "loss": 1.0019, | |
| "objective": 1.0119422674179077, | |
| "ranking_idealized": 0.5708333253860474, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.3508855998516083, | |
| "step": 240 | |
| }, | |
| { | |
| "dpo_loss": 0.6545276045799255, | |
| "epoch": 0.9256926952141058, | |
| "grad_norm": 55.62927449157849, | |
| "learning_rate": 6.482049903865769e-08, | |
| "logits": -1.3181027173995972, | |
| "logps": -92.43726348876953, | |
| "loss": 1.0099, | |
| "objective": 0.9958193302154541, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.341291606426239, | |
| "step": 245 | |
| }, | |
| { | |
| "dpo_loss": 0.6624430418014526, | |
| "epoch": 0.9445843828715366, | |
| "grad_norm": 45.52279742153591, | |
| "learning_rate": 6.323351964932908e-08, | |
| "logits": -1.363103985786438, | |
| "logps": -91.6707992553711, | |
| "loss": 1.0173, | |
| "objective": 1.0481780767440796, | |
| "ranking_idealized": 0.6083333492279053, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3857349753379822, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9445843828715366, | |
| "eval_dpo_loss": 0.6845124959945679, | |
| "eval_logits": -1.3585822582244873, | |
| "eval_logps": -99.33744812011719, | |
| "eval_loss": 1.0862815380096436, | |
| "eval_objective": 1.0919297933578491, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.519336998462677, | |
| "eval_regularize": 0.40741726756095886, | |
| "eval_runtime": 235.789, | |
| "eval_samples_per_second": 24.556, | |
| "eval_steps_per_second": 1.535, | |
| "step": 250 | |
| }, | |
| { | |
| "dpo_loss": 0.6532657742500305, | |
| "epoch": 0.9634760705289672, | |
| "grad_norm": 51.00987164026873, | |
| "learning_rate": 6.163206960055652e-08, | |
| "logits": -1.460981011390686, | |
| "logps": -94.54005432128906, | |
| "loss": 1.0046, | |
| "objective": 1.0130220651626587, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.359756201505661, | |
| "step": 255 | |
| }, | |
| { | |
| "dpo_loss": 0.6577326059341431, | |
| "epoch": 0.982367758186398, | |
| "grad_norm": 50.576016219867064, | |
| "learning_rate": 6.001790005445606e-08, | |
| "logits": -1.362568974494934, | |
| "logps": -93.19829559326172, | |
| "loss": 1.01, | |
| "objective": 0.9904889464378357, | |
| "ranking_idealized": 0.6208333373069763, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.33275625109672546, | |
| "step": 260 | |
| }, | |
| { | |
| "dpo_loss": 0.6527519226074219, | |
| "epoch": 1.0012594458438286, | |
| "grad_norm": 50.78293400345161, | |
| "learning_rate": 5.839277608172738e-08, | |
| "logits": -1.3711644411087036, | |
| "logps": -93.6723403930664, | |
| "loss": 1.0071, | |
| "objective": 1.026816487312317, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.3740644156932831, | |
| "step": 265 | |
| }, | |
| { | |
| "dpo_loss": 0.6553524136543274, | |
| "epoch": 1.0201511335012594, | |
| "grad_norm": 52.08673841750398, | |
| "learning_rate": 5.675847473157485e-08, | |
| "logits": -1.4852278232574463, | |
| "logps": -94.36783599853516, | |
| "loss": 0.9818, | |
| "objective": 0.9954887628555298, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.34013625979423523, | |
| "step": 270 | |
| }, | |
| { | |
| "dpo_loss": 0.6415101289749146, | |
| "epoch": 1.0390428211586902, | |
| "grad_norm": 44.617567180749866, | |
| "learning_rate": 5.511678308853025e-08, | |
| "logits": -1.4626718759536743, | |
| "logps": -93.63671112060547, | |
| "loss": 0.9878, | |
| "objective": 0.9779664278030396, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.33645620942115784, | |
| "step": 275 | |
| }, | |
| { | |
| "dpo_loss": 0.6500685811042786, | |
| "epoch": 1.057934508816121, | |
| "grad_norm": 47.25187912639425, | |
| "learning_rate": 5.3469496318302197e-08, | |
| "logits": -1.40652334690094, | |
| "logps": -92.5146255493164, | |
| "loss": 0.9659, | |
| "objective": 0.9707435369491577, | |
| "ranking_idealized": 0.5708333253860474, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.3206748366355896, | |
| "step": 280 | |
| }, | |
| { | |
| "dpo_loss": 0.6382409930229187, | |
| "epoch": 1.0768261964735517, | |
| "grad_norm": 47.30764627956772, | |
| "learning_rate": 5.1818415704788724e-08, | |
| "logits": -1.4587684869766235, | |
| "logps": -94.23441314697266, | |
| "loss": 0.964, | |
| "objective": 0.9715672135353088, | |
| "ranking_idealized": 0.6666666865348816, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.6083333492279053, | |
| "regularize": 0.3333263099193573, | |
| "step": 285 | |
| }, | |
| { | |
| "dpo_loss": 0.6455419063568115, | |
| "epoch": 1.0957178841309823, | |
| "grad_norm": 45.41530383871643, | |
| "learning_rate": 5.016534668039975e-08, | |
| "logits": -1.3359150886535645, | |
| "logps": -91.27457427978516, | |
| "loss": 0.9783, | |
| "objective": 0.9912244081497192, | |
| "ranking_idealized": 0.6208333373069763, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.3456825315952301, | |
| "step": 290 | |
| }, | |
| { | |
| "dpo_loss": 0.6454907655715942, | |
| "epoch": 1.114609571788413, | |
| "grad_norm": 52.660264254351745, | |
| "learning_rate": 4.8512096851843375e-08, | |
| "logits": -1.414397120475769, | |
| "logps": -92.17057037353516, | |
| "loss": 0.9657, | |
| "objective": 0.964141309261322, | |
| "ranking_idealized": 0.625, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 0.31865036487579346, | |
| "step": 295 | |
| }, | |
| { | |
| "dpo_loss": 0.6478288173675537, | |
| "epoch": 1.1335012594458438, | |
| "grad_norm": 55.18319202572926, | |
| "learning_rate": 4.686047402353433e-08, | |
| "logits": -1.482143521308899, | |
| "logps": -93.26191711425781, | |
| "loss": 0.9755, | |
| "objective": 0.9665765762329102, | |
| "ranking_idealized": 0.612500011920929, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.31874755024909973, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1335012594458438, | |
| "eval_dpo_loss": 0.6826499700546265, | |
| "eval_logits": -1.3663146495819092, | |
| "eval_logps": -99.1143798828125, | |
| "eval_loss": 1.0829436779022217, | |
| "eval_objective": 1.08933687210083, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5214088559150696, | |
| "eval_regularize": 0.40668678283691406, | |
| "eval_runtime": 235.9585, | |
| "eval_samples_per_second": 24.538, | |
| "eval_steps_per_second": 1.534, | |
| "step": 300 | |
| }, | |
| { | |
| "dpo_loss": 0.6493499875068665, | |
| "epoch": 1.1523929471032746, | |
| "grad_norm": 45.82300619637504, | |
| "learning_rate": 4.521228422078649e-08, | |
| "logits": -1.386905550956726, | |
| "logps": -92.7632064819336, | |
| "loss": 0.9773, | |
| "objective": 0.9832318425178528, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.33388179540634155, | |
| "step": 305 | |
| }, | |
| { | |
| "dpo_loss": 0.643381655216217, | |
| "epoch": 1.1712846347607053, | |
| "grad_norm": 43.431490164469025, | |
| "learning_rate": 4.3569329714950706e-08, | |
| "logits": -1.4345782995224, | |
| "logps": -92.81407928466797, | |
| "loss": 0.9734, | |
| "objective": 0.9835326671600342, | |
| "ranking_idealized": 0.625, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.3401508927345276, | |
| "step": 310 | |
| }, | |
| { | |
| "dpo_loss": 0.6410504579544067, | |
| "epoch": 1.190176322418136, | |
| "grad_norm": 47.92691613928351, | |
| "learning_rate": 4.1933407052657454e-08, | |
| "logits": -1.406593918800354, | |
| "logps": -94.20066833496094, | |
| "loss": 0.9734, | |
| "objective": 0.9850514531135559, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 0.3440004587173462, | |
| "step": 315 | |
| }, | |
| { | |
| "dpo_loss": 0.6493880748748779, | |
| "epoch": 1.2090680100755669, | |
| "grad_norm": 49.39982876479276, | |
| "learning_rate": 4.030630509131959e-08, | |
| "logits": -1.4532923698425293, | |
| "logps": -92.71504974365234, | |
| "loss": 0.9679, | |
| "objective": 0.961419939994812, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.31203168630599976, | |
| "step": 320 | |
| }, | |
| { | |
| "dpo_loss": 0.6389760971069336, | |
| "epoch": 1.2279596977329974, | |
| "grad_norm": 50.23647074169398, | |
| "learning_rate": 3.8689803043042996e-08, | |
| "logits": -1.3649462461471558, | |
| "logps": -94.10885620117188, | |
| "loss": 0.9416, | |
| "objective": 0.9290289282798767, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.29005277156829834, | |
| "step": 325 | |
| }, | |
| { | |
| "dpo_loss": 0.6403206586837769, | |
| "epoch": 1.2468513853904282, | |
| "grad_norm": 48.23351344804798, | |
| "learning_rate": 3.708566852908418e-08, | |
| "logits": -1.4762166738510132, | |
| "logps": -92.44864654541016, | |
| "loss": 0.9562, | |
| "objective": 0.942139744758606, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.30181899666786194, | |
| "step": 330 | |
| }, | |
| { | |
| "dpo_loss": 0.6417604088783264, | |
| "epoch": 1.265743073047859, | |
| "grad_norm": 54.445785967306286, | |
| "learning_rate": 3.54956556469825e-08, | |
| "logits": -1.446532130241394, | |
| "logps": -92.52754974365234, | |
| "loss": 0.9746, | |
| "objective": 0.9908974766731262, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.3491368889808655, | |
| "step": 335 | |
| }, | |
| { | |
| "dpo_loss": 0.6481935977935791, | |
| "epoch": 1.2846347607052897, | |
| "grad_norm": 43.30323866491239, | |
| "learning_rate": 3.392150305248024e-08, | |
| "logits": -1.416210651397705, | |
| "logps": -92.2144775390625, | |
| "loss": 0.9605, | |
| "objective": 0.9179424047470093, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.2697486877441406, | |
| "step": 340 | |
| }, | |
| { | |
| "dpo_loss": 0.6473791599273682, | |
| "epoch": 1.3035264483627205, | |
| "grad_norm": 46.63436114142443, | |
| "learning_rate": 3.236493205832794e-08, | |
| "logits": -1.4570471048355103, | |
| "logps": -93.01604461669922, | |
| "loss": 0.9614, | |
| "objective": 0.9653963446617126, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.4749999940395355, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 0.3180171847343445, | |
| "step": 345 | |
| }, | |
| { | |
| "dpo_loss": 0.6448249816894531, | |
| "epoch": 1.322418136020151, | |
| "grad_norm": 48.68063353112031, | |
| "learning_rate": 3.082764475205442e-08, | |
| "logits": -1.4262375831604004, | |
| "logps": -92.78704071044922, | |
| "loss": 0.9708, | |
| "objective": 0.9741830229759216, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.32935774326324463, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.322418136020151, | |
| "eval_dpo_loss": 0.6832794547080994, | |
| "eval_logits": -1.3642104864120483, | |
| "eval_logps": -99.26692962646484, | |
| "eval_loss": 1.082851767539978, | |
| "eval_objective": 1.0900707244873047, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5220994353294373, | |
| "eval_regularize": 0.4067910611629486, | |
| "eval_runtime": 235.8313, | |
| "eval_samples_per_second": 24.551, | |
| "eval_steps_per_second": 1.535, | |
| "step": 350 | |
| }, | |
| { | |
| "dpo_loss": 0.6446579098701477, | |
| "epoch": 1.341309823677582, | |
| "grad_norm": 44.218673755410634, | |
| "learning_rate": 2.9311322134758836e-08, | |
| "logits": -1.4359955787658691, | |
| "logps": -91.50603485107422, | |
| "loss": 0.9673, | |
| "objective": 0.991865873336792, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3472079634666443, | |
| "step": 355 | |
| }, | |
| { | |
| "dpo_loss": 0.6347137689590454, | |
| "epoch": 1.3602015113350125, | |
| "grad_norm": 46.29947549071951, | |
| "learning_rate": 2.7817622282960812e-08, | |
| "logits": -1.430161714553833, | |
| "logps": -94.7589340209961, | |
| "loss": 0.9565, | |
| "objective": 0.9400501251220703, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.3053361773490906, | |
| "step": 360 | |
| }, | |
| { | |
| "dpo_loss": 0.6405820250511169, | |
| "epoch": 1.3790931989924433, | |
| "grad_norm": 47.287766896204, | |
| "learning_rate": 2.6348178535517962e-08, | |
| "logits": -1.4314912557601929, | |
| "logps": -93.3971939086914, | |
| "loss": 0.9446, | |
| "objective": 0.9586126208305359, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.31803005933761597, | |
| "step": 365 | |
| }, | |
| { | |
| "dpo_loss": 0.6399816274642944, | |
| "epoch": 1.397984886649874, | |
| "grad_norm": 47.51527080377652, | |
| "learning_rate": 2.4904597707593977e-08, | |
| "logits": -1.4040166139602661, | |
| "logps": -93.57247924804688, | |
| "loss": 0.9651, | |
| "objective": 0.9659475088119507, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.32596567273139954, | |
| "step": 370 | |
| }, | |
| { | |
| "dpo_loss": 0.6420871615409851, | |
| "epoch": 1.4168765743073048, | |
| "grad_norm": 45.177183298812, | |
| "learning_rate": 2.3488458333629773e-08, | |
| "logits": -1.362334966659546, | |
| "logps": -94.35677337646484, | |
| "loss": 0.9574, | |
| "objective": 0.9662178754806519, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.4749999940395355, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.3241305351257324, | |
| "step": 375 | |
| }, | |
| { | |
| "dpo_loss": 0.6476179957389832, | |
| "epoch": 1.4357682619647356, | |
| "grad_norm": 42.51629065900929, | |
| "learning_rate": 2.21013089412392e-08, | |
| "logits": -1.4462828636169434, | |
| "logps": -93.31849670410156, | |
| "loss": 0.9686, | |
| "objective": 0.9625572562217712, | |
| "ranking_idealized": 0.6499999761581421, | |
| "ranking_idealized_expo": 0.5791666507720947, | |
| "ranking_simple": 0.6000000238418579, | |
| "regularize": 0.31493937969207764, | |
| "step": 380 | |
| }, | |
| { | |
| "dpo_loss": 0.6483107209205627, | |
| "epoch": 1.4546599496221662, | |
| "grad_norm": 50.494940681761314, | |
| "learning_rate": 2.0744666357916923e-08, | |
| "logits": -1.3696460723876953, | |
| "logps": -93.16974639892578, | |
| "loss": 0.9674, | |
| "objective": 0.9437763690948486, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 0.2954654395580292, | |
| "step": 385 | |
| }, | |
| { | |
| "dpo_loss": 0.6387519836425781, | |
| "epoch": 1.473551637279597, | |
| "grad_norm": 48.0772224935248, | |
| "learning_rate": 1.942001405240979e-08, | |
| "logits": -1.4375274181365967, | |
| "logps": -93.4893798828125, | |
| "loss": 0.951, | |
| "objective": 0.9498422145843506, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 0.3110901117324829, | |
| "step": 390 | |
| }, | |
| { | |
| "dpo_loss": 0.6425871253013611, | |
| "epoch": 1.4924433249370277, | |
| "grad_norm": 50.00168420726503, | |
| "learning_rate": 1.8128800512565513e-08, | |
| "logits": -1.4277968406677246, | |
| "logps": -92.17271423339844, | |
| "loss": 0.9714, | |
| "objective": 0.9966804385185242, | |
| "ranking_idealized": 0.6083333492279053, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.3540932834148407, | |
| "step": 395 | |
| }, | |
| { | |
| "dpo_loss": 0.6415828466415405, | |
| "epoch": 1.5113350125944585, | |
| "grad_norm": 47.59483679412131, | |
| "learning_rate": 1.6872437661432515e-08, | |
| "logits": -1.4341318607330322, | |
| "logps": -94.33455657958984, | |
| "loss": 0.968, | |
| "objective": 0.9913274049758911, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.3497445285320282, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5113350125944585, | |
| "eval_dpo_loss": 0.6832301616668701, | |
| "eval_logits": -1.3681507110595703, | |
| "eval_logps": -99.251953125, | |
| "eval_loss": 1.0828720331192017, | |
| "eval_objective": 1.090120792388916, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5214088559150696, | |
| "eval_regularize": 0.4068906307220459, | |
| "eval_runtime": 235.2584, | |
| "eval_samples_per_second": 24.611, | |
| "eval_steps_per_second": 1.539, | |
| "step": 400 | |
| }, | |
| { | |
| "dpo_loss": 0.645163357257843, | |
| "epoch": 1.5302267002518892, | |
| "grad_norm": 47.34605640737913, | |
| "learning_rate": 1.5652299313342772e-08, | |
| "logits": -1.4498835802078247, | |
| "logps": -94.09959411621094, | |
| "loss": 0.9622, | |
| "objective": 0.9742183089256287, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.3290548026561737, | |
| "step": 405 | |
| }, | |
| { | |
| "dpo_loss": 0.6364944577217102, | |
| "epoch": 1.5491183879093198, | |
| "grad_norm": 49.26429643097485, | |
| "learning_rate": 1.4469719671666041e-08, | |
| "logits": -1.3392701148986816, | |
| "logps": -93.40227508544922, | |
| "loss": 0.9596, | |
| "objective": 0.970222532749176, | |
| "ranking_idealized": 0.612500011920929, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.33372804522514343, | |
| "step": 410 | |
| }, | |
| { | |
| "dpo_loss": 0.6261176466941833, | |
| "epoch": 1.5680100755667508, | |
| "grad_norm": 47.529393740735564, | |
| "learning_rate": 1.3325991869878012e-08, | |
| "logits": -1.398348093032837, | |
| "logps": -93.3853530883789, | |
| "loss": 0.9724, | |
| "objective": 0.9390360116958618, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.4583333432674408, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.3129182457923889, | |
| "step": 415 | |
| }, | |
| { | |
| "dpo_loss": 0.6377049088478088, | |
| "epoch": 1.5869017632241813, | |
| "grad_norm": 57.43753774126503, | |
| "learning_rate": 1.222236655753791e-08, | |
| "logits": -1.4465179443359375, | |
| "logps": -92.30918884277344, | |
| "loss": 0.9625, | |
| "objective": 0.9549995064735413, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.49166667461395264, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.31729447841644287, | |
| "step": 420 | |
| }, | |
| { | |
| "dpo_loss": 0.6409914493560791, | |
| "epoch": 1.605793450881612, | |
| "grad_norm": 48.69171129272466, | |
| "learning_rate": 1.1160050532721526e-08, | |
| "logits": -1.4412765502929688, | |
| "logps": -92.86591339111328, | |
| "loss": 0.9543, | |
| "objective": 0.9196628928184509, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.46666666865348816, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.27867138385772705, | |
| "step": 425 | |
| }, | |
| { | |
| "dpo_loss": 0.6387853622436523, | |
| "epoch": 1.6246851385390428, | |
| "grad_norm": 45.275230591779184, | |
| "learning_rate": 1.0140205422405212e-08, | |
| "logits": -1.4535937309265137, | |
| "logps": -92.8780517578125, | |
| "loss": 0.9509, | |
| "objective": 0.9649083018302917, | |
| "ranking_idealized": 0.6208333373069763, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.3261227011680603, | |
| "step": 430 | |
| }, | |
| { | |
| "dpo_loss": 0.6288203001022339, | |
| "epoch": 1.6435768261964736, | |
| "grad_norm": 44.1758284644859, | |
| "learning_rate": 9.163946412243894e-09, | |
| "logits": -1.4540935754776, | |
| "logps": -93.5007553100586, | |
| "loss": 0.9616, | |
| "objective": 0.9557725191116333, | |
| "ranking_idealized": 0.6416666507720947, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.32695215940475464, | |
| "step": 435 | |
| }, | |
| { | |
| "dpo_loss": 0.6319211721420288, | |
| "epoch": 1.6624685138539044, | |
| "grad_norm": 47.357355663292644, | |
| "learning_rate": 8.232341027131883e-09, | |
| "logits": -1.4391558170318604, | |
| "logps": -92.56800079345703, | |
| "loss": 0.9484, | |
| "objective": 0.9231205582618713, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.2911991775035858, | |
| "step": 440 | |
| }, | |
| { | |
| "dpo_loss": 0.6425955295562744, | |
| "epoch": 1.681360201511335, | |
| "grad_norm": 46.443814611627765, | |
| "learning_rate": 7.346407963880136e-09, | |
| "logits": -1.4678118228912354, | |
| "logps": -92.27074432373047, | |
| "loss": 0.9664, | |
| "objective": 0.949626624584198, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.3070310652256012, | |
| "step": 445 | |
| }, | |
| { | |
| "dpo_loss": 0.6390999555587769, | |
| "epoch": 1.700251889168766, | |
| "grad_norm": 49.0414360352724, | |
| "learning_rate": 6.507115977286143e-09, | |
| "logits": -1.393315076828003, | |
| "logps": -93.02438354492188, | |
| "loss": 0.9495, | |
| "objective": 0.9575169682502747, | |
| "ranking_idealized": 0.6083333492279053, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.31841692328453064, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.700251889168766, | |
| "eval_dpo_loss": 0.6830143928527832, | |
| "eval_logits": -1.3688236474990845, | |
| "eval_logps": -99.32157135009766, | |
| "eval_loss": 1.0825797319412231, | |
| "eval_objective": 1.0900212526321411, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5214088559150696, | |
| "eval_regularize": 0.4070066511631012, | |
| "eval_runtime": 235.5467, | |
| "eval_samples_per_second": 24.581, | |
| "eval_steps_per_second": 1.537, | |
| "step": 450 | |
| }, | |
| { | |
| "dpo_loss": 0.6333620548248291, | |
| "epoch": 1.7191435768261965, | |
| "grad_norm": 51.27723544337059, | |
| "learning_rate": 5.715382820814885e-09, | |
| "logits": -1.463624358177185, | |
| "logps": -95.06629180908203, | |
| "loss": 0.9525, | |
| "objective": 0.981126070022583, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.34776392579078674, | |
| "step": 455 | |
| }, | |
| { | |
| "dpo_loss": 0.6484085321426392, | |
| "epoch": 1.7380352644836272, | |
| "grad_norm": 46.95616030148672, | |
| "learning_rate": 4.972074243048896e-09, | |
| "logits": -1.4587913751602173, | |
| "logps": -91.8313217163086, | |
| "loss": 0.9569, | |
| "objective": 0.9492828845977783, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.30087435245513916, | |
| "step": 460 | |
| }, | |
| { | |
| "dpo_loss": 0.6343809962272644, | |
| "epoch": 1.756926952141058, | |
| "grad_norm": 47.70310058715592, | |
| "learning_rate": 4.278003041004779e-09, | |
| "logits": -1.3960585594177246, | |
| "logps": -92.12299346923828, | |
| "loss": 0.9624, | |
| "objective": 0.9703618884086609, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.33598068356513977, | |
| "step": 465 | |
| }, | |
| { | |
| "dpo_loss": 0.6355936527252197, | |
| "epoch": 1.7758186397984885, | |
| "grad_norm": 48.36448652264405, | |
| "learning_rate": 3.63392817135173e-09, | |
| "logits": -1.4002528190612793, | |
| "logps": -92.8111343383789, | |
| "loss": 0.9635, | |
| "objective": 0.9694804549217224, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3338867127895355, | |
| "step": 470 | |
| }, | |
| { | |
| "dpo_loss": 0.6456637978553772, | |
| "epoch": 1.7947103274559195, | |
| "grad_norm": 48.611216917206725, | |
| "learning_rate": 3.0405539205035023e-09, | |
| "logits": -1.4299951791763306, | |
| "logps": -92.8174819946289, | |
| "loss": 0.9564, | |
| "objective": 0.9828375577926636, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.3371736407279968, | |
| "step": 475 | |
| }, | |
| { | |
| "dpo_loss": 0.6420384645462036, | |
| "epoch": 1.81360201511335, | |
| "grad_norm": 46.76089409325827, | |
| "learning_rate": 2.4985291344915673e-09, | |
| "logits": -1.4936844110488892, | |
| "logps": -93.58541870117188, | |
| "loss": 0.967, | |
| "objective": 0.9691051244735718, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.4583333432674408, | |
| "ranking_simple": 0.46666666865348816, | |
| "regularize": 0.3270666301250458, | |
| "step": 480 | |
| }, | |
| { | |
| "dpo_loss": 0.6282381415367126, | |
| "epoch": 1.8324937027707808, | |
| "grad_norm": 52.789969586082115, | |
| "learning_rate": 2.0084465094614976e-09, | |
| "logits": -1.3527694940567017, | |
| "logps": -94.94033813476562, | |
| "loss": 0.9721, | |
| "objective": 0.9742304682731628, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.34599223732948303, | |
| "step": 485 | |
| }, | |
| { | |
| "dpo_loss": 0.6374282836914062, | |
| "epoch": 1.8513853904282116, | |
| "grad_norm": 47.2028496146643, | |
| "learning_rate": 1.570841943568446e-09, | |
| "logits": -1.430949091911316, | |
| "logps": -93.16008758544922, | |
| "loss": 0.9587, | |
| "objective": 0.9563023447990417, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3188740313053131, | |
| "step": 490 | |
| }, | |
| { | |
| "dpo_loss": 0.6446228623390198, | |
| "epoch": 1.8702770780856424, | |
| "grad_norm": 47.63126102742745, | |
| "learning_rate": 1.1861939509803687e-09, | |
| "logits": -1.4252774715423584, | |
| "logps": -93.47933959960938, | |
| "loss": 0.9498, | |
| "objective": 0.9525411128997803, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.30791813135147095, | |
| "step": 495 | |
| }, | |
| { | |
| "dpo_loss": 0.6408995985984802, | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 47.540323720333014, | |
| "learning_rate": 8.54923138629815e-10, | |
| "logits": -1.4415290355682373, | |
| "logps": -92.98784637451172, | |
| "loss": 0.9463, | |
| "objective": 0.9464013576507568, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.3055017292499542, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "eval_dpo_loss": 0.6829814314842224, | |
| "eval_logits": -1.3685972690582275, | |
| "eval_logps": -99.3604736328125, | |
| "eval_loss": 1.0826866626739502, | |
| "eval_objective": 1.0901830196380615, | |
| "eval_ranking_idealized": 0.5925414562225342, | |
| "eval_ranking_idealized_expo": 0.5165745615959167, | |
| "eval_ranking_simple": 0.5220994353294373, | |
| "eval_regularize": 0.40720152854919434, | |
| "eval_runtime": 235.7418, | |
| "eval_samples_per_second": 24.561, | |
| "eval_steps_per_second": 1.536, | |
| "step": 500 | |
| }, | |
| { | |
| "dpo_loss": 0.6330133080482483, | |
| "epoch": 1.9080604534005037, | |
| "grad_norm": 44.27516380619689, | |
| "learning_rate": 5.773917462864264e-10, | |
| "logits": -1.3644249439239502, | |
| "logps": -91.88590240478516, | |
| "loss": 0.954, | |
| "objective": 0.951960027217865, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.3189466595649719, | |
| "step": 505 | |
| }, | |
| { | |
| "dpo_loss": 0.6388721466064453, | |
| "epoch": 1.9269521410579347, | |
| "grad_norm": 51.31132661100108, | |
| "learning_rate": 3.53903250453047e-10, | |
| "logits": -1.4425408840179443, | |
| "logps": -95.0220718383789, | |
| "loss": 0.9495, | |
| "objective": 0.9448812007904053, | |
| "ranking_idealized": 0.6625000238418579, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.6000000238418579, | |
| "regularize": 0.30600884556770325, | |
| "step": 510 | |
| }, | |
| { | |
| "dpo_loss": 0.638128936290741, | |
| "epoch": 1.9458438287153652, | |
| "grad_norm": 49.79971218067957, | |
| "learning_rate": 1.8470203251865768e-10, | |
| "logits": -1.4008522033691406, | |
| "logps": -95.15757751464844, | |
| "loss": 0.9567, | |
| "objective": 0.9624292850494385, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.324300080537796, | |
| "step": 515 | |
| }, | |
| { | |
| "dpo_loss": 0.6457895636558533, | |
| "epoch": 1.964735516372796, | |
| "grad_norm": 55.67667071934356, | |
| "learning_rate": 6.997311153086882e-11, | |
| "logits": -1.464538335800171, | |
| "logps": -93.64386749267578, | |
| "loss": 0.9495, | |
| "objective": 0.9239124059677124, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5583333373069763, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 0.2781226336956024, | |
| "step": 520 | |
| }, | |
| { | |
| "dpo_loss": 0.6334381103515625, | |
| "epoch": 1.9836272040302267, | |
| "grad_norm": 50.356520924506064, | |
| "learning_rate": 9.841941880361915e-12, | |
| "logits": -1.4272739887237549, | |
| "logps": -93.79329681396484, | |
| "loss": 0.966, | |
| "objective": 0.9610148072242737, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.3275766670703888, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.994962216624685, | |
| "step": 528, | |
| "total_flos": 0.0, | |
| "train_loss": 1.010300681672313, | |
| "train_runtime": 13850.696, | |
| "train_samples_per_second": 7.336, | |
| "train_steps_per_second": 0.038 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 528, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |