| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9981298423724285, | |
| "eval_steps": 400, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010686615014694095, | |
| "grad_norm": 56.81669148706694, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0553786754608154, | |
| "logits/rejected": -0.9845818281173706, | |
| "logps/chosen": -0.7874154448509216, | |
| "logps/rejected": -0.7125983834266663, | |
| "loss": 3.9428, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -7.874154567718506, | |
| "rewards/margins": -0.7481697797775269, | |
| "rewards/rejected": -7.125983238220215, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137323002938819, | |
| "grad_norm": 90.64309336283279, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -0.9677240252494812, | |
| "logits/rejected": -0.9489688873291016, | |
| "logps/chosen": -0.8535294532775879, | |
| "logps/rejected": -1.0613572597503662, | |
| "loss": 3.6689, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -8.535294532775879, | |
| "rewards/margins": 2.0782790184020996, | |
| "rewards/rejected": -10.613574028015137, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03205984504408229, | |
| "grad_norm": 110.04849731267365, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.9973424673080444, | |
| "logits/rejected": -0.9682930111885071, | |
| "logps/chosen": -0.8421398401260376, | |
| "logps/rejected": -0.9425029754638672, | |
| "loss": 3.9663, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -8.42139720916748, | |
| "rewards/margins": 1.0036319494247437, | |
| "rewards/rejected": -9.425029754638672, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04274646005877638, | |
| "grad_norm": 89.89735367257849, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -1.0042946338653564, | |
| "logits/rejected": -0.9698360562324524, | |
| "logps/chosen": -0.6765571236610413, | |
| "logps/rejected": -0.7212239503860474, | |
| "loss": 3.798, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -6.765571594238281, | |
| "rewards/margins": 0.44666871428489685, | |
| "rewards/rejected": -7.212240695953369, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053433075073470476, | |
| "grad_norm": 40.5523084584386, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.042263388633728, | |
| "logits/rejected": -1.020334005355835, | |
| "logps/chosen": -0.6275808215141296, | |
| "logps/rejected": -0.6092665195465088, | |
| "loss": 3.5027, | |
| "rewards/accuracies": 0.38749998807907104, | |
| "rewards/chosen": -6.275807857513428, | |
| "rewards/margins": -0.18314316868782043, | |
| "rewards/rejected": -6.092665195465088, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06411969008816458, | |
| "grad_norm": 82.97126019183185, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -1.0934265851974487, | |
| "logits/rejected": -1.0273423194885254, | |
| "logps/chosen": -0.4222745895385742, | |
| "logps/rejected": -0.4437929689884186, | |
| "loss": 3.2915, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -4.2227463722229, | |
| "rewards/margins": 0.21518389880657196, | |
| "rewards/rejected": -4.437930107116699, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07480630510285867, | |
| "grad_norm": 45.671980780198346, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.1021089553833008, | |
| "logits/rejected": -1.016642689704895, | |
| "logps/chosen": -0.40351754426956177, | |
| "logps/rejected": -0.4182681143283844, | |
| "loss": 3.0336, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -4.035175800323486, | |
| "rewards/margins": 0.14750559628009796, | |
| "rewards/rejected": -4.182681083679199, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08549292011755276, | |
| "grad_norm": 56.834561606796534, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -1.0844147205352783, | |
| "logits/rejected": -1.0052152872085571, | |
| "logps/chosen": -0.3840138018131256, | |
| "logps/rejected": -0.41540414094924927, | |
| "loss": 2.9873, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -3.8401379585266113, | |
| "rewards/margins": 0.3139037489891052, | |
| "rewards/rejected": -4.154041290283203, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09617953513224686, | |
| "grad_norm": 57.81609876406988, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.0846834182739258, | |
| "logits/rejected": -0.9838830828666687, | |
| "logps/chosen": -0.3058851957321167, | |
| "logps/rejected": -0.3447844386100769, | |
| "loss": 2.9473, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -3.058851957321167, | |
| "rewards/margins": 0.3889924883842468, | |
| "rewards/rejected": -3.4478442668914795, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10686615014694095, | |
| "grad_norm": 107.63293409046761, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.183288335800171, | |
| "logits/rejected": -1.0858242511749268, | |
| "logps/chosen": -0.3091445863246918, | |
| "logps/rejected": -0.3627066910266876, | |
| "loss": 2.9167, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -3.0914459228515625, | |
| "rewards/margins": 0.5356208682060242, | |
| "rewards/rejected": -3.6270668506622314, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11755276516163506, | |
| "grad_norm": 40.363889487849875, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.1698521375656128, | |
| "logits/rejected": -1.1601616144180298, | |
| "logps/chosen": -0.29625993967056274, | |
| "logps/rejected": -0.33069461584091187, | |
| "loss": 2.9262, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -2.962599515914917, | |
| "rewards/margins": 0.34434670209884644, | |
| "rewards/rejected": -3.306946277618408, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12823938017632916, | |
| "grad_norm": 49.68454575475429, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.2926113605499268, | |
| "logits/rejected": -1.2354081869125366, | |
| "logps/chosen": -0.34236031770706177, | |
| "logps/rejected": -0.370178759098053, | |
| "loss": 2.828, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -3.423603057861328, | |
| "rewards/margins": 0.2781844735145569, | |
| "rewards/rejected": -3.7017874717712402, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13892599519102325, | |
| "grad_norm": 64.97695318174799, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -1.2277723550796509, | |
| "logits/rejected": -1.1402448415756226, | |
| "logps/chosen": -0.33444857597351074, | |
| "logps/rejected": -0.38238364458084106, | |
| "loss": 2.8768, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -3.3444857597351074, | |
| "rewards/margins": 0.47935038805007935, | |
| "rewards/rejected": -3.8238365650177, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14961261020571734, | |
| "grad_norm": 42.426764348756556, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.2491295337677002, | |
| "logits/rejected": -1.2098782062530518, | |
| "logps/chosen": -0.2992488443851471, | |
| "logps/rejected": -0.34425443410873413, | |
| "loss": 2.8068, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.992488145828247, | |
| "rewards/margins": 0.45005664229393005, | |
| "rewards/rejected": -3.442544460296631, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16029922522041143, | |
| "grad_norm": 55.41018623107141, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -1.224936604499817, | |
| "logits/rejected": -1.1226717233657837, | |
| "logps/chosen": -0.3549841642379761, | |
| "logps/rejected": -0.42627596855163574, | |
| "loss": 2.6663, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -3.5498414039611816, | |
| "rewards/margins": 0.7129180431365967, | |
| "rewards/rejected": -4.262759685516357, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17098584023510552, | |
| "grad_norm": 67.19768319790357, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -1.22861647605896, | |
| "logits/rejected": -1.173981785774231, | |
| "logps/chosen": -0.35033124685287476, | |
| "logps/rejected": -0.3842136263847351, | |
| "loss": 2.6237, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -3.503312349319458, | |
| "rewards/margins": 0.3388238847255707, | |
| "rewards/rejected": -3.8421363830566406, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18167245524979964, | |
| "grad_norm": 56.06934270502333, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -1.1919827461242676, | |
| "logits/rejected": -1.145477056503296, | |
| "logps/chosen": -0.3422418534755707, | |
| "logps/rejected": -0.4096073508262634, | |
| "loss": 2.6001, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -3.4224185943603516, | |
| "rewards/margins": 0.6736553311347961, | |
| "rewards/rejected": -4.096073627471924, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19235907026449373, | |
| "grad_norm": 66.32485026269077, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.2551718950271606, | |
| "logits/rejected": -1.1607410907745361, | |
| "logps/chosen": -0.39799433946609497, | |
| "logps/rejected": -0.47051042318344116, | |
| "loss": 2.6013, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -3.9799437522888184, | |
| "rewards/margins": 0.7251602411270142, | |
| "rewards/rejected": -4.705103397369385, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20304568527918782, | |
| "grad_norm": 88.72645076406972, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.202864646911621, | |
| "logits/rejected": -1.0979254245758057, | |
| "logps/chosen": -0.40985170006752014, | |
| "logps/rejected": -0.4939080774784088, | |
| "loss": 2.47, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -4.098517417907715, | |
| "rewards/margins": 0.8405635952949524, | |
| "rewards/rejected": -4.939080238342285, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2137323002938819, | |
| "grad_norm": 78.62356168132955, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -1.210082769393921, | |
| "logits/rejected": -1.19521963596344, | |
| "logps/chosen": -0.4767809808254242, | |
| "logps/rejected": -0.6341355443000793, | |
| "loss": 2.4364, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -4.767809867858887, | |
| "rewards/margins": 1.5735454559326172, | |
| "rewards/rejected": -6.341355323791504, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.224418915308576, | |
| "grad_norm": 182.17113716520097, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -1.318027138710022, | |
| "logits/rejected": -1.2231770753860474, | |
| "logps/chosen": -0.5009247064590454, | |
| "logps/rejected": -0.6060206890106201, | |
| "loss": 2.4925, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -5.009247779846191, | |
| "rewards/margins": 1.050958514213562, | |
| "rewards/rejected": -6.060206413269043, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2351055303232701, | |
| "grad_norm": 97.52898918968008, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -1.3112456798553467, | |
| "logits/rejected": -1.2372304201126099, | |
| "logps/chosen": -0.4988152086734772, | |
| "logps/rejected": -0.6063084602355957, | |
| "loss": 2.4096, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -4.988151550292969, | |
| "rewards/margins": 1.074933409690857, | |
| "rewards/rejected": -6.063085079193115, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2457921453379642, | |
| "grad_norm": 65.2666290162174, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.2346036434173584, | |
| "logits/rejected": -1.155144453048706, | |
| "logps/chosen": -0.46027320623397827, | |
| "logps/rejected": -0.539958655834198, | |
| "loss": 2.4476, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -4.602732181549072, | |
| "rewards/margins": 0.7968543767929077, | |
| "rewards/rejected": -5.3995866775512695, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2564787603526583, | |
| "grad_norm": 94.23205420362748, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -1.2661981582641602, | |
| "logits/rejected": -1.2356774806976318, | |
| "logps/chosen": -0.5204485058784485, | |
| "logps/rejected": -0.6347986459732056, | |
| "loss": 2.313, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -5.204484939575195, | |
| "rewards/margins": 1.1435019969940186, | |
| "rewards/rejected": -6.347987174987793, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2671653753673524, | |
| "grad_norm": 104.425486907994, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.4145832061767578, | |
| "logits/rejected": -1.3696868419647217, | |
| "logps/chosen": -0.5677964687347412, | |
| "logps/rejected": -0.7819775342941284, | |
| "loss": 2.0573, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -5.677964210510254, | |
| "rewards/margins": 2.1418113708496094, | |
| "rewards/rejected": -7.8197760581970215, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2778519903820465, | |
| "grad_norm": 118.15756723487034, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.2262728214263916, | |
| "logits/rejected": -1.1968516111373901, | |
| "logps/chosen": -0.5714520812034607, | |
| "logps/rejected": -0.7978314757347107, | |
| "loss": 2.229, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -5.7145209312438965, | |
| "rewards/margins": 2.2637939453125, | |
| "rewards/rejected": -7.9783148765563965, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2885386053967406, | |
| "grad_norm": 71.01619814345214, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.423404574394226, | |
| "logits/rejected": -1.3302305936813354, | |
| "logps/chosen": -0.6018441915512085, | |
| "logps/rejected": -0.7836464047431946, | |
| "loss": 1.9147, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -6.018441677093506, | |
| "rewards/margins": 1.8180221319198608, | |
| "rewards/rejected": -7.836463928222656, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2992252204114347, | |
| "grad_norm": 66.78115766091884, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.384155511856079, | |
| "logits/rejected": -1.3765170574188232, | |
| "logps/chosen": -0.6419236660003662, | |
| "logps/rejected": -0.8896235227584839, | |
| "loss": 1.9382, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -6.419236660003662, | |
| "rewards/margins": 2.476998805999756, | |
| "rewards/rejected": -8.896234512329102, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30991183542612877, | |
| "grad_norm": 102.33456775014218, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.3982570171356201, | |
| "logits/rejected": -1.3454999923706055, | |
| "logps/chosen": -0.6346415281295776, | |
| "logps/rejected": -0.8345499038696289, | |
| "loss": 1.8347, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -6.346415042877197, | |
| "rewards/margins": 1.9990848302841187, | |
| "rewards/rejected": -8.345499038696289, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32059845044082286, | |
| "grad_norm": 104.88848048977833, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.3800594806671143, | |
| "logits/rejected": -1.3424351215362549, | |
| "logps/chosen": -0.7959817051887512, | |
| "logps/rejected": -1.088423252105713, | |
| "loss": 1.9956, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -7.959817409515381, | |
| "rewards/margins": 2.9244155883789062, | |
| "rewards/rejected": -10.884233474731445, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33128506545551695, | |
| "grad_norm": 86.74090140921074, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.289271354675293, | |
| "logits/rejected": -1.221858263015747, | |
| "logps/chosen": -0.7765235900878906, | |
| "logps/rejected": -1.0151869058609009, | |
| "loss": 1.905, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -7.76523494720459, | |
| "rewards/margins": 2.386632204055786, | |
| "rewards/rejected": -10.15186882019043, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34197168047021104, | |
| "grad_norm": 103.52900771453592, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.3143701553344727, | |
| "logits/rejected": -1.2429687976837158, | |
| "logps/chosen": -0.8146079182624817, | |
| "logps/rejected": -1.076160192489624, | |
| "loss": 1.9347, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -8.146079063415527, | |
| "rewards/margins": 2.6155219078063965, | |
| "rewards/rejected": -10.761602401733398, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3526582954849052, | |
| "grad_norm": 91.78991869467103, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.277517557144165, | |
| "logits/rejected": -1.2467105388641357, | |
| "logps/chosen": -0.8164845705032349, | |
| "logps/rejected": -1.0702334642410278, | |
| "loss": 1.8253, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -8.16484546661377, | |
| "rewards/margins": 2.5374886989593506, | |
| "rewards/rejected": -10.7023344039917, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.36334491049959927, | |
| "grad_norm": 146.18164794432158, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.4038187265396118, | |
| "logits/rejected": -1.32179856300354, | |
| "logps/chosen": -0.9452447891235352, | |
| "logps/rejected": -1.285444974899292, | |
| "loss": 1.8459, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -9.452447891235352, | |
| "rewards/margins": 3.4020028114318848, | |
| "rewards/rejected": -12.854451179504395, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37403152551429336, | |
| "grad_norm": 74.04178740704424, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.2986555099487305, | |
| "logits/rejected": -1.2864700555801392, | |
| "logps/chosen": -0.8545284271240234, | |
| "logps/rejected": -1.2237508296966553, | |
| "loss": 1.7393, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -8.545283317565918, | |
| "rewards/margins": 3.6922249794006348, | |
| "rewards/rejected": -12.237508773803711, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38471814052898745, | |
| "grad_norm": 146.93436823237963, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.4434401988983154, | |
| "logits/rejected": -1.3974043130874634, | |
| "logps/chosen": -1.084814190864563, | |
| "logps/rejected": -1.4313297271728516, | |
| "loss": 1.5783, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -10.84814167022705, | |
| "rewards/margins": 3.4651541709899902, | |
| "rewards/rejected": -14.3132963180542, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39540475554368154, | |
| "grad_norm": 300.3576490829542, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.331994891166687, | |
| "logits/rejected": -1.3323677778244019, | |
| "logps/chosen": -1.0099122524261475, | |
| "logps/rejected": -1.400909185409546, | |
| "loss": 1.6356, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -10.099123001098633, | |
| "rewards/margins": 3.9099678993225098, | |
| "rewards/rejected": -14.0090913772583, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.40609137055837563, | |
| "grad_norm": 117.205754582783, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.3903909921646118, | |
| "logits/rejected": -1.395379662513733, | |
| "logps/chosen": -0.9584177136421204, | |
| "logps/rejected": -1.3904134035110474, | |
| "loss": 1.5868, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -9.584177017211914, | |
| "rewards/margins": 4.319957256317139, | |
| "rewards/rejected": -13.904134750366211, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4167779855730697, | |
| "grad_norm": 193.5248033898295, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.479113221168518, | |
| "logits/rejected": -1.431293249130249, | |
| "logps/chosen": -1.1905192136764526, | |
| "logps/rejected": -1.5327824354171753, | |
| "loss": 1.4974, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -11.905191421508789, | |
| "rewards/margins": 3.4226326942443848, | |
| "rewards/rejected": -15.327824592590332, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4274646005877638, | |
| "grad_norm": 72.90690697186486, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.464138388633728, | |
| "logits/rejected": -1.4378821849822998, | |
| "logps/chosen": -1.2057116031646729, | |
| "logps/rejected": -1.6115760803222656, | |
| "loss": 1.6772, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -12.05711555480957, | |
| "rewards/margins": 4.058643341064453, | |
| "rewards/rejected": -16.115758895874023, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4381512156024579, | |
| "grad_norm": 73.64255724761823, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.463689923286438, | |
| "logits/rejected": -1.4153610467910767, | |
| "logps/chosen": -0.9573764801025391, | |
| "logps/rejected": -1.4560494422912598, | |
| "loss": 1.5083, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -9.57376480102539, | |
| "rewards/margins": 4.986730098724365, | |
| "rewards/rejected": -14.560495376586914, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.448837830617152, | |
| "grad_norm": 95.73515995991542, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.5166422128677368, | |
| "logits/rejected": -1.5037015676498413, | |
| "logps/chosen": -1.1732789278030396, | |
| "logps/rejected": -1.5742861032485962, | |
| "loss": 1.5533, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -11.732789993286133, | |
| "rewards/margins": 4.010071754455566, | |
| "rewards/rejected": -15.7428617477417, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45952444563184613, | |
| "grad_norm": 111.72733375525473, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.5549886226654053, | |
| "logits/rejected": -1.4790352582931519, | |
| "logps/chosen": -1.1950467824935913, | |
| "logps/rejected": -1.602651596069336, | |
| "loss": 1.5756, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -11.950468063354492, | |
| "rewards/margins": 4.076048851013184, | |
| "rewards/rejected": -16.02651596069336, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4702110606465402, | |
| "grad_norm": 95.02342077059286, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.483459711074829, | |
| "logits/rejected": -1.454347014427185, | |
| "logps/chosen": -1.2087767124176025, | |
| "logps/rejected": -1.5687811374664307, | |
| "loss": 1.434, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -12.087766647338867, | |
| "rewards/margins": 3.6000449657440186, | |
| "rewards/rejected": -15.687810897827148, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4808976756612343, | |
| "grad_norm": 80.04963040030957, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.4368016719818115, | |
| "logits/rejected": -1.4539368152618408, | |
| "logps/chosen": -1.1799414157867432, | |
| "logps/rejected": -1.7078040838241577, | |
| "loss": 1.2062, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -11.799413681030273, | |
| "rewards/margins": 5.278626918792725, | |
| "rewards/rejected": -17.078041076660156, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4915842906759284, | |
| "grad_norm": 90.27822366362969, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.5053470134735107, | |
| "logits/rejected": -1.4647208452224731, | |
| "logps/chosen": -1.3405472040176392, | |
| "logps/rejected": -1.8345695734024048, | |
| "loss": 1.1551, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -13.405473709106445, | |
| "rewards/margins": 4.9402241706848145, | |
| "rewards/rejected": -18.3456974029541, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5022709056906225, | |
| "grad_norm": 198.7780967148136, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.463410496711731, | |
| "logits/rejected": -1.4734344482421875, | |
| "logps/chosen": -1.4616682529449463, | |
| "logps/rejected": -1.8893461227416992, | |
| "loss": 1.5181, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -14.616681098937988, | |
| "rewards/margins": 4.276780128479004, | |
| "rewards/rejected": -18.893463134765625, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5129575207053166, | |
| "grad_norm": 92.23381750381444, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.5198612213134766, | |
| "logits/rejected": -1.5063666105270386, | |
| "logps/chosen": -1.4451749324798584, | |
| "logps/rejected": -2.1483402252197266, | |
| "loss": 1.3406, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -14.451748847961426, | |
| "rewards/margins": 7.031652927398682, | |
| "rewards/rejected": -21.483402252197266, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5236441357200107, | |
| "grad_norm": 101.7965922889438, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.4860137701034546, | |
| "logits/rejected": -1.4609707593917847, | |
| "logps/chosen": -1.2388508319854736, | |
| "logps/rejected": -1.7417356967926025, | |
| "loss": 1.2849, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -12.388508796691895, | |
| "rewards/margins": 5.0288496017456055, | |
| "rewards/rejected": -17.4173583984375, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5343307507347048, | |
| "grad_norm": 112.84840492502885, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.545151710510254, | |
| "logits/rejected": -1.48394775390625, | |
| "logps/chosen": -1.2914670705795288, | |
| "logps/rejected": -1.867653489112854, | |
| "loss": 1.226, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -12.9146728515625, | |
| "rewards/margins": 5.761862754821777, | |
| "rewards/rejected": -18.676532745361328, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450173657493989, | |
| "grad_norm": 81.89729359110314, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.5118135213851929, | |
| "logits/rejected": -1.5217331647872925, | |
| "logps/chosen": -1.5564649105072021, | |
| "logps/rejected": -1.9826581478118896, | |
| "loss": 1.2119, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.564648628234863, | |
| "rewards/margins": 4.261934757232666, | |
| "rewards/rejected": -19.826580047607422, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.555703980764093, | |
| "grad_norm": 97.54502724467275, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.5081006288528442, | |
| "logits/rejected": -1.4816151857376099, | |
| "logps/chosen": -1.337573766708374, | |
| "logps/rejected": -1.8502023220062256, | |
| "loss": 1.226, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -13.375738143920898, | |
| "rewards/margins": 5.126286029815674, | |
| "rewards/rejected": -18.502025604248047, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.566390595778787, | |
| "grad_norm": 99.34071724726068, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.4894455671310425, | |
| "logits/rejected": -1.4260051250457764, | |
| "logps/chosen": -1.388643503189087, | |
| "logps/rejected": -1.9015836715698242, | |
| "loss": 1.1655, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -13.886438369750977, | |
| "rewards/margins": 5.129402160644531, | |
| "rewards/rejected": -19.015838623046875, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5770772107934812, | |
| "grad_norm": 150.30388856950947, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.5251694917678833, | |
| "logits/rejected": -1.510664701461792, | |
| "logps/chosen": -1.352906346321106, | |
| "logps/rejected": -1.9253631830215454, | |
| "loss": 1.2737, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -13.529065132141113, | |
| "rewards/margins": 5.724569320678711, | |
| "rewards/rejected": -19.253633499145508, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5877638258081752, | |
| "grad_norm": 106.62043935215682, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.6284313201904297, | |
| "logits/rejected": -1.5873377323150635, | |
| "logps/chosen": -1.4256664514541626, | |
| "logps/rejected": -2.05672025680542, | |
| "loss": 1.356, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -14.256665229797363, | |
| "rewards/margins": 6.310537338256836, | |
| "rewards/rejected": -20.567203521728516, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5984504408228694, | |
| "grad_norm": 84.26686987382098, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.4645414352416992, | |
| "logits/rejected": -1.4171415567398071, | |
| "logps/chosen": -1.4661189317703247, | |
| "logps/rejected": -1.9972292184829712, | |
| "loss": 1.1226, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -14.661189079284668, | |
| "rewards/margins": 5.311104774475098, | |
| "rewards/rejected": -19.972293853759766, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6091370558375635, | |
| "grad_norm": 111.79495099216508, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.580186128616333, | |
| "logits/rejected": -1.4852710962295532, | |
| "logps/chosen": -1.6370970010757446, | |
| "logps/rejected": -2.1605658531188965, | |
| "loss": 1.3708, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -16.370969772338867, | |
| "rewards/margins": 5.234689235687256, | |
| "rewards/rejected": -21.605661392211914, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6198236708522575, | |
| "grad_norm": 89.00502902448149, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.6441500186920166, | |
| "logits/rejected": -1.6167094707489014, | |
| "logps/chosen": -1.5200397968292236, | |
| "logps/rejected": -2.0216360092163086, | |
| "loss": 1.19, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.200396537780762, | |
| "rewards/margins": 5.01596212387085, | |
| "rewards/rejected": -20.216358184814453, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305102858669517, | |
| "grad_norm": 101.58972547942741, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.5900800228118896, | |
| "logits/rejected": -1.5477546453475952, | |
| "logps/chosen": -1.4291937351226807, | |
| "logps/rejected": -2.104898691177368, | |
| "loss": 1.0996, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -14.291936874389648, | |
| "rewards/margins": 6.757050514221191, | |
| "rewards/rejected": -21.04898452758789, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6411969008816457, | |
| "grad_norm": 114.7765696570134, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.5715553760528564, | |
| "logits/rejected": -1.5747449398040771, | |
| "logps/chosen": -1.4967474937438965, | |
| "logps/rejected": -2.078972101211548, | |
| "loss": 1.0805, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -14.967473983764648, | |
| "rewards/margins": 5.8222455978393555, | |
| "rewards/rejected": -20.789722442626953, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6518835158963399, | |
| "grad_norm": 123.5878416310842, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.4123915433883667, | |
| "logits/rejected": -1.4187986850738525, | |
| "logps/chosen": -1.2821505069732666, | |
| "logps/rejected": -2.120354413986206, | |
| "loss": 1.1007, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -12.821507453918457, | |
| "rewards/margins": 8.382037162780762, | |
| "rewards/rejected": -21.203542709350586, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6625701309110339, | |
| "grad_norm": 94.61228245151283, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.5294448137283325, | |
| "logits/rejected": -1.5259336233139038, | |
| "logps/chosen": -1.5088483095169067, | |
| "logps/rejected": -2.108046531677246, | |
| "loss": 1.147, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -15.088482856750488, | |
| "rewards/margins": 5.991982460021973, | |
| "rewards/rejected": -21.080463409423828, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.673256745925728, | |
| "grad_norm": 88.47117948255506, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.4750919342041016, | |
| "logits/rejected": -1.4806705713272095, | |
| "logps/chosen": -1.588282823562622, | |
| "logps/rejected": -2.2189738750457764, | |
| "loss": 1.0766, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -15.882827758789062, | |
| "rewards/margins": 6.306912899017334, | |
| "rewards/rejected": -22.189741134643555, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6839433609404221, | |
| "grad_norm": 125.62338182234869, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.6073205471038818, | |
| "logits/rejected": -1.538448691368103, | |
| "logps/chosen": -1.5898334980010986, | |
| "logps/rejected": -2.3282947540283203, | |
| "loss": 1.2473, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.898333549499512, | |
| "rewards/margins": 7.3846116065979, | |
| "rewards/rejected": -23.282947540283203, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6946299759551162, | |
| "grad_norm": 111.24354162328734, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.5743417739868164, | |
| "logits/rejected": -1.566840410232544, | |
| "logps/chosen": -1.6232945919036865, | |
| "logps/rejected": -2.2983360290527344, | |
| "loss": 1.0508, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -16.232946395874023, | |
| "rewards/margins": 6.750412940979004, | |
| "rewards/rejected": -22.983360290527344, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7053165909698104, | |
| "grad_norm": 85.89351794884949, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.5692135095596313, | |
| "logits/rejected": -1.5725057125091553, | |
| "logps/chosen": -1.542747974395752, | |
| "logps/rejected": -2.168196678161621, | |
| "loss": 1.0903, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.42747974395752, | |
| "rewards/margins": 6.254487037658691, | |
| "rewards/rejected": -21.68196678161621, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160032059845044, | |
| "grad_norm": 116.80439013939746, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.55947744846344, | |
| "logits/rejected": -1.501947283744812, | |
| "logps/chosen": -1.5338480472564697, | |
| "logps/rejected": -2.0976195335388184, | |
| "loss": 1.0239, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -15.338479995727539, | |
| "rewards/margins": 5.637716770172119, | |
| "rewards/rejected": -20.976198196411133, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7266898209991985, | |
| "grad_norm": 78.12195704672156, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.4831820726394653, | |
| "logits/rejected": -1.5086932182312012, | |
| "logps/chosen": -1.5225416421890259, | |
| "logps/rejected": -2.1293392181396484, | |
| "loss": 0.9575, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.22541618347168, | |
| "rewards/margins": 6.0679755210876465, | |
| "rewards/rejected": -21.293392181396484, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7373764360138926, | |
| "grad_norm": 91.10910799077911, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.5832130908966064, | |
| "logits/rejected": -1.5328218936920166, | |
| "logps/chosen": -1.4637863636016846, | |
| "logps/rejected": -2.059370517730713, | |
| "loss": 1.0804, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -14.637863159179688, | |
| "rewards/margins": 5.95583963394165, | |
| "rewards/rejected": -20.593706130981445, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7480630510285867, | |
| "grad_norm": 120.29698335176921, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.519425868988037, | |
| "logits/rejected": -1.5000859498977661, | |
| "logps/chosen": -1.5877223014831543, | |
| "logps/rejected": -2.4219236373901367, | |
| "loss": 1.282, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.877223014831543, | |
| "rewards/margins": 8.342010498046875, | |
| "rewards/rejected": -24.219234466552734, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7587496660432808, | |
| "grad_norm": 114.7323293702361, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.5452375411987305, | |
| "logits/rejected": -1.477497935295105, | |
| "logps/chosen": -1.8073337078094482, | |
| "logps/rejected": -2.5182600021362305, | |
| "loss": 1.2039, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -18.073339462280273, | |
| "rewards/margins": 7.109262943267822, | |
| "rewards/rejected": -25.182600021362305, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7694362810579749, | |
| "grad_norm": 101.20181237745652, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.5502760410308838, | |
| "logits/rejected": -1.5708996057510376, | |
| "logps/chosen": -1.6630090475082397, | |
| "logps/rejected": -2.226963520050049, | |
| "loss": 1.0971, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -16.63009262084961, | |
| "rewards/margins": 5.639544486999512, | |
| "rewards/rejected": -22.269634246826172, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7801228960726689, | |
| "grad_norm": 139.38000134354874, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.5358211994171143, | |
| "logits/rejected": -1.5561330318450928, | |
| "logps/chosen": -1.6269845962524414, | |
| "logps/rejected": -2.3564658164978027, | |
| "loss": 1.1184, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -16.269845962524414, | |
| "rewards/margins": 7.294811248779297, | |
| "rewards/rejected": -23.564655303955078, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7908095110873631, | |
| "grad_norm": 104.011040382236, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.5664985179901123, | |
| "logits/rejected": -1.4946684837341309, | |
| "logps/chosen": -1.5413891077041626, | |
| "logps/rejected": -2.3655648231506348, | |
| "loss": 1.0455, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.413891792297363, | |
| "rewards/margins": 8.241754531860352, | |
| "rewards/rejected": -23.6556453704834, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8014961261020572, | |
| "grad_norm": 98.02139267959178, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.5228099822998047, | |
| "logits/rejected": -1.5087474584579468, | |
| "logps/chosen": -1.480344295501709, | |
| "logps/rejected": -2.425346851348877, | |
| "loss": 1.2231, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -14.803441047668457, | |
| "rewards/margins": 9.45002555847168, | |
| "rewards/rejected": -24.253467559814453, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8121827411167513, | |
| "grad_norm": 134.01729655272072, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.497368574142456, | |
| "logits/rejected": -1.5044596195220947, | |
| "logps/chosen": -1.594871163368225, | |
| "logps/rejected": -2.4201245307922363, | |
| "loss": 1.2771, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -15.948710441589355, | |
| "rewards/margins": 8.252533912658691, | |
| "rewards/rejected": -24.201244354248047, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8228693561314454, | |
| "grad_norm": 74.68961856468854, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.551857352256775, | |
| "logits/rejected": -1.535230278968811, | |
| "logps/chosen": -1.6363334655761719, | |
| "logps/rejected": -2.4195470809936523, | |
| "loss": 1.0466, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -16.363332748413086, | |
| "rewards/margins": 7.832139015197754, | |
| "rewards/rejected": -24.195472717285156, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8335559711461394, | |
| "grad_norm": 107.5882104761143, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.5829932689666748, | |
| "logits/rejected": -1.5114977359771729, | |
| "logps/chosen": -1.5258924961090088, | |
| "logps/rejected": -2.1448683738708496, | |
| "loss": 1.0185, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.25892448425293, | |
| "rewards/margins": 6.189759254455566, | |
| "rewards/rejected": -21.448684692382812, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8442425861608336, | |
| "grad_norm": 102.80663837958856, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.4724600315093994, | |
| "logits/rejected": -1.4828989505767822, | |
| "logps/chosen": -1.5901472568511963, | |
| "logps/rejected": -2.1642448902130127, | |
| "loss": 1.1312, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.901473999023438, | |
| "rewards/margins": 5.740975379943848, | |
| "rewards/rejected": -21.642446517944336, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8549292011755276, | |
| "grad_norm": 101.28973665181952, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.561643123626709, | |
| "logits/rejected": -1.5329078435897827, | |
| "logps/chosen": -1.514060616493225, | |
| "logps/rejected": -2.152096748352051, | |
| "loss": 1.1687, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.140606880187988, | |
| "rewards/margins": 6.380361080169678, | |
| "rewards/rejected": -21.520967483520508, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8549292011755276, | |
| "eval_logits/chosen": -1.7677618265151978, | |
| "eval_logits/rejected": -1.773621916770935, | |
| "eval_logps/chosen": -1.601662278175354, | |
| "eval_logps/rejected": -2.307917356491089, | |
| "eval_loss": 1.0876977443695068, | |
| "eval_rewards/accuracies": 0.8870967626571655, | |
| "eval_rewards/chosen": -16.016620635986328, | |
| "eval_rewards/margins": 7.062554836273193, | |
| "eval_rewards/rejected": -23.07917594909668, | |
| "eval_runtime": 96.7009, | |
| "eval_samples_per_second": 20.279, | |
| "eval_steps_per_second": 0.641, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8656158161902218, | |
| "grad_norm": 133.15081807524635, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.5344595909118652, | |
| "logits/rejected": -1.494829773902893, | |
| "logps/chosen": -1.7153148651123047, | |
| "logps/rejected": -2.2892117500305176, | |
| "loss": 1.2171, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -17.153148651123047, | |
| "rewards/margins": 5.7389678955078125, | |
| "rewards/rejected": -22.89211654663086, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8763024312049158, | |
| "grad_norm": 158.42453976320837, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.5358643531799316, | |
| "logits/rejected": -1.4975647926330566, | |
| "logps/chosen": -1.5330779552459717, | |
| "logps/rejected": -2.2891881465911865, | |
| "loss": 1.1582, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.330778121948242, | |
| "rewards/margins": 7.561104774475098, | |
| "rewards/rejected": -22.89188575744629, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.88698904621961, | |
| "grad_norm": 104.59131788649188, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.5346996784210205, | |
| "logits/rejected": -1.5254827737808228, | |
| "logps/chosen": -1.5777432918548584, | |
| "logps/rejected": -2.2442333698272705, | |
| "loss": 1.0981, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -15.777432441711426, | |
| "rewards/margins": 6.664898872375488, | |
| "rewards/rejected": -22.442333221435547, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.897675661234304, | |
| "grad_norm": 95.79743506562838, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.5502920150756836, | |
| "logits/rejected": -1.5145201683044434, | |
| "logps/chosen": -1.432101845741272, | |
| "logps/rejected": -2.1306354999542236, | |
| "loss": 1.072, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -14.321019172668457, | |
| "rewards/margins": 6.985334873199463, | |
| "rewards/rejected": -21.306354522705078, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9083622762489981, | |
| "grad_norm": 91.78736413360674, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.5086685419082642, | |
| "logits/rejected": -1.5284610986709595, | |
| "logps/chosen": -1.5403655767440796, | |
| "logps/rejected": -2.4292709827423096, | |
| "loss": 0.9504, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.403654098510742, | |
| "rewards/margins": 8.889054298400879, | |
| "rewards/rejected": -24.292707443237305, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9190488912636923, | |
| "grad_norm": 84.03123746989313, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.6175905466079712, | |
| "logits/rejected": -1.6218225955963135, | |
| "logps/chosen": -1.5030295848846436, | |
| "logps/rejected": -2.3066420555114746, | |
| "loss": 1.0077, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.030294418334961, | |
| "rewards/margins": 8.036127090454102, | |
| "rewards/rejected": -23.066423416137695, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9297355062783863, | |
| "grad_norm": 117.932317769854, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.5319820642471313, | |
| "logits/rejected": -1.5626695156097412, | |
| "logps/chosen": -1.5170115232467651, | |
| "logps/rejected": -2.136791229248047, | |
| "loss": 1.0213, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -15.170114517211914, | |
| "rewards/margins": 6.197798252105713, | |
| "rewards/rejected": -21.36791229248047, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9404221212930804, | |
| "grad_norm": 126.32296519295083, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.5235329866409302, | |
| "logits/rejected": -1.5282210111618042, | |
| "logps/chosen": -1.6138379573822021, | |
| "logps/rejected": -2.333554744720459, | |
| "loss": 1.203, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -16.13838005065918, | |
| "rewards/margins": 7.197165489196777, | |
| "rewards/rejected": -23.335546493530273, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9511087363077745, | |
| "grad_norm": 102.89001505284573, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.6163692474365234, | |
| "logits/rejected": -1.5660312175750732, | |
| "logps/chosen": -1.5261099338531494, | |
| "logps/rejected": -2.2487072944641113, | |
| "loss": 1.1582, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.261096000671387, | |
| "rewards/margins": 7.225976467132568, | |
| "rewards/rejected": -22.487075805664062, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9617953513224686, | |
| "grad_norm": 119.74095319029244, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.5374672412872314, | |
| "logits/rejected": -1.4871861934661865, | |
| "logps/chosen": -1.5755971670150757, | |
| "logps/rejected": -2.061048984527588, | |
| "loss": 1.073, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -15.75597095489502, | |
| "rewards/margins": 4.854517459869385, | |
| "rewards/rejected": -20.610488891601562, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9724819663371627, | |
| "grad_norm": 76.57788575531883, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.591564416885376, | |
| "logits/rejected": -1.5509247779846191, | |
| "logps/chosen": -1.5886285305023193, | |
| "logps/rejected": -2.421255588531494, | |
| "loss": 0.9676, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -15.886283874511719, | |
| "rewards/margins": 8.326271057128906, | |
| "rewards/rejected": -24.212556838989258, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9831685813518568, | |
| "grad_norm": 109.6328495391863, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.5486748218536377, | |
| "logits/rejected": -1.549953818321228, | |
| "logps/chosen": -1.5361286401748657, | |
| "logps/rejected": -2.2728710174560547, | |
| "loss": 1.1484, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -15.361285209655762, | |
| "rewards/margins": 7.367424964904785, | |
| "rewards/rejected": -22.728710174560547, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9938551963665508, | |
| "grad_norm": 101.61409866072276, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.5434294939041138, | |
| "logits/rejected": -1.5330185890197754, | |
| "logps/chosen": -1.7000032663345337, | |
| "logps/rejected": -2.3660502433776855, | |
| "loss": 1.0908, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -17.000032424926758, | |
| "rewards/margins": 6.660471439361572, | |
| "rewards/rejected": -23.66050148010254, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9981298423724285, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.7411203522243133, | |
| "train_runtime": 12577.1262, | |
| "train_samples_per_second": 4.761, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 467, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 120, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |