{ "best_metric": 0.23926551640033722, "best_model_checkpoint": "models/qwen2.5-3b-dpo-mini/checkpoint-10000", "epoch": 1.9999360981532366, "eval_steps": 10000, "global_step": 15648, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00012780369352674293, "grad_norm": 1.583091404664238, "learning_rate": 6.385696040868454e-10, "logits/chosen": -2.140625, "logits/rejected": -2.015625, "logps/chosen": -141.0, "logps/rejected": -147.0, "loss": 0.6914, "rewards/accuracies": 0.0, "rewards/chosen": 0.0, "rewards/margins": 0.0, "rewards/rejected": 0.0, "step": 1 }, { "epoch": 0.0012780369352674292, "grad_norm": 1.831193607031671, "learning_rate": 6.3856960408684546e-09, "logits/chosen": -2.484375, "logits/rejected": -1.9375, "logps/chosen": -156.0, "logps/rejected": -136.0, "loss": 0.6924, "rewards/accuracies": 0.2361111044883728, "rewards/chosen": 8.869171142578125e-05, "rewards/margins": 8.726119995117188e-05, "rewards/rejected": -4.246830940246582e-07, "step": 10 }, { "epoch": 0.0025560738705348585, "grad_norm": 1.9051082450552785, "learning_rate": 1.2771392081736909e-08, "logits/chosen": -2.53125, "logits/rejected": -2.125, "logps/chosen": -167.0, "logps/rejected": -134.0, "loss": 0.6928, "rewards/accuracies": 0.33125001192092896, "rewards/chosen": -0.00012493133544921875, "rewards/margins": -0.000560760498046875, "rewards/rejected": 0.000438690185546875, "step": 20 }, { "epoch": 0.0038341108058022877, "grad_norm": 2.0951178941868793, "learning_rate": 1.915708812260536e-08, "logits/chosen": -2.546875, "logits/rejected": -2.125, "logps/chosen": -172.0, "logps/rejected": -144.0, "loss": 0.6925, "rewards/accuracies": 0.3812499940395355, "rewards/chosen": -0.00020313262939453125, "rewards/margins": 0.000377655029296875, "rewards/rejected": -0.000579833984375, "step": 30 }, { "epoch": 0.005112147741069717, "grad_norm": 1.8582175367673581, "learning_rate": 2.5542784163473818e-08, "logits/chosen": -2.4375, "logits/rejected": -2.0625, "logps/chosen": -153.0, "logps/rejected": -133.0, "loss": 0.6926, "rewards/accuracies": 0.34375, "rewards/chosen": -0.00213623046875, "rewards/margins": -0.0004711151123046875, "rewards/rejected": -0.00165557861328125, "step": 40 }, { "epoch": 0.006390184676337146, "grad_norm": 1.8191369002557825, "learning_rate": 3.192848020434227e-08, "logits/chosen": -2.484375, "logits/rejected": -2.03125, "logps/chosen": -165.0, "logps/rejected": -135.0, "loss": 0.6929, "rewards/accuracies": 0.3499999940395355, "rewards/chosen": 6.29425048828125e-05, "rewards/margins": -0.0002498626708984375, "rewards/rejected": 0.0003147125244140625, "step": 50 }, { "epoch": 0.007668221611604575, "grad_norm": 1.8104389933900382, "learning_rate": 3.831417624521072e-08, "logits/chosen": -2.4375, "logits/rejected": -2.03125, "logps/chosen": -154.0, "logps/rejected": -126.0, "loss": 0.6929, "rewards/accuracies": 0.375, "rewards/chosen": -7.82012939453125e-05, "rewards/margins": -0.00019168853759765625, "rewards/rejected": 0.00010967254638671875, "step": 60 }, { "epoch": 0.008946258546872005, "grad_norm": 1.9080659587821094, "learning_rate": 4.469987228607918e-08, "logits/chosen": -2.390625, "logits/rejected": -2.015625, "logps/chosen": -164.0, "logps/rejected": -141.0, "loss": 0.6928, "rewards/accuracies": 0.3687500059604645, "rewards/chosen": 0.000797271728515625, "rewards/margins": 0.0012969970703125, "rewards/rejected": -0.000499725341796875, "step": 70 }, { "epoch": 0.010224295482139434, "grad_norm": 1.7284420631380606, "learning_rate": 5.1085568326947637e-08, "logits/chosen": -2.4375, "logits/rejected": -2.0, "logps/chosen": -163.0, "logps/rejected": -135.0, "loss": 0.6926, "rewards/accuracies": 0.42500001192092896, "rewards/chosen": 0.0013580322265625, "rewards/margins": 0.0010833740234375, "rewards/rejected": 0.00028228759765625, "step": 80 }, { "epoch": 0.011502332417406863, "grad_norm": 1.9985804873605706, "learning_rate": 5.747126436781609e-08, "logits/chosen": -2.40625, "logits/rejected": -2.03125, "logps/chosen": -153.0, "logps/rejected": -142.0, "loss": 0.6925, "rewards/accuracies": 0.36250001192092896, "rewards/chosen": 7.867813110351562e-05, "rewards/margins": 0.0002193450927734375, "rewards/rejected": -0.00014019012451171875, "step": 90 }, { "epoch": 0.012780369352674292, "grad_norm": 1.7673732788906307, "learning_rate": 6.385696040868454e-08, "logits/chosen": -2.53125, "logits/rejected": -1.9296875, "logps/chosen": -157.0, "logps/rejected": -138.0, "loss": 0.6923, "rewards/accuracies": 0.3687500059604645, "rewards/chosen": -0.000423431396484375, "rewards/margins": -0.0003070831298828125, "rewards/rejected": -0.00011539459228515625, "step": 100 }, { "epoch": 0.014058406287941722, "grad_norm": 1.8348538098786262, "learning_rate": 7.024265644955301e-08, "logits/chosen": -2.53125, "logits/rejected": -1.96875, "logps/chosen": -151.0, "logps/rejected": -132.0, "loss": 0.692, "rewards/accuracies": 0.4124999940395355, "rewards/chosen": 0.001129150390625, "rewards/margins": 0.000797271728515625, "rewards/rejected": 0.00032806396484375, "step": 110 }, { "epoch": 0.01533644322320915, "grad_norm": 2.0218021664474373, "learning_rate": 7.662835249042144e-08, "logits/chosen": -2.40625, "logits/rejected": -1.859375, "logps/chosen": -152.0, "logps/rejected": -137.0, "loss": 0.6918, "rewards/accuracies": 0.48124998807907104, "rewards/chosen": 0.002960205078125, "rewards/margins": 0.0024261474609375, "rewards/rejected": 0.000530242919921875, "step": 120 }, { "epoch": 0.01661448015847658, "grad_norm": 1.82922758998825, "learning_rate": 8.301404853128991e-08, "logits/chosen": -2.421875, "logits/rejected": -1.9140625, "logps/chosen": -144.0, "logps/rejected": -158.0, "loss": 0.6915, "rewards/accuracies": 0.543749988079071, "rewards/chosen": 0.006317138671875, "rewards/margins": 0.00396728515625, "rewards/rejected": 0.002349853515625, "step": 130 }, { "epoch": 0.01789251709374401, "grad_norm": 1.7861618345539088, "learning_rate": 8.939974457215836e-08, "logits/chosen": -2.53125, "logits/rejected": -1.984375, "logps/chosen": -165.0, "logps/rejected": -141.0, "loss": 0.6909, "rewards/accuracies": 0.53125, "rewards/chosen": 0.0064697265625, "rewards/margins": 0.0040283203125, "rewards/rejected": 0.0024261474609375, "step": 140 }, { "epoch": 0.01917055402901144, "grad_norm": 1.8958781010879984, "learning_rate": 9.578544061302682e-08, "logits/chosen": -2.46875, "logits/rejected": -2.203125, "logps/chosen": -154.0, "logps/rejected": -132.0, "loss": 0.6903, "rewards/accuracies": 0.5249999761581421, "rewards/chosen": 0.0076904296875, "rewards/margins": 0.0042724609375, "rewards/rejected": 0.00341796875, "step": 150 }, { "epoch": 0.020448590964278868, "grad_norm": 1.689453237922255, "learning_rate": 1.0217113665389527e-07, "logits/chosen": -2.4375, "logits/rejected": -1.9140625, "logps/chosen": -168.0, "logps/rejected": -152.0, "loss": 0.6903, "rewards/accuracies": 0.4937500059604645, "rewards/chosen": 0.005950927734375, "rewards/margins": 0.0031585693359375, "rewards/rejected": 0.0027923583984375, "step": 160 }, { "epoch": 0.021726627899546297, "grad_norm": 1.6908724447283379, "learning_rate": 1.0855683269476372e-07, "logits/chosen": -2.578125, "logits/rejected": -1.890625, "logps/chosen": -153.0, "logps/rejected": -151.0, "loss": 0.6894, "rewards/accuracies": 0.574999988079071, "rewards/chosen": 0.01123046875, "rewards/margins": 0.00665283203125, "rewards/rejected": 0.004608154296875, "step": 170 }, { "epoch": 0.023004664834813726, "grad_norm": 1.9217126730360887, "learning_rate": 1.1494252873563217e-07, "logits/chosen": -2.453125, "logits/rejected": -1.8828125, "logps/chosen": -155.0, "logps/rejected": -146.0, "loss": 0.6875, "rewards/accuracies": 0.6499999761581421, "rewards/chosen": 0.01275634765625, "rewards/margins": 0.00927734375, "rewards/rejected": 0.003509521484375, "step": 180 }, { "epoch": 0.024282701770081155, "grad_norm": 1.7796346205464415, "learning_rate": 1.2132822477650062e-07, "logits/chosen": -2.5, "logits/rejected": -2.15625, "logps/chosen": -165.0, "logps/rejected": -144.0, "loss": 0.6877, "rewards/accuracies": 0.612500011920929, "rewards/chosen": 0.016845703125, "rewards/margins": 0.00970458984375, "rewards/rejected": 0.007080078125, "step": 190 }, { "epoch": 0.025560738705348585, "grad_norm": 1.8062803830724468, "learning_rate": 1.277139208173691e-07, "logits/chosen": -2.484375, "logits/rejected": -2.125, "logps/chosen": -128.0, "logps/rejected": -117.0, "loss": 0.6868, "rewards/accuracies": 0.675000011920929, "rewards/chosen": 0.0159912109375, "rewards/margins": 0.010986328125, "rewards/rejected": 0.0050048828125, "step": 200 }, { "epoch": 0.026838775640616014, "grad_norm": 1.8109623854322845, "learning_rate": 1.3409961685823753e-07, "logits/chosen": -2.53125, "logits/rejected": -2.125, "logps/chosen": -152.0, "logps/rejected": -135.0, "loss": 0.6859, "rewards/accuracies": 0.675000011920929, "rewards/chosen": 0.0263671875, "rewards/margins": 0.01507568359375, "rewards/rejected": 0.01129150390625, "step": 210 }, { "epoch": 0.028116812575883443, "grad_norm": 1.7555569281404872, "learning_rate": 1.4048531289910602e-07, "logits/chosen": -2.4375, "logits/rejected": -1.9609375, "logps/chosen": -168.0, "logps/rejected": -162.0, "loss": 0.6846, "rewards/accuracies": 0.7250000238418579, "rewards/chosen": 0.0260009765625, "rewards/margins": 0.0228271484375, "rewards/rejected": 0.003173828125, "step": 220 }, { "epoch": 0.029394849511150872, "grad_norm": 1.6784346166574386, "learning_rate": 1.4687100893997445e-07, "logits/chosen": -2.546875, "logits/rejected": -2.046875, "logps/chosen": -156.0, "logps/rejected": -145.0, "loss": 0.6833, "rewards/accuracies": 0.762499988079071, "rewards/chosen": 0.024658203125, "rewards/margins": 0.0225830078125, "rewards/rejected": 0.002105712890625, "step": 230 }, { "epoch": 0.0306728864464183, "grad_norm": 2.017021630939051, "learning_rate": 1.532567049808429e-07, "logits/chosen": -2.53125, "logits/rejected": -2.125, "logps/chosen": -169.0, "logps/rejected": -139.0, "loss": 0.6819, "rewards/accuracies": 0.6937500238418579, "rewards/chosen": 0.0257568359375, "rewards/margins": 0.0235595703125, "rewards/rejected": 0.0020751953125, "step": 240 }, { "epoch": 0.03195092338168573, "grad_norm": 1.7084660802314153, "learning_rate": 1.5964240102171135e-07, "logits/chosen": -2.546875, "logits/rejected": -2.15625, "logps/chosen": -154.0, "logps/rejected": -145.0, "loss": 0.6796, "rewards/accuracies": 0.699999988079071, "rewards/chosen": 0.0361328125, "rewards/margins": 0.0294189453125, "rewards/rejected": 0.006744384765625, "step": 250 }, { "epoch": 0.03322896031695316, "grad_norm": 1.8365627743934905, "learning_rate": 1.6602809706257982e-07, "logits/chosen": -2.515625, "logits/rejected": -2.0, "logps/chosen": -157.0, "logps/rejected": -152.0, "loss": 0.6764, "rewards/accuracies": 0.737500011920929, "rewards/chosen": 0.032470703125, "rewards/margins": 0.036865234375, "rewards/rejected": -0.004364013671875, "step": 260 }, { "epoch": 0.03450699725222059, "grad_norm": 1.7335796846774567, "learning_rate": 1.7241379310344828e-07, "logits/chosen": -2.546875, "logits/rejected": -1.9765625, "logps/chosen": -158.0, "logps/rejected": -139.0, "loss": 0.6742, "rewards/accuracies": 0.7437499761581421, "rewards/chosen": 0.028564453125, "rewards/margins": 0.046630859375, "rewards/rejected": -0.0179443359375, "step": 270 }, { "epoch": 0.03578503418748802, "grad_norm": 1.6763039372574249, "learning_rate": 1.7879948914431672e-07, "logits/chosen": -2.53125, "logits/rejected": -2.0, "logps/chosen": -146.0, "logps/rejected": -144.0, "loss": 0.6716, "rewards/accuracies": 0.8062499761581421, "rewards/chosen": 0.03759765625, "rewards/margins": 0.05615234375, "rewards/rejected": -0.0184326171875, "step": 280 }, { "epoch": 0.03706307112275545, "grad_norm": 1.7476768718236924, "learning_rate": 1.8518518518518516e-07, "logits/chosen": -2.53125, "logits/rejected": -2.015625, "logps/chosen": -155.0, "logps/rejected": -131.0, "loss": 0.6693, "rewards/accuracies": 0.699999988079071, "rewards/chosen": 0.0230712890625, "rewards/margins": 0.03955078125, "rewards/rejected": -0.016357421875, "step": 290 }, { "epoch": 0.03834110805802288, "grad_norm": 1.9904155949768438, "learning_rate": 1.9157088122605365e-07, "logits/chosen": -2.625, "logits/rejected": -2.15625, "logps/chosen": -165.0, "logps/rejected": -154.0, "loss": 0.6654, "rewards/accuracies": 0.7562500238418579, "rewards/chosen": 0.01300048828125, "rewards/margins": 0.05810546875, "rewards/rejected": -0.045166015625, "step": 300 }, { "epoch": 0.039619144993290306, "grad_norm": 2.1919340946635275, "learning_rate": 1.9795657726692208e-07, "logits/chosen": -2.65625, "logits/rejected": -2.09375, "logps/chosen": -150.0, "logps/rejected": -153.0, "loss": 0.6579, "rewards/accuracies": 0.768750011920929, "rewards/chosen": 0.00860595703125, "rewards/margins": 0.08154296875, "rewards/rejected": -0.07275390625, "step": 310 }, { "epoch": 0.040897181928557735, "grad_norm": 2.050188028710248, "learning_rate": 2.0434227330779055e-07, "logits/chosen": -2.578125, "logits/rejected": -2.25, "logps/chosen": -151.0, "logps/rejected": -155.0, "loss": 0.654, "rewards/accuracies": 0.78125, "rewards/chosen": 0.0015411376953125, "rewards/margins": 0.08837890625, "rewards/rejected": -0.0869140625, "step": 320 }, { "epoch": 0.042175218863825165, "grad_norm": 2.016463993421717, "learning_rate": 2.1072796934865898e-07, "logits/chosen": -2.734375, "logits/rejected": -2.265625, "logps/chosen": -164.0, "logps/rejected": -165.0, "loss": 0.6475, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -0.03125, "rewards/margins": 0.10693359375, "rewards/rejected": -0.1376953125, "step": 330 }, { "epoch": 0.043453255799092594, "grad_norm": 2.4323566534644714, "learning_rate": 2.1711366538952745e-07, "logits/chosen": -2.6875, "logits/rejected": -2.171875, "logps/chosen": -171.0, "logps/rejected": -153.0, "loss": 0.6386, "rewards/accuracies": 0.8062499761581421, "rewards/chosen": -0.05078125, "rewards/margins": 0.1279296875, "rewards/rejected": -0.177734375, "step": 340 }, { "epoch": 0.04473129273436002, "grad_norm": 2.202903045169722, "learning_rate": 2.234993614303959e-07, "logits/chosen": -2.734375, "logits/rejected": -2.21875, "logps/chosen": -150.0, "logps/rejected": -169.0, "loss": 0.6285, "rewards/accuracies": 0.7749999761581421, "rewards/chosen": -0.0888671875, "rewards/margins": 0.1650390625, "rewards/rejected": -0.25390625, "step": 350 }, { "epoch": 0.04600932966962745, "grad_norm": 2.206602901584152, "learning_rate": 2.2988505747126435e-07, "logits/chosen": -2.703125, "logits/rejected": -2.296875, "logps/chosen": -181.0, "logps/rejected": -158.0, "loss": 0.6227, "rewards/accuracies": 0.75, "rewards/chosen": -0.1357421875, "rewards/margins": 0.1533203125, "rewards/rejected": -0.2890625, "step": 360 }, { "epoch": 0.04728736660489488, "grad_norm": 2.077125523450064, "learning_rate": 2.3627075351213284e-07, "logits/chosen": -2.65625, "logits/rejected": -2.203125, "logps/chosen": -181.0, "logps/rejected": -185.0, "loss": 0.6089, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -0.1953125, "rewards/margins": 0.2177734375, "rewards/rejected": -0.4140625, "step": 370 }, { "epoch": 0.04856540354016231, "grad_norm": 2.286616738197532, "learning_rate": 2.4265644955300125e-07, "logits/chosen": -2.71875, "logits/rejected": -2.21875, "logps/chosen": -165.0, "logps/rejected": -192.0, "loss": 0.5919, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -0.2578125, "rewards/margins": 0.2734375, "rewards/rejected": -0.53125, "step": 380 }, { "epoch": 0.04984344047542974, "grad_norm": 2.3378312139893755, "learning_rate": 2.490421455938697e-07, "logits/chosen": -2.765625, "logits/rejected": -2.34375, "logps/chosen": -205.0, "logps/rejected": -184.0, "loss": 0.5987, "rewards/accuracies": 0.6812499761581421, "rewards/chosen": -0.38671875, "rewards/margins": 0.19921875, "rewards/rejected": -0.5859375, "step": 390 }, { "epoch": 0.05112147741069717, "grad_norm": 2.0376352660492505, "learning_rate": 2.554278416347382e-07, "logits/chosen": -2.734375, "logits/rejected": -2.4375, "logps/chosen": -205.0, "logps/rejected": -225.0, "loss": 0.5894, "rewards/accuracies": 0.7437499761581421, "rewards/chosen": -0.423828125, "rewards/margins": 0.25, "rewards/rejected": -0.671875, "step": 400 }, { "epoch": 0.0523995143459646, "grad_norm": 2.198540810606561, "learning_rate": 2.6181353767560667e-07, "logits/chosen": -2.796875, "logits/rejected": -2.375, "logps/chosen": -203.0, "logps/rejected": -228.0, "loss": 0.5725, "rewards/accuracies": 0.7562500238418579, "rewards/chosen": -0.423828125, "rewards/margins": 0.341796875, "rewards/rejected": -0.765625, "step": 410 }, { "epoch": 0.05367755128123203, "grad_norm": 2.5064207498697306, "learning_rate": 2.6819923371647505e-07, "logits/chosen": -2.84375, "logits/rejected": -2.453125, "logps/chosen": -203.0, "logps/rejected": -231.0, "loss": 0.5589, "rewards/accuracies": 0.737500011920929, "rewards/chosen": -0.45703125, "rewards/margins": 0.408203125, "rewards/rejected": -0.86328125, "step": 420 }, { "epoch": 0.05495558821649946, "grad_norm": 2.7332223174524604, "learning_rate": 2.7458492975734354e-07, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -207.0, "logps/rejected": -217.0, "loss": 0.5532, "rewards/accuracies": 0.7250000238418579, "rewards/chosen": -0.53515625, "rewards/margins": 0.384765625, "rewards/rejected": -0.91796875, "step": 430 }, { "epoch": 0.056233625151766886, "grad_norm": 3.160820809244754, "learning_rate": 2.8097062579821203e-07, "logits/chosen": -3.0625, "logits/rejected": -2.59375, "logps/chosen": -224.0, "logps/rejected": -276.0, "loss": 0.5351, "rewards/accuracies": 0.78125, "rewards/chosen": -0.63671875, "rewards/margins": 0.609375, "rewards/rejected": -1.25, "step": 440 }, { "epoch": 0.057511662087034315, "grad_norm": 3.1305847824157005, "learning_rate": 2.873563218390804e-07, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -225.0, "logps/rejected": -266.0, "loss": 0.5226, "rewards/accuracies": 0.75, "rewards/chosen": -0.69921875, "rewards/margins": 0.5859375, "rewards/rejected": -1.2890625, "step": 450 }, { "epoch": 0.058789699022301745, "grad_norm": 3.2458132009733975, "learning_rate": 2.937420178799489e-07, "logits/chosen": -3.21875, "logits/rejected": -2.890625, "logps/chosen": -212.0, "logps/rejected": -284.0, "loss": 0.5138, "rewards/accuracies": 0.78125, "rewards/chosen": -0.734375, "rewards/margins": 0.7421875, "rewards/rejected": -1.4765625, "step": 460 }, { "epoch": 0.060067735957569174, "grad_norm": 3.590934206396528, "learning_rate": 3.001277139208174e-07, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -227.0, "logps/rejected": -286.0, "loss": 0.5261, "rewards/accuracies": 0.78125, "rewards/chosen": -0.78515625, "rewards/margins": 0.65234375, "rewards/rejected": -1.4375, "step": 470 }, { "epoch": 0.0613457728928366, "grad_norm": 4.090558879807313, "learning_rate": 3.065134099616858e-07, "logits/chosen": -3.171875, "logits/rejected": -2.8125, "logps/chosen": -241.0, "logps/rejected": -296.0, "loss": 0.5063, "rewards/accuracies": 0.7875000238418579, "rewards/chosen": -0.71484375, "rewards/margins": 0.80078125, "rewards/rejected": -1.515625, "step": 480 }, { "epoch": 0.06262380982810403, "grad_norm": 3.738095666639755, "learning_rate": 3.1289910600255427e-07, "logits/chosen": -3.28125, "logits/rejected": -2.875, "logps/chosen": -232.0, "logps/rejected": -306.0, "loss": 0.5051, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -0.78125, "rewards/margins": 0.87890625, "rewards/rejected": -1.6640625, "step": 490 }, { "epoch": 0.06390184676337146, "grad_norm": 4.03887506227773, "learning_rate": 3.192848020434227e-07, "logits/chosen": -3.3125, "logits/rejected": -2.859375, "logps/chosen": -227.0, "logps/rejected": -286.0, "loss": 0.5208, "rewards/accuracies": 0.7437499761581421, "rewards/chosen": -0.8046875, "rewards/margins": 0.63671875, "rewards/rejected": -1.4453125, "step": 500 }, { "epoch": 0.06517988369863889, "grad_norm": 3.7784530781629444, "learning_rate": 3.2567049808429114e-07, "logits/chosen": -3.234375, "logits/rejected": -2.9375, "logps/chosen": -240.0, "logps/rejected": -296.0, "loss": 0.4996, "rewards/accuracies": 0.7875000238418579, "rewards/chosen": -0.8515625, "rewards/margins": 0.80078125, "rewards/rejected": -1.6484375, "step": 510 }, { "epoch": 0.06645792063390632, "grad_norm": 4.907395725764965, "learning_rate": 3.3205619412515963e-07, "logits/chosen": -3.1875, "logits/rejected": -3.0, "logps/chosen": -236.0, "logps/rejected": -312.0, "loss": 0.4853, "rewards/accuracies": 0.7250000238418579, "rewards/chosen": -0.9765625, "rewards/margins": 0.734375, "rewards/rejected": -1.7109375, "step": 520 }, { "epoch": 0.06773595756917375, "grad_norm": 6.328166419969473, "learning_rate": 3.3844189016602807e-07, "logits/chosen": -3.203125, "logits/rejected": -2.90625, "logps/chosen": -274.0, "logps/rejected": -344.0, "loss": 0.4789, "rewards/accuracies": 0.762499988079071, "rewards/chosen": -1.1015625, "rewards/margins": 0.8984375, "rewards/rejected": -2.0, "step": 530 }, { "epoch": 0.06901399450444118, "grad_norm": 7.41169890134317, "learning_rate": 3.4482758620689656e-07, "logits/chosen": -3.171875, "logits/rejected": -2.890625, "logps/chosen": -264.0, "logps/rejected": -350.0, "loss": 0.4664, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -0.9921875, "rewards/margins": 0.97265625, "rewards/rejected": -1.96875, "step": 540 }, { "epoch": 0.07029203143970861, "grad_norm": 6.329307259348903, "learning_rate": 3.51213282247765e-07, "logits/chosen": -3.140625, "logits/rejected": -2.84375, "logps/chosen": -242.0, "logps/rejected": -334.0, "loss": 0.4756, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -0.97265625, "rewards/margins": 0.92578125, "rewards/rejected": -1.8984375, "step": 550 }, { "epoch": 0.07157006837497604, "grad_norm": 5.5382587319010605, "learning_rate": 3.5759897828863344e-07, "logits/chosen": -3.1875, "logits/rejected": -2.875, "logps/chosen": -290.0, "logps/rejected": -372.0, "loss": 0.4816, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.21875, "rewards/margins": 0.94140625, "rewards/rejected": -2.15625, "step": 560 }, { "epoch": 0.07284810531024347, "grad_norm": 5.59071348439989, "learning_rate": 3.639846743295019e-07, "logits/chosen": -3.21875, "logits/rejected": -2.890625, "logps/chosen": -268.0, "logps/rejected": -336.0, "loss": 0.4814, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -1.1953125, "rewards/margins": 0.87109375, "rewards/rejected": -2.0625, "step": 570 }, { "epoch": 0.0741261422455109, "grad_norm": 7.500668748088703, "learning_rate": 3.703703703703703e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -278.0, "logps/rejected": -346.0, "loss": 0.4581, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.25, "rewards/margins": 0.87890625, "rewards/rejected": -2.125, "step": 580 }, { "epoch": 0.07540417918077832, "grad_norm": 7.815381329656392, "learning_rate": 3.767560664112388e-07, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -286.0, "logps/rejected": -366.0, "loss": 0.4516, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.15625, "rewards/margins": 0.953125, "rewards/rejected": -2.109375, "step": 590 }, { "epoch": 0.07668221611604575, "grad_norm": 7.560002445698683, "learning_rate": 3.831417624521073e-07, "logits/chosen": -3.09375, "logits/rejected": -2.765625, "logps/chosen": -290.0, "logps/rejected": -414.0, "loss": 0.4415, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -1.1796875, "rewards/margins": 1.328125, "rewards/rejected": -2.5, "step": 600 }, { "epoch": 0.07796025305131318, "grad_norm": 7.597044745493579, "learning_rate": 3.895274584929757e-07, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -290.0, "logps/rejected": -404.0, "loss": 0.4573, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -1.28125, "rewards/margins": 1.171875, "rewards/rejected": -2.453125, "step": 610 }, { "epoch": 0.07923828998658061, "grad_norm": 6.0294238622536085, "learning_rate": 3.9591315453384417e-07, "logits/chosen": -2.984375, "logits/rejected": -2.84375, "logps/chosen": -288.0, "logps/rejected": -352.0, "loss": 0.433, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -1.171875, "rewards/margins": 0.93359375, "rewards/rejected": -2.109375, "step": 620 }, { "epoch": 0.08051632692184804, "grad_norm": 6.57580363708218, "learning_rate": 4.0229885057471266e-07, "logits/chosen": -2.875, "logits/rejected": -2.65625, "logps/chosen": -288.0, "logps/rejected": -370.0, "loss": 0.4397, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -1.34375, "rewards/margins": 1.0546875, "rewards/rejected": -2.40625, "step": 630 }, { "epoch": 0.08179436385711547, "grad_norm": 8.126812055550822, "learning_rate": 4.086845466155811e-07, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -290.0, "logps/rejected": -396.0, "loss": 0.4337, "rewards/accuracies": 0.793749988079071, "rewards/chosen": -1.2421875, "rewards/margins": 1.1484375, "rewards/rejected": -2.390625, "step": 640 }, { "epoch": 0.0830724007923829, "grad_norm": 7.662460966992449, "learning_rate": 4.1507024265644953e-07, "logits/chosen": -2.921875, "logits/rejected": -2.609375, "logps/chosen": -286.0, "logps/rejected": -436.0, "loss": 0.4231, "rewards/accuracies": 0.8125, "rewards/chosen": -1.4140625, "rewards/margins": 1.4140625, "rewards/rejected": -2.828125, "step": 650 }, { "epoch": 0.08435043772765033, "grad_norm": 6.26765621154057, "learning_rate": 4.2145593869731797e-07, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -268.0, "logps/rejected": -368.0, "loss": 0.4428, "rewards/accuracies": 0.78125, "rewards/chosen": -1.1796875, "rewards/margins": 1.171875, "rewards/rejected": -2.34375, "step": 660 }, { "epoch": 0.08562847466291776, "grad_norm": 10.966780568835159, "learning_rate": 4.2784163473818646e-07, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -272.0, "logps/rejected": -384.0, "loss": 0.4744, "rewards/accuracies": 0.7437499761581421, "rewards/chosen": -1.1484375, "rewards/margins": 1.171875, "rewards/rejected": -2.328125, "step": 670 }, { "epoch": 0.08690651159818519, "grad_norm": 7.995403962637887, "learning_rate": 4.342273307790549e-07, "logits/chosen": -2.90625, "logits/rejected": -2.546875, "logps/chosen": -308.0, "logps/rejected": -406.0, "loss": 0.446, "rewards/accuracies": 0.78125, "rewards/chosen": -1.328125, "rewards/margins": 1.1328125, "rewards/rejected": -2.46875, "step": 680 }, { "epoch": 0.08818454853345262, "grad_norm": 7.767114033697263, "learning_rate": 4.4061302681992333e-07, "logits/chosen": -2.90625, "logits/rejected": -2.765625, "logps/chosen": -294.0, "logps/rejected": -386.0, "loss": 0.4252, "rewards/accuracies": 0.7749999761581421, "rewards/chosen": -1.359375, "rewards/margins": 1.09375, "rewards/rejected": -2.453125, "step": 690 }, { "epoch": 0.08946258546872005, "grad_norm": 10.782647668236367, "learning_rate": 4.469987228607918e-07, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -266.0, "logps/rejected": -372.0, "loss": 0.4396, "rewards/accuracies": 0.84375, "rewards/chosen": -1.1484375, "rewards/margins": 1.1796875, "rewards/rejected": -2.328125, "step": 700 }, { "epoch": 0.09074062240398748, "grad_norm": 13.262314383907762, "learning_rate": 4.5338441890166026e-07, "logits/chosen": -2.875, "logits/rejected": -2.640625, "logps/chosen": -296.0, "logps/rejected": -408.0, "loss": 0.412, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -1.4140625, "rewards/margins": 1.2734375, "rewards/rejected": -2.6875, "step": 710 }, { "epoch": 0.0920186593392549, "grad_norm": 8.452988029757332, "learning_rate": 4.597701149425287e-07, "logits/chosen": -2.890625, "logits/rejected": -2.75, "logps/chosen": -302.0, "logps/rejected": -428.0, "loss": 0.4053, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.390625, "rewards/margins": 1.4296875, "rewards/rejected": -2.8125, "step": 720 }, { "epoch": 0.09329669627452233, "grad_norm": 10.451528356808828, "learning_rate": 4.661558109833972e-07, "logits/chosen": -3.078125, "logits/rejected": -2.78125, "logps/chosen": -286.0, "logps/rejected": -418.0, "loss": 0.4013, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -1.3671875, "rewards/margins": 1.59375, "rewards/rejected": -2.96875, "step": 730 }, { "epoch": 0.09457473320978976, "grad_norm": 11.00358677164108, "learning_rate": 4.725415070242657e-07, "logits/chosen": -3.03125, "logits/rejected": -2.859375, "logps/chosen": -328.0, "logps/rejected": -432.0, "loss": 0.387, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.515625, "rewards/margins": 1.3203125, "rewards/rejected": -2.828125, "step": 740 }, { "epoch": 0.09585277014505719, "grad_norm": 13.575988451351572, "learning_rate": 4.789272030651341e-07, "logits/chosen": -2.9375, "logits/rejected": -2.84375, "logps/chosen": -290.0, "logps/rejected": -408.0, "loss": 0.4079, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -1.4609375, "rewards/margins": 1.3203125, "rewards/rejected": -2.78125, "step": 750 }, { "epoch": 0.09713080708032462, "grad_norm": 9.325954891766974, "learning_rate": 4.853128991060025e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -302.0, "logps/rejected": -454.0, "loss": 0.4176, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -1.34375, "rewards/margins": 1.5546875, "rewards/rejected": -2.90625, "step": 760 }, { "epoch": 0.09840884401559205, "grad_norm": 7.52371689034163, "learning_rate": 4.91698595146871e-07, "logits/chosen": -2.84375, "logits/rejected": -2.6875, "logps/chosen": -330.0, "logps/rejected": -448.0, "loss": 0.4036, "rewards/accuracies": 0.75, "rewards/chosen": -1.7265625, "rewards/margins": 1.21875, "rewards/rejected": -2.953125, "step": 770 }, { "epoch": 0.09968688095085948, "grad_norm": 7.847386893700331, "learning_rate": 4.980842911877394e-07, "logits/chosen": -2.765625, "logits/rejected": -2.734375, "logps/chosen": -308.0, "logps/rejected": -450.0, "loss": 0.374, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -1.5, "rewards/margins": 1.53125, "rewards/rejected": -3.03125, "step": 780 }, { "epoch": 0.10096491788612691, "grad_norm": 8.270767140969047, "learning_rate": 4.999987806263758e-07, "logits/chosen": -2.859375, "logits/rejected": -2.78125, "logps/chosen": -298.0, "logps/rejected": -418.0, "loss": 0.4074, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.390625, "rewards/margins": 1.34375, "rewards/rejected": -2.734375, "step": 790 }, { "epoch": 0.10224295482139434, "grad_norm": 9.313330271911113, "learning_rate": 4.999928082127696e-07, "logits/chosen": -3.1875, "logits/rejected": -2.96875, "logps/chosen": -306.0, "logps/rejected": -446.0, "loss": 0.4058, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -1.484375, "rewards/margins": 1.5234375, "rewards/rejected": -3.0, "step": 800 }, { "epoch": 0.10352099175666177, "grad_norm": 10.98658746507854, "learning_rate": 4.999818589113488e-07, "logits/chosen": -3.15625, "logits/rejected": -3.0625, "logps/chosen": -282.0, "logps/rejected": -398.0, "loss": 0.3954, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.25, "rewards/margins": 1.40625, "rewards/rejected": -2.65625, "step": 810 }, { "epoch": 0.1047990286919292, "grad_norm": 10.59961166689991, "learning_rate": 4.99965932940093e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -300.0, "logps/rejected": -470.0, "loss": 0.3963, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.5234375, "rewards/margins": 1.703125, "rewards/rejected": -3.21875, "step": 820 }, { "epoch": 0.10607706562719663, "grad_norm": 13.077749744872387, "learning_rate": 4.999450306160585e-07, "logits/chosen": -3.15625, "logits/rejected": -2.984375, "logps/chosen": -310.0, "logps/rejected": -472.0, "loss": 0.382, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -1.5546875, "rewards/margins": 1.8359375, "rewards/rejected": -3.390625, "step": 830 }, { "epoch": 0.10735510256246406, "grad_norm": 10.551892493562612, "learning_rate": 4.999191523553715e-07, "logits/chosen": -3.21875, "logits/rejected": -3.015625, "logps/chosen": -304.0, "logps/rejected": -468.0, "loss": 0.3833, "rewards/accuracies": 0.875, "rewards/chosen": -1.3984375, "rewards/margins": 1.734375, "rewards/rejected": -3.125, "step": 840 }, { "epoch": 0.10863313949773148, "grad_norm": 12.356141382859514, "learning_rate": 4.998882986732195e-07, "logits/chosen": -2.9375, "logits/rejected": -2.875, "logps/chosen": -316.0, "logps/rejected": -482.0, "loss": 0.3878, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -1.5390625, "rewards/margins": 1.578125, "rewards/rejected": -3.109375, "step": 850 }, { "epoch": 0.10991117643299891, "grad_norm": 15.20620984625719, "learning_rate": 4.998524701838414e-07, "logits/chosen": -3.0625, "logits/rejected": -3.0, "logps/chosen": -306.0, "logps/rejected": -498.0, "loss": 0.3599, "rewards/accuracies": 0.875, "rewards/chosen": -1.515625, "rewards/margins": 1.921875, "rewards/rejected": -3.4375, "step": 860 }, { "epoch": 0.11118921336826634, "grad_norm": 14.444326022124622, "learning_rate": 4.998116676005154e-07, "logits/chosen": -3.234375, "logits/rejected": -3.21875, "logps/chosen": -306.0, "logps/rejected": -452.0, "loss": 0.4075, "rewards/accuracies": 0.800000011920929, "rewards/chosen": -1.5703125, "rewards/margins": 1.6015625, "rewards/rejected": -3.171875, "step": 870 }, { "epoch": 0.11246725030353377, "grad_norm": 7.828143338550613, "learning_rate": 4.997658917355441e-07, "logits/chosen": -3.1875, "logits/rejected": -2.953125, "logps/chosen": -338.0, "logps/rejected": -500.0, "loss": 0.3833, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -1.7578125, "rewards/margins": 1.6796875, "rewards/rejected": -3.4375, "step": 880 }, { "epoch": 0.1137452872388012, "grad_norm": 18.788227330963323, "learning_rate": 4.997151435002394e-07, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -338.0, "logps/rejected": -488.0, "loss": 0.3749, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -1.5234375, "rewards/margins": 1.7109375, "rewards/rejected": -3.234375, "step": 890 }, { "epoch": 0.11502332417406863, "grad_norm": 11.597065629717148, "learning_rate": 4.996594239049032e-07, "logits/chosen": -2.96875, "logits/rejected": -2.84375, "logps/chosen": -300.0, "logps/rejected": -456.0, "loss": 0.4025, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -1.4765625, "rewards/margins": 1.59375, "rewards/rejected": -3.0625, "step": 900 }, { "epoch": 0.11630136110933606, "grad_norm": 10.476293056540518, "learning_rate": 4.995987340588082e-07, "logits/chosen": -3.109375, "logits/rejected": -2.90625, "logps/chosen": -324.0, "logps/rejected": -486.0, "loss": 0.3728, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -1.53125, "rewards/margins": 1.7890625, "rewards/rejected": -3.3125, "step": 910 }, { "epoch": 0.11757939804460349, "grad_norm": 9.355309673884802, "learning_rate": 4.995330751701755e-07, "logits/chosen": -3.25, "logits/rejected": -3.0625, "logps/chosen": -280.0, "logps/rejected": -406.0, "loss": 0.4006, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -1.359375, "rewards/margins": 1.46875, "rewards/rejected": -2.828125, "step": 920 }, { "epoch": 0.11885743497987092, "grad_norm": 10.550972596004549, "learning_rate": 4.994624485461503e-07, "logits/chosen": -3.21875, "logits/rejected": -2.9375, "logps/chosen": -310.0, "logps/rejected": -492.0, "loss": 0.3842, "rewards/accuracies": 0.8062499761581421, "rewards/chosen": -1.5859375, "rewards/margins": 1.796875, "rewards/rejected": -3.390625, "step": 930 }, { "epoch": 0.12013547191513835, "grad_norm": 7.584312388304975, "learning_rate": 4.99386855592776e-07, "logits/chosen": -2.84375, "logits/rejected": -2.890625, "logps/chosen": -306.0, "logps/rejected": -454.0, "loss": 0.3669, "rewards/accuracies": 0.8125, "rewards/chosen": -1.46875, "rewards/margins": 1.59375, "rewards/rejected": -3.0625, "step": 940 }, { "epoch": 0.12141350885040578, "grad_norm": 14.923871887916482, "learning_rate": 4.993062978149668e-07, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -308.0, "logps/rejected": -508.0, "loss": 0.3733, "rewards/accuracies": 0.875, "rewards/chosen": -1.4921875, "rewards/margins": 1.984375, "rewards/rejected": -3.46875, "step": 950 }, { "epoch": 0.1226915457856732, "grad_norm": 12.368485690167017, "learning_rate": 4.992207768164769e-07, "logits/chosen": -2.71875, "logits/rejected": -2.640625, "logps/chosen": -274.0, "logps/rejected": -396.0, "loss": 0.4143, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -1.3125, "rewards/margins": 1.296875, "rewards/rejected": -2.609375, "step": 960 }, { "epoch": 0.12396958272094064, "grad_norm": 12.872066642334858, "learning_rate": 4.991302942998686e-07, "logits/chosen": -2.953125, "logits/rejected": -2.875, "logps/chosen": -324.0, "logps/rejected": -492.0, "loss": 0.3478, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -1.71875, "rewards/margins": 1.75, "rewards/rejected": -3.46875, "step": 970 }, { "epoch": 0.12524761965620806, "grad_norm": 8.76843030472801, "learning_rate": 4.990348520664794e-07, "logits/chosen": -3.03125, "logits/rejected": -2.90625, "logps/chosen": -310.0, "logps/rejected": -438.0, "loss": 0.3752, "rewards/accuracies": 0.8125, "rewards/chosen": -1.421875, "rewards/margins": 1.5546875, "rewards/rejected": -2.96875, "step": 980 }, { "epoch": 0.1265256565914755, "grad_norm": 10.846462610025974, "learning_rate": 4.989344520163849e-07, "logits/chosen": -2.9375, "logits/rejected": -2.78125, "logps/chosen": -294.0, "logps/rejected": -452.0, "loss": 0.3621, "rewards/accuracies": 0.8062499761581421, "rewards/chosen": -1.640625, "rewards/margins": 1.53125, "rewards/rejected": -3.171875, "step": 990 }, { "epoch": 0.12780369352674292, "grad_norm": 24.815238960975694, "learning_rate": 4.98829096148362e-07, "logits/chosen": -3.0625, "logits/rejected": -2.9375, "logps/chosen": -316.0, "logps/rejected": -494.0, "loss": 0.3789, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -1.5703125, "rewards/margins": 1.8515625, "rewards/rejected": -3.421875, "step": 1000 }, { "epoch": 0.12908173046201035, "grad_norm": 12.469684746555005, "learning_rate": 4.987187865598482e-07, "logits/chosen": -3.03125, "logits/rejected": -3.046875, "logps/chosen": -312.0, "logps/rejected": -480.0, "loss": 0.354, "rewards/accuracies": 0.875, "rewards/chosen": -1.578125, "rewards/margins": 1.96875, "rewards/rejected": -3.546875, "step": 1010 }, { "epoch": 0.13035976739727778, "grad_norm": 10.651703453569814, "learning_rate": 4.986035254469005e-07, "logits/chosen": -3.09375, "logits/rejected": -2.96875, "logps/chosen": -280.0, "logps/rejected": -460.0, "loss": 0.3777, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -1.3671875, "rewards/margins": 1.84375, "rewards/rejected": -3.21875, "step": 1020 }, { "epoch": 0.1316378043325452, "grad_norm": 10.169153192152455, "learning_rate": 4.984833151041512e-07, "logits/chosen": -2.984375, "logits/rejected": -2.96875, "logps/chosen": -322.0, "logps/rejected": -524.0, "loss": 0.3529, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -1.734375, "rewards/margins": 2.0625, "rewards/rejected": -3.796875, "step": 1030 }, { "epoch": 0.13291584126781264, "grad_norm": 17.70616690114393, "learning_rate": 4.98358157924763e-07, "logits/chosen": -3.234375, "logits/rejected": -3.125, "logps/chosen": -344.0, "logps/rejected": -584.0, "loss": 0.344, "rewards/accuracies": 0.84375, "rewards/chosen": -1.9921875, "rewards/margins": 2.328125, "rewards/rejected": -4.3125, "step": 1040 }, { "epoch": 0.13419387820308007, "grad_norm": 14.910681564668302, "learning_rate": 4.982280564003806e-07, "logits/chosen": -2.90625, "logits/rejected": -2.90625, "logps/chosen": -342.0, "logps/rejected": -454.0, "loss": 0.3676, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -1.640625, "rewards/margins": 1.4140625, "rewards/rejected": -3.0625, "step": 1050 }, { "epoch": 0.1354719151383475, "grad_norm": 9.401234339612973, "learning_rate": 4.98093013121081e-07, "logits/chosen": -2.90625, "logits/rejected": -2.96875, "logps/chosen": -336.0, "logps/rejected": -488.0, "loss": 0.3497, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -1.7109375, "rewards/margins": 1.734375, "rewards/rejected": -3.453125, "step": 1060 }, { "epoch": 0.13674995207361493, "grad_norm": 11.021445735908344, "learning_rate": 4.979530307753227e-07, "logits/chosen": -2.984375, "logits/rejected": -2.9375, "logps/chosen": -324.0, "logps/rejected": -472.0, "loss": 0.3518, "rewards/accuracies": 0.8062499761581421, "rewards/chosen": -1.6328125, "rewards/margins": 1.6875, "rewards/rejected": -3.3125, "step": 1070 }, { "epoch": 0.13802798900888236, "grad_norm": 13.943751456778854, "learning_rate": 4.978081121498916e-07, "logits/chosen": -3.21875, "logits/rejected": -3.046875, "logps/chosen": -344.0, "logps/rejected": -516.0, "loss": 0.3489, "rewards/accuracies": 0.875, "rewards/chosen": -1.9609375, "rewards/margins": 1.7578125, "rewards/rejected": -3.71875, "step": 1080 }, { "epoch": 0.13930602594414979, "grad_norm": 10.338244132399032, "learning_rate": 4.976582601298456e-07, "logits/chosen": -2.953125, "logits/rejected": -2.9375, "logps/chosen": -332.0, "logps/rejected": -510.0, "loss": 0.3353, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -1.7109375, "rewards/margins": 1.9453125, "rewards/rejected": -3.65625, "step": 1090 }, { "epoch": 0.14058406287941722, "grad_norm": 8.71888191275971, "learning_rate": 4.975034776984573e-07, "logits/chosen": -3.09375, "logits/rejected": -3.015625, "logps/chosen": -338.0, "logps/rejected": -486.0, "loss": 0.3553, "rewards/accuracies": 0.8125, "rewards/chosen": -1.8203125, "rewards/margins": 1.65625, "rewards/rejected": -3.46875, "step": 1100 }, { "epoch": 0.14186209981468464, "grad_norm": 12.216205581090218, "learning_rate": 4.973437679371546e-07, "logits/chosen": -2.984375, "logits/rejected": -2.9375, "logps/chosen": -292.0, "logps/rejected": -456.0, "loss": 0.3496, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -1.5, "rewards/margins": 1.7890625, "rewards/rejected": -3.296875, "step": 1110 }, { "epoch": 0.14314013674995207, "grad_norm": 12.423722309973517, "learning_rate": 4.971791340254592e-07, "logits/chosen": -2.859375, "logits/rejected": -2.875, "logps/chosen": -320.0, "logps/rejected": -462.0, "loss": 0.3644, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -1.6171875, "rewards/margins": 1.640625, "rewards/rejected": -3.25, "step": 1120 }, { "epoch": 0.1444181736852195, "grad_norm": 11.14194864610794, "learning_rate": 4.970095792409233e-07, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -326.0, "logps/rejected": -454.0, "loss": 0.3456, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.5859375, "rewards/margins": 1.625, "rewards/rejected": -3.21875, "step": 1130 }, { "epoch": 0.14569621062048693, "grad_norm": 11.464691538377055, "learning_rate": 4.968351069590646e-07, "logits/chosen": -3.0, "logits/rejected": -2.90625, "logps/chosen": -356.0, "logps/rejected": -506.0, "loss": 0.3693, "rewards/accuracies": 0.8125, "rewards/chosen": -1.9453125, "rewards/margins": 1.703125, "rewards/rejected": -3.65625, "step": 1140 }, { "epoch": 0.14697424755575436, "grad_norm": 9.505788032958687, "learning_rate": 4.966557206532992e-07, "logits/chosen": -2.875, "logits/rejected": -2.796875, "logps/chosen": -352.0, "logps/rejected": -474.0, "loss": 0.3805, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.8515625, "rewards/margins": 1.4140625, "rewards/rejected": -3.265625, "step": 1150 }, { "epoch": 0.1482522844910218, "grad_norm": 9.689853960676805, "learning_rate": 4.964714238948715e-07, "logits/chosen": -3.078125, "logits/rejected": -2.953125, "logps/chosen": -328.0, "logps/rejected": -520.0, "loss": 0.3429, "rewards/accuracies": 0.875, "rewards/chosen": -1.7265625, "rewards/margins": 1.9609375, "rewards/rejected": -3.6875, "step": 1160 }, { "epoch": 0.14953032142628922, "grad_norm": 8.325164199298303, "learning_rate": 4.962822203527845e-07, "logits/chosen": -3.3125, "logits/rejected": -3.1875, "logps/chosen": -326.0, "logps/rejected": -564.0, "loss": 0.3277, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -1.7578125, "rewards/margins": 2.515625, "rewards/rejected": -4.25, "step": 1170 }, { "epoch": 0.15080835836155665, "grad_norm": 15.52096274687861, "learning_rate": 4.960881137937256e-07, "logits/chosen": -3.203125, "logits/rejected": -3.09375, "logps/chosen": -312.0, "logps/rejected": -492.0, "loss": 0.3235, "rewards/accuracies": 0.84375, "rewards/chosen": -1.546875, "rewards/margins": 1.921875, "rewards/rejected": -3.46875, "step": 1180 }, { "epoch": 0.15208639529682408, "grad_norm": 12.639275691252218, "learning_rate": 4.958891080819923e-07, "logits/chosen": -3.34375, "logits/rejected": -3.34375, "logps/chosen": -356.0, "logps/rejected": -552.0, "loss": 0.3541, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -2.125, "rewards/margins": 2.0, "rewards/rejected": -4.125, "step": 1190 }, { "epoch": 0.1533644322320915, "grad_norm": 8.869328760760212, "learning_rate": 4.956852071794151e-07, "logits/chosen": -3.109375, "logits/rejected": -2.953125, "logps/chosen": -324.0, "logps/rejected": -470.0, "loss": 0.3544, "rewards/accuracies": 0.875, "rewards/chosen": -1.5703125, "rewards/margins": 1.71875, "rewards/rejected": -3.28125, "step": 1200 }, { "epoch": 0.15464246916735894, "grad_norm": 10.63346925167416, "learning_rate": 4.954764151452782e-07, "logits/chosen": -3.109375, "logits/rejected": -3.046875, "logps/chosen": -332.0, "logps/rejected": -512.0, "loss": 0.3265, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -1.828125, "rewards/margins": 1.8125, "rewards/rejected": -3.640625, "step": 1210 }, { "epoch": 0.15592050610262637, "grad_norm": 11.334624897642897, "learning_rate": 4.952627361362395e-07, "logits/chosen": -3.375, "logits/rejected": -3.234375, "logps/chosen": -340.0, "logps/rejected": -540.0, "loss": 0.3435, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -1.9765625, "rewards/margins": 1.921875, "rewards/rejected": -3.90625, "step": 1220 }, { "epoch": 0.1571985430378938, "grad_norm": 11.849129116829852, "learning_rate": 4.950441744062471e-07, "logits/chosen": -3.375, "logits/rejected": -3.28125, "logps/chosen": -364.0, "logps/rejected": -564.0, "loss": 0.3435, "rewards/accuracies": 0.84375, "rewards/chosen": -2.046875, "rewards/margins": 1.9453125, "rewards/rejected": -3.984375, "step": 1230 }, { "epoch": 0.15847657997316122, "grad_norm": 9.577666300992249, "learning_rate": 4.948207343064551e-07, "logits/chosen": -3.234375, "logits/rejected": -3.21875, "logps/chosen": -374.0, "logps/rejected": -596.0, "loss": 0.3173, "rewards/accuracies": 0.875, "rewards/chosen": -2.203125, "rewards/margins": 2.34375, "rewards/rejected": -4.5625, "step": 1240 }, { "epoch": 0.15975461690842865, "grad_norm": 10.910196783813214, "learning_rate": 4.945924202851366e-07, "logits/chosen": -3.234375, "logits/rejected": -3.125, "logps/chosen": -398.0, "logps/rejected": -616.0, "loss": 0.3307, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.328125, "rewards/margins": 2.359375, "rewards/rejected": -4.6875, "step": 1250 }, { "epoch": 0.16103265384369608, "grad_norm": 10.57466378686154, "learning_rate": 4.943592368875955e-07, "logits/chosen": -3.078125, "logits/rejected": -2.875, "logps/chosen": -414.0, "logps/rejected": -592.0, "loss": 0.3491, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -2.59375, "rewards/margins": 1.921875, "rewards/rejected": -4.53125, "step": 1260 }, { "epoch": 0.1623106907789635, "grad_norm": 11.230589436335967, "learning_rate": 4.941211887560757e-07, "logits/chosen": -3.15625, "logits/rejected": -3.046875, "logps/chosen": -458.0, "logps/rejected": -668.0, "loss": 0.3023, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -2.859375, "rewards/margins": 2.234375, "rewards/rejected": -5.09375, "step": 1270 }, { "epoch": 0.16358872771423094, "grad_norm": 11.961477038376033, "learning_rate": 4.938782806296691e-07, "logits/chosen": -3.203125, "logits/rejected": -3.25, "logps/chosen": -444.0, "logps/rejected": -628.0, "loss": 0.3028, "rewards/accuracies": 0.78125, "rewards/chosen": -2.875, "rewards/margins": 2.078125, "rewards/rejected": -4.96875, "step": 1280 }, { "epoch": 0.16486676464949837, "grad_norm": 9.972580283679875, "learning_rate": 4.936305173442206e-07, "logits/chosen": -3.078125, "logits/rejected": -3.0, "logps/chosen": -412.0, "logps/rejected": -608.0, "loss": 0.3068, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.5, "rewards/margins": 2.203125, "rewards/rejected": -4.71875, "step": 1290 }, { "epoch": 0.1661448015847658, "grad_norm": 10.211364589134622, "learning_rate": 4.933779038322324e-07, "logits/chosen": -3.125, "logits/rejected": -3.078125, "logps/chosen": -420.0, "logps/rejected": -636.0, "loss": 0.3251, "rewards/accuracies": 0.875, "rewards/chosen": -2.59375, "rewards/margins": 2.234375, "rewards/rejected": -4.84375, "step": 1300 }, { "epoch": 0.16742283852003323, "grad_norm": 8.909898711500084, "learning_rate": 4.931204451227658e-07, "logits/chosen": -3.140625, "logits/rejected": -3.125, "logps/chosen": -386.0, "logps/rejected": -572.0, "loss": 0.3113, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -2.234375, "rewards/margins": 2.0, "rewards/rejected": -4.25, "step": 1310 }, { "epoch": 0.16870087545530066, "grad_norm": 13.917027005009777, "learning_rate": 4.928581463413405e-07, "logits/chosen": -3.375, "logits/rejected": -3.125, "logps/chosen": -354.0, "logps/rejected": -584.0, "loss": 0.3264, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -2.140625, "rewards/margins": 2.265625, "rewards/rejected": -4.40625, "step": 1320 }, { "epoch": 0.1699789123905681, "grad_norm": 9.075174861427325, "learning_rate": 4.925910127098333e-07, "logits/chosen": -3.1875, "logits/rejected": -3.078125, "logps/chosen": -342.0, "logps/rejected": -560.0, "loss": 0.3096, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -1.90625, "rewards/margins": 2.390625, "rewards/rejected": -4.28125, "step": 1330 }, { "epoch": 0.17125694932583552, "grad_norm": 10.492320216709347, "learning_rate": 4.923190495463736e-07, "logits/chosen": -3.171875, "logits/rejected": -3.0625, "logps/chosen": -390.0, "logps/rejected": -624.0, "loss": 0.2988, "rewards/accuracies": 0.90625, "rewards/chosen": -2.296875, "rewards/margins": 2.453125, "rewards/rejected": -4.75, "step": 1340 }, { "epoch": 0.17253498626110295, "grad_norm": 11.104892250574451, "learning_rate": 4.920422622652377e-07, "logits/chosen": -3.125, "logits/rejected": -2.96875, "logps/chosen": -376.0, "logps/rejected": -576.0, "loss": 0.3202, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -2.078125, "rewards/margins": 2.140625, "rewards/rejected": -4.21875, "step": 1350 }, { "epoch": 0.17381302319637038, "grad_norm": 8.174183970674996, "learning_rate": 4.917606563767411e-07, "logits/chosen": -3.0625, "logits/rejected": -2.953125, "logps/chosen": -420.0, "logps/rejected": -648.0, "loss": 0.3258, "rewards/accuracies": 0.90625, "rewards/chosen": -2.359375, "rewards/margins": 2.578125, "rewards/rejected": -4.9375, "step": 1360 }, { "epoch": 0.1750910601316378, "grad_norm": 9.822967964793463, "learning_rate": 4.914742374871289e-07, "logits/chosen": -3.203125, "logits/rejected": -3.046875, "logps/chosen": -440.0, "logps/rejected": -648.0, "loss": 0.2794, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.90625, "rewards/margins": 2.21875, "rewards/rejected": -5.125, "step": 1370 }, { "epoch": 0.17636909706690523, "grad_norm": 13.046396762006776, "learning_rate": 4.911830112984638e-07, "logits/chosen": -3.234375, "logits/rejected": -3.28125, "logps/chosen": -404.0, "logps/rejected": -616.0, "loss": 0.3158, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -2.515625, "rewards/margins": 2.21875, "rewards/rejected": -4.75, "step": 1380 }, { "epoch": 0.17764713400217266, "grad_norm": 9.634796199540869, "learning_rate": 4.908869836085127e-07, "logits/chosen": -3.234375, "logits/rejected": -3.109375, "logps/chosen": -400.0, "logps/rejected": -620.0, "loss": 0.2978, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -2.453125, "rewards/margins": 2.375, "rewards/rejected": -4.84375, "step": 1390 }, { "epoch": 0.1789251709374401, "grad_norm": 10.036265985004444, "learning_rate": 4.905861603106318e-07, "logits/chosen": -3.40625, "logits/rejected": -3.328125, "logps/chosen": -404.0, "logps/rejected": -644.0, "loss": 0.3018, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -2.296875, "rewards/margins": 2.609375, "rewards/rejected": -4.90625, "step": 1400 }, { "epoch": 0.18020320787270752, "grad_norm": 18.756659281595155, "learning_rate": 4.902805473936483e-07, "logits/chosen": -3.4375, "logits/rejected": -3.296875, "logps/chosen": -420.0, "logps/rejected": -676.0, "loss": 0.2875, "rewards/accuracies": 0.875, "rewards/chosen": -2.5625, "rewards/margins": 2.578125, "rewards/rejected": -5.15625, "step": 1410 }, { "epoch": 0.18148124480797495, "grad_norm": 11.083399006007372, "learning_rate": 4.899701509417423e-07, "logits/chosen": -3.21875, "logits/rejected": -3.203125, "logps/chosen": -418.0, "logps/rejected": -664.0, "loss": 0.2897, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -2.609375, "rewards/margins": 2.546875, "rewards/rejected": -5.15625, "step": 1420 }, { "epoch": 0.18275928174324238, "grad_norm": 9.072585092087222, "learning_rate": 4.896549771343247e-07, "logits/chosen": -3.390625, "logits/rejected": -3.171875, "logps/chosen": -456.0, "logps/rejected": -732.0, "loss": 0.2857, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.109375, "rewards/margins": 2.828125, "rewards/rejected": -5.9375, "step": 1430 }, { "epoch": 0.1840373186785098, "grad_norm": 7.519453357476137, "learning_rate": 4.893350322459149e-07, "logits/chosen": -3.078125, "logits/rejected": -3.03125, "logps/chosen": -444.0, "logps/rejected": -676.0, "loss": 0.2972, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -2.71875, "rewards/margins": 2.5, "rewards/rejected": -5.21875, "step": 1440 }, { "epoch": 0.18531535561377724, "grad_norm": 10.926070436917199, "learning_rate": 4.890103226460152e-07, "logits/chosen": -3.078125, "logits/rejected": -2.921875, "logps/chosen": -444.0, "logps/rejected": -656.0, "loss": 0.3027, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -2.828125, "rewards/margins": 2.21875, "rewards/rejected": -5.0625, "step": 1450 }, { "epoch": 0.18659339254904467, "grad_norm": 9.889333734788899, "learning_rate": 4.886808547989846e-07, "logits/chosen": -3.3125, "logits/rejected": -3.1875, "logps/chosen": -504.0, "logps/rejected": -700.0, "loss": 0.3162, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -3.515625, "rewards/margins": 2.109375, "rewards/rejected": -5.625, "step": 1460 }, { "epoch": 0.1878714294843121, "grad_norm": 11.48343624861157, "learning_rate": 4.883466352639099e-07, "logits/chosen": -3.296875, "logits/rejected": -3.234375, "logps/chosen": -440.0, "logps/rejected": -656.0, "loss": 0.3164, "rewards/accuracies": 0.84375, "rewards/chosen": -2.625, "rewards/margins": 2.359375, "rewards/rejected": -4.96875, "step": 1470 }, { "epoch": 0.18914946641957953, "grad_norm": 11.028607347761977, "learning_rate": 4.88007670694475e-07, "logits/chosen": -3.421875, "logits/rejected": -3.21875, "logps/chosen": -440.0, "logps/rejected": -672.0, "loss": 0.292, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -2.734375, "rewards/margins": 2.5, "rewards/rejected": -5.25, "step": 1480 }, { "epoch": 0.19042750335484696, "grad_norm": 9.137715141189384, "learning_rate": 4.876639678388285e-07, "logits/chosen": -3.21875, "logits/rejected": -3.0, "logps/chosen": -420.0, "logps/rejected": -628.0, "loss": 0.3172, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -2.515625, "rewards/margins": 2.3125, "rewards/rejected": -4.84375, "step": 1490 }, { "epoch": 0.19170554029011438, "grad_norm": 8.392316890933923, "learning_rate": 4.873155335394497e-07, "logits/chosen": -3.234375, "logits/rejected": -3.0625, "logps/chosen": -430.0, "logps/rejected": -656.0, "loss": 0.2895, "rewards/accuracies": 0.84375, "rewards/chosen": -2.765625, "rewards/margins": 2.453125, "rewards/rejected": -5.21875, "step": 1500 }, { "epoch": 0.1929835772253818, "grad_norm": 9.776765970586137, "learning_rate": 4.869623747330116e-07, "logits/chosen": -3.265625, "logits/rejected": -3.09375, "logps/chosen": -474.0, "logps/rejected": -696.0, "loss": 0.3101, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.15625, "rewards/margins": 2.578125, "rewards/rejected": -5.75, "step": 1510 }, { "epoch": 0.19426161416064924, "grad_norm": 10.113772712494775, "learning_rate": 4.866044984502436e-07, "logits/chosen": -3.015625, "logits/rejected": -2.921875, "logps/chosen": -434.0, "logps/rejected": -696.0, "loss": 0.2819, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.765625, "rewards/margins": 2.5625, "rewards/rejected": -5.34375, "step": 1520 }, { "epoch": 0.19553965109591667, "grad_norm": 12.373484208276247, "learning_rate": 4.862419118157909e-07, "logits/chosen": -3.15625, "logits/rejected": -3.0, "logps/chosen": -424.0, "logps/rejected": -648.0, "loss": 0.2809, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.75, "rewards/margins": 2.328125, "rewards/rejected": -5.09375, "step": 1530 }, { "epoch": 0.1968176880311841, "grad_norm": 11.78683291379729, "learning_rate": 4.858746220480734e-07, "logits/chosen": -3.234375, "logits/rejected": -3.140625, "logps/chosen": -412.0, "logps/rejected": -632.0, "loss": 0.3016, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -2.453125, "rewards/margins": 2.4375, "rewards/rejected": -4.875, "step": 1540 }, { "epoch": 0.19809572496645153, "grad_norm": 12.160856594312754, "learning_rate": 4.855026364591413e-07, "logits/chosen": -3.375, "logits/rejected": -3.234375, "logps/chosen": -482.0, "logps/rejected": -724.0, "loss": 0.2759, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.078125, "rewards/margins": 2.53125, "rewards/rejected": -5.625, "step": 1550 }, { "epoch": 0.19937376190171896, "grad_norm": 7.635186752817528, "learning_rate": 4.851259624545297e-07, "logits/chosen": -3.296875, "logits/rejected": -3.203125, "logps/chosen": -470.0, "logps/rejected": -700.0, "loss": 0.2999, "rewards/accuracies": 0.90625, "rewards/chosen": -3.0625, "rewards/margins": 2.421875, "rewards/rejected": -5.46875, "step": 1560 }, { "epoch": 0.2006517988369864, "grad_norm": 10.580854781397509, "learning_rate": 4.847446075331114e-07, "logits/chosen": -3.28125, "logits/rejected": -3.046875, "logps/chosen": -436.0, "logps/rejected": -672.0, "loss": 0.298, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -2.8125, "rewards/margins": 2.453125, "rewards/rejected": -5.28125, "step": 1570 }, { "epoch": 0.20192983577225382, "grad_norm": 15.508397767776305, "learning_rate": 4.843585792869476e-07, "logits/chosen": -3.375, "logits/rejected": -3.265625, "logps/chosen": -532.0, "logps/rejected": -796.0, "loss": 0.3035, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -3.875, "rewards/margins": 2.765625, "rewards/rejected": -6.625, "step": 1580 }, { "epoch": 0.20320787270752125, "grad_norm": 12.29502011282903, "learning_rate": 4.839678854011362e-07, "logits/chosen": -3.21875, "logits/rejected": -3.15625, "logps/chosen": -484.0, "logps/rejected": -680.0, "loss": 0.3106, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -3.21875, "rewards/margins": 2.109375, "rewards/rejected": -5.34375, "step": 1590 }, { "epoch": 0.20448590964278868, "grad_norm": 9.221266332248716, "learning_rate": 4.835725336536598e-07, "logits/chosen": -3.328125, "logits/rejected": -3.25, "logps/chosen": -476.0, "logps/rejected": -724.0, "loss": 0.2795, "rewards/accuracies": 0.90625, "rewards/chosen": -3.265625, "rewards/margins": 2.609375, "rewards/rejected": -5.875, "step": 1600 }, { "epoch": 0.2057639465780561, "grad_norm": 21.949201663234852, "learning_rate": 4.831725319152298e-07, "logits/chosen": -3.328125, "logits/rejected": -3.359375, "logps/chosen": -496.0, "logps/rejected": -772.0, "loss": 0.2822, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.484375, "rewards/margins": 2.796875, "rewards/rejected": -6.28125, "step": 1610 }, { "epoch": 0.20704198351332354, "grad_norm": 8.390379977679709, "learning_rate": 4.827678881491305e-07, "logits/chosen": -3.25, "logits/rejected": -3.125, "logps/chosen": -496.0, "logps/rejected": -780.0, "loss": 0.2692, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -3.359375, "rewards/margins": 2.65625, "rewards/rejected": -6.0, "step": 1620 }, { "epoch": 0.20832002044859096, "grad_norm": 9.085937257418248, "learning_rate": 4.8235861041106e-07, "logits/chosen": -3.0, "logits/rejected": -2.96875, "logps/chosen": -502.0, "logps/rejected": -716.0, "loss": 0.2757, "rewards/accuracies": 0.84375, "rewards/chosen": -3.46875, "rewards/margins": 2.3125, "rewards/rejected": -5.78125, "step": 1630 }, { "epoch": 0.2095980573838584, "grad_norm": 10.915204034593046, "learning_rate": 4.819447068489705e-07, "logits/chosen": -3.15625, "logits/rejected": -3.0625, "logps/chosen": -516.0, "logps/rejected": -752.0, "loss": 0.2838, "rewards/accuracies": 0.84375, "rewards/chosen": -3.46875, "rewards/margins": 2.484375, "rewards/rejected": -5.96875, "step": 1640 }, { "epoch": 0.21087609431912582, "grad_norm": 10.623276883304097, "learning_rate": 4.815261857029052e-07, "logits/chosen": -3.125, "logits/rejected": -3.140625, "logps/chosen": -462.0, "logps/rejected": -744.0, "loss": 0.2661, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.0625, "rewards/margins": 3.09375, "rewards/rejected": -6.1875, "step": 1650 }, { "epoch": 0.21215413125439325, "grad_norm": 15.529572832655484, "learning_rate": 4.811030553048351e-07, "logits/chosen": -3.25, "logits/rejected": -3.0625, "logps/chosen": -464.0, "logps/rejected": -696.0, "loss": 0.2895, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -2.96875, "rewards/margins": 2.484375, "rewards/rejected": -5.4375, "step": 1660 }, { "epoch": 0.21343216818966068, "grad_norm": 8.986111656832815, "learning_rate": 4.806753240784924e-07, "logits/chosen": -3.09375, "logits/rejected": -2.890625, "logps/chosen": -452.0, "logps/rejected": -724.0, "loss": 0.2822, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -2.796875, "rewards/margins": 2.8125, "rewards/rejected": -5.59375, "step": 1670 }, { "epoch": 0.2147102051249281, "grad_norm": 10.662265185336627, "learning_rate": 4.802430005392037e-07, "logits/chosen": -3.140625, "logits/rejected": -3.15625, "logps/chosen": -434.0, "logps/rejected": -668.0, "loss": 0.3085, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -2.96875, "rewards/margins": 2.53125, "rewards/rejected": -5.5, "step": 1680 }, { "epoch": 0.21598824206019554, "grad_norm": 18.32782419922393, "learning_rate": 4.798060932937194e-07, "logits/chosen": -2.984375, "logits/rejected": -3.0625, "logps/chosen": -464.0, "logps/rejected": -684.0, "loss": 0.2825, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.0, "rewards/margins": 2.453125, "rewards/rejected": -5.46875, "step": 1690 }, { "epoch": 0.21726627899546297, "grad_norm": 8.37402052085016, "learning_rate": 4.79364611040043e-07, "logits/chosen": -3.046875, "logits/rejected": -2.984375, "logps/chosen": -456.0, "logps/rejected": -756.0, "loss": 0.2776, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.0, "rewards/margins": 3.15625, "rewards/rejected": -6.15625, "step": 1700 }, { "epoch": 0.2185443159307304, "grad_norm": 7.388938797100764, "learning_rate": 4.789185625672583e-07, "logits/chosen": -3.03125, "logits/rejected": -2.96875, "logps/chosen": -510.0, "logps/rejected": -736.0, "loss": 0.2688, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.484375, "rewards/margins": 2.421875, "rewards/rejected": -5.90625, "step": 1710 }, { "epoch": 0.21982235286599783, "grad_norm": 11.18964791369844, "learning_rate": 4.784679567553531e-07, "logits/chosen": -3.3125, "logits/rejected": -3.140625, "logps/chosen": -470.0, "logps/rejected": -780.0, "loss": 0.2807, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.125, "rewards/margins": 3.234375, "rewards/rejected": -6.375, "step": 1720 }, { "epoch": 0.22110038980126526, "grad_norm": 12.788195666651463, "learning_rate": 4.780128025750442e-07, "logits/chosen": -3.09375, "logits/rejected": -3.0, "logps/chosen": -466.0, "logps/rejected": -716.0, "loss": 0.2667, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.15625, "rewards/margins": 2.578125, "rewards/rejected": -5.75, "step": 1730 }, { "epoch": 0.22237842673653269, "grad_norm": 13.452879203275968, "learning_rate": 4.775531090875971e-07, "logits/chosen": -3.15625, "logits/rejected": -3.015625, "logps/chosen": -504.0, "logps/rejected": -724.0, "loss": 0.2936, "rewards/accuracies": 0.84375, "rewards/chosen": -3.390625, "rewards/margins": 2.34375, "rewards/rejected": -5.75, "step": 1740 }, { "epoch": 0.22365646367180012, "grad_norm": 13.978917804241819, "learning_rate": 4.770888854446471e-07, "logits/chosen": -3.078125, "logits/rejected": -3.09375, "logps/chosen": -462.0, "logps/rejected": -696.0, "loss": 0.2665, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -3.09375, "rewards/margins": 2.625, "rewards/rejected": -5.71875, "step": 1750 }, { "epoch": 0.22493450060706754, "grad_norm": 14.827300418751848, "learning_rate": 4.766201408880156e-07, "logits/chosen": -3.09375, "logits/rejected": -3.09375, "logps/chosen": -556.0, "logps/rejected": -752.0, "loss": 0.3005, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.875, "rewards/margins": 2.25, "rewards/rejected": -6.125, "step": 1760 }, { "epoch": 0.22621253754233497, "grad_norm": 8.780665034348093, "learning_rate": 4.761468847495277e-07, "logits/chosen": -3.0, "logits/rejected": -2.796875, "logps/chosen": -516.0, "logps/rejected": -748.0, "loss": 0.268, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -3.46875, "rewards/margins": 2.46875, "rewards/rejected": -5.9375, "step": 1770 }, { "epoch": 0.2274905744776024, "grad_norm": 8.913295529168444, "learning_rate": 4.756691264508251e-07, "logits/chosen": -3.078125, "logits/rejected": -3.078125, "logps/chosen": -468.0, "logps/rejected": -740.0, "loss": 0.2597, "rewards/accuracies": 0.84375, "rewards/chosen": -3.140625, "rewards/margins": 2.703125, "rewards/rejected": -5.84375, "step": 1780 }, { "epoch": 0.22876861141286983, "grad_norm": 9.149720146692632, "learning_rate": 4.751868755031793e-07, "logits/chosen": -3.3125, "logits/rejected": -3.234375, "logps/chosen": -528.0, "logps/rejected": -880.0, "loss": 0.2978, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -3.796875, "rewards/margins": 3.5, "rewards/rejected": -7.3125, "step": 1790 }, { "epoch": 0.23004664834813726, "grad_norm": 9.864161436507205, "learning_rate": 4.747001415073018e-07, "logits/chosen": -3.203125, "logits/rejected": -3.1875, "logps/chosen": -488.0, "logps/rejected": -684.0, "loss": 0.2801, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.171875, "rewards/margins": 2.25, "rewards/rejected": -5.40625, "step": 1800 }, { "epoch": 0.2313246852834047, "grad_norm": 12.049630916813516, "learning_rate": 4.742089341531535e-07, "logits/chosen": -3.15625, "logits/rejected": -3.109375, "logps/chosen": -488.0, "logps/rejected": -744.0, "loss": 0.2725, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.3125, "rewards/margins": 2.75, "rewards/rejected": -6.0625, "step": 1810 }, { "epoch": 0.23260272221867212, "grad_norm": 9.245650817902728, "learning_rate": 4.7371326321975127e-07, "logits/chosen": -3.203125, "logits/rejected": -3.09375, "logps/chosen": -496.0, "logps/rejected": -800.0, "loss": 0.2838, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.484375, "rewards/margins": 3.03125, "rewards/rejected": -6.53125, "step": 1820 }, { "epoch": 0.23388075915393955, "grad_norm": 10.565374823511835, "learning_rate": 4.7321313857497336e-07, "logits/chosen": -2.984375, "logits/rejected": -3.0625, "logps/chosen": -544.0, "logps/rejected": -832.0, "loss": 0.3001, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.90625, "rewards/margins": 2.84375, "rewards/rejected": -6.75, "step": 1830 }, { "epoch": 0.23515879608920698, "grad_norm": 11.728663046788858, "learning_rate": 4.7270857017536335e-07, "logits/chosen": -2.90625, "logits/rejected": -2.8125, "logps/chosen": -544.0, "logps/rejected": -836.0, "loss": 0.2755, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.875, "rewards/margins": 3.0625, "rewards/rejected": -6.9375, "step": 1840 }, { "epoch": 0.2364368330244744, "grad_norm": 13.400694062249938, "learning_rate": 4.7219956806593143e-07, "logits/chosen": -3.046875, "logits/rejected": -2.9375, "logps/chosen": -604.0, "logps/rejected": -848.0, "loss": 0.2901, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.40625, "rewards/margins": 2.671875, "rewards/rejected": -7.0625, "step": 1850 }, { "epoch": 0.23771486995974184, "grad_norm": 8.737541772532301, "learning_rate": 4.716861423799546e-07, "logits/chosen": -2.859375, "logits/rejected": -2.828125, "logps/chosen": -612.0, "logps/rejected": -832.0, "loss": 0.2871, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -4.40625, "rewards/margins": 2.546875, "rewards/rejected": -6.9375, "step": 1860 }, { "epoch": 0.23899290689500927, "grad_norm": 11.454317171770168, "learning_rate": 4.711683033387752e-07, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -572.0, "logps/rejected": -872.0, "loss": 0.2612, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.1875, "rewards/margins": 2.875, "rewards/rejected": -7.0625, "step": 1870 }, { "epoch": 0.2402709438302767, "grad_norm": 8.019796544220595, "learning_rate": 4.70646061251597e-07, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -556.0, "logps/rejected": -848.0, "loss": 0.2603, "rewards/accuracies": 0.90625, "rewards/chosen": -3.953125, "rewards/margins": 3.046875, "rewards/rejected": -7.0, "step": 1880 }, { "epoch": 0.24154898076554412, "grad_norm": 9.374363546593047, "learning_rate": 4.701194265152802e-07, "logits/chosen": -2.796875, "logits/rejected": -2.75, "logps/chosen": -490.0, "logps/rejected": -740.0, "loss": 0.2726, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.375, "rewards/margins": 2.734375, "rewards/rejected": -6.125, "step": 1890 }, { "epoch": 0.24282701770081155, "grad_norm": 13.860826584264853, "learning_rate": 4.6958840961413447e-07, "logits/chosen": -2.890625, "logits/rejected": -2.8125, "logps/chosen": -506.0, "logps/rejected": -724.0, "loss": 0.2811, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -3.546875, "rewards/margins": 2.4375, "rewards/rejected": -6.0, "step": 1900 }, { "epoch": 0.24410505463607898, "grad_norm": 10.57554715572998, "learning_rate": 4.6905302111971004e-07, "logits/chosen": -3.109375, "logits/rejected": -3.03125, "logps/chosen": -428.0, "logps/rejected": -712.0, "loss": 0.2829, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -2.609375, "rewards/margins": 3.0, "rewards/rejected": -5.59375, "step": 1910 }, { "epoch": 0.2453830915713464, "grad_norm": 14.259221382970665, "learning_rate": 4.6851327169058735e-07, "logits/chosen": -3.21875, "logits/rejected": -3.03125, "logps/chosen": -516.0, "logps/rejected": -732.0, "loss": 0.2912, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -3.390625, "rewards/margins": 2.484375, "rewards/rejected": -5.875, "step": 1920 }, { "epoch": 0.24666112850661384, "grad_norm": 7.366111663404775, "learning_rate": 4.679691720721651e-07, "logits/chosen": -3.171875, "logits/rejected": -2.984375, "logps/chosen": -540.0, "logps/rejected": -740.0, "loss": 0.2693, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -3.734375, "rewards/margins": 2.28125, "rewards/rejected": -6.0, "step": 1930 }, { "epoch": 0.24793916544188127, "grad_norm": 11.096803393586505, "learning_rate": 4.674207330964458e-07, "logits/chosen": -3.15625, "logits/rejected": -3.0625, "logps/chosen": -564.0, "logps/rejected": -856.0, "loss": 0.2789, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.09375, "rewards/margins": 2.96875, "rewards/rejected": -7.0625, "step": 1940 }, { "epoch": 0.2492172023771487, "grad_norm": 8.956111080301984, "learning_rate": 4.668679656818207e-07, "logits/chosen": -3.046875, "logits/rejected": -2.9375, "logps/chosen": -576.0, "logps/rejected": -840.0, "loss": 0.2746, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.1875, "rewards/margins": 2.71875, "rewards/rejected": -6.90625, "step": 1950 }, { "epoch": 0.25049523931241613, "grad_norm": 9.513508988130903, "learning_rate": 4.663108808328519e-07, "logits/chosen": -2.984375, "logits/rejected": -2.90625, "logps/chosen": -596.0, "logps/rejected": -804.0, "loss": 0.2819, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -4.4375, "rewards/margins": 2.296875, "rewards/rejected": -6.71875, "step": 1960 }, { "epoch": 0.25177327624768353, "grad_norm": 10.234008316257968, "learning_rate": 4.6574948964005367e-07, "logits/chosen": -3.0, "logits/rejected": -2.984375, "logps/chosen": -584.0, "logps/rejected": -816.0, "loss": 0.2871, "rewards/accuracies": 0.875, "rewards/chosen": -4.34375, "rewards/margins": 2.546875, "rewards/rejected": -6.875, "step": 1970 }, { "epoch": 0.253051313182951, "grad_norm": 8.01274333751014, "learning_rate": 4.6518380327967145e-07, "logits/chosen": -3.203125, "logits/rejected": -3.1875, "logps/chosen": -564.0, "logps/rejected": -820.0, "loss": 0.247, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.1875, "rewards/margins": 2.515625, "rewards/rejected": -6.71875, "step": 1980 }, { "epoch": 0.2543293501182184, "grad_norm": 12.33132352378727, "learning_rate": 4.6461383301345947e-07, "logits/chosen": -3.328125, "logits/rejected": -3.265625, "logps/chosen": -548.0, "logps/rejected": -824.0, "loss": 0.2821, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.96875, "rewards/margins": 2.703125, "rewards/rejected": -6.6875, "step": 1990 }, { "epoch": 0.25560738705348585, "grad_norm": 10.395762625615612, "learning_rate": 4.6403959018845637e-07, "logits/chosen": -3.109375, "logits/rejected": -3.09375, "logps/chosen": -524.0, "logps/rejected": -804.0, "loss": 0.2645, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.703125, "rewards/margins": 2.8125, "rewards/rejected": -6.5, "step": 2000 }, { "epoch": 0.25688542398875325, "grad_norm": 15.793711134680974, "learning_rate": 4.6346108623675954e-07, "logits/chosen": -2.984375, "logits/rejected": -2.875, "logps/chosen": -544.0, "logps/rejected": -820.0, "loss": 0.2914, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.8125, "rewards/margins": 2.90625, "rewards/rejected": -6.71875, "step": 2010 }, { "epoch": 0.2581634609240207, "grad_norm": 10.102379635568527, "learning_rate": 4.628783326752974e-07, "logits/chosen": -3.125, "logits/rejected": -3.078125, "logps/chosen": -544.0, "logps/rejected": -788.0, "loss": 0.261, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.828125, "rewards/margins": 2.609375, "rewards/rejected": -6.4375, "step": 2020 }, { "epoch": 0.2594414978592881, "grad_norm": 11.19917017437885, "learning_rate": 4.622913411056e-07, "logits/chosen": -3.25, "logits/rejected": -3.171875, "logps/chosen": -572.0, "logps/rejected": -944.0, "loss": 0.2563, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.25, "rewards/margins": 3.765625, "rewards/rejected": -8.0, "step": 2030 }, { "epoch": 0.26071953479455556, "grad_norm": 9.591252819586105, "learning_rate": 4.617001232135684e-07, "logits/chosen": -3.203125, "logits/rejected": -3.046875, "logps/chosen": -496.0, "logps/rejected": -776.0, "loss": 0.2804, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -3.5, "rewards/margins": 2.828125, "rewards/rejected": -6.3125, "step": 2040 }, { "epoch": 0.26199757172982296, "grad_norm": 10.927105525570921, "learning_rate": 4.6110469076924153e-07, "logits/chosen": -3.125, "logits/rejected": -3.0625, "logps/chosen": -532.0, "logps/rejected": -804.0, "loss": 0.2661, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.5625, "rewards/margins": 3.0, "rewards/rejected": -6.5625, "step": 2050 }, { "epoch": 0.2632756086650904, "grad_norm": 8.081281860104301, "learning_rate": 4.605050556265624e-07, "logits/chosen": -2.9375, "logits/rejected": -2.9375, "logps/chosen": -572.0, "logps/rejected": -800.0, "loss": 0.2664, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.0, "rewards/margins": 2.5, "rewards/rejected": -6.5, "step": 2060 }, { "epoch": 0.2645536456003578, "grad_norm": 8.719157157796769, "learning_rate": 4.599012297231417e-07, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -490.0, "logps/rejected": -804.0, "loss": 0.2605, "rewards/accuracies": 0.90625, "rewards/chosen": -3.328125, "rewards/margins": 3.015625, "rewards/rejected": -6.34375, "step": 2070 }, { "epoch": 0.2658316825356253, "grad_norm": 9.889601643083903, "learning_rate": 4.5929322508002045e-07, "logits/chosen": -3.125, "logits/rejected": -2.828125, "logps/chosen": -532.0, "logps/rejected": -776.0, "loss": 0.264, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -3.609375, "rewards/margins": 2.90625, "rewards/rejected": -6.53125, "step": 2080 }, { "epoch": 0.2671097194708927, "grad_norm": 9.01425592242334, "learning_rate": 4.586810538014304e-07, "logits/chosen": -3.09375, "logits/rejected": -3.109375, "logps/chosen": -484.0, "logps/rejected": -740.0, "loss": 0.2496, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -3.28125, "rewards/margins": 2.8125, "rewards/rejected": -6.09375, "step": 2090 }, { "epoch": 0.26838775640616014, "grad_norm": 9.917816992815437, "learning_rate": 4.580647280745532e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -490.0, "logps/rejected": -744.0, "loss": 0.2651, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.3125, "rewards/margins": 2.75, "rewards/rejected": -6.0625, "step": 2100 }, { "epoch": 0.26966579334142754, "grad_norm": 9.08237633986445, "learning_rate": 4.5744426016927783e-07, "logits/chosen": -2.859375, "logits/rejected": -2.796875, "logps/chosen": -596.0, "logps/rejected": -840.0, "loss": 0.2618, "rewards/accuracies": 0.8187500238418579, "rewards/chosen": -4.34375, "rewards/margins": 2.6875, "rewards/rejected": -7.03125, "step": 2110 }, { "epoch": 0.270943830276695, "grad_norm": 10.374914225435086, "learning_rate": 4.5681966243795645e-07, "logits/chosen": -2.8125, "logits/rejected": -2.6875, "logps/chosen": -516.0, "logps/rejected": -832.0, "loss": 0.2563, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -3.65625, "rewards/margins": 3.09375, "rewards/rejected": -6.75, "step": 2120 }, { "epoch": 0.2722218672119624, "grad_norm": 9.39552251607615, "learning_rate": 4.5619094731515783e-07, "logits/chosen": -3.046875, "logits/rejected": -2.90625, "logps/chosen": -520.0, "logps/rejected": -812.0, "loss": 0.2488, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.515625, "rewards/margins": 3.21875, "rewards/rejected": -6.71875, "step": 2130 }, { "epoch": 0.27349990414722986, "grad_norm": 9.255807709157049, "learning_rate": 4.5555812731742085e-07, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -512.0, "logps/rejected": -784.0, "loss": 0.2436, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.71875, "rewards/margins": 2.71875, "rewards/rejected": -6.4375, "step": 2140 }, { "epoch": 0.27477794108249726, "grad_norm": 13.023248862747112, "learning_rate": 4.549212150430042e-07, "logits/chosen": -3.078125, "logits/rejected": -2.96875, "logps/chosen": -506.0, "logps/rejected": -788.0, "loss": 0.2483, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.453125, "rewards/margins": 2.921875, "rewards/rejected": -6.375, "step": 2150 }, { "epoch": 0.2760559780177647, "grad_norm": 10.069695705444243, "learning_rate": 4.5428022317163654e-07, "logits/chosen": -3.09375, "logits/rejected": -3.015625, "logps/chosen": -504.0, "logps/rejected": -812.0, "loss": 0.2538, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -3.390625, "rewards/margins": 3.140625, "rewards/rejected": -6.53125, "step": 2160 }, { "epoch": 0.2773340149530321, "grad_norm": 7.905785619069278, "learning_rate": 4.5363516446426353e-07, "logits/chosen": -3.0625, "logits/rejected": -2.953125, "logps/chosen": -476.0, "logps/rejected": -716.0, "loss": 0.2553, "rewards/accuracies": 0.875, "rewards/chosen": -3.21875, "rewards/margins": 2.59375, "rewards/rejected": -5.8125, "step": 2170 }, { "epoch": 0.27861205188829957, "grad_norm": 14.110926580452388, "learning_rate": 4.5298605176279383e-07, "logits/chosen": -2.984375, "logits/rejected": -2.84375, "logps/chosen": -506.0, "logps/rejected": -808.0, "loss": 0.2546, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.40625, "rewards/margins": 3.171875, "rewards/rejected": -6.59375, "step": 2180 }, { "epoch": 0.279890088823567, "grad_norm": 11.437656708234213, "learning_rate": 4.5233289798984355e-07, "logits/chosen": -3.15625, "logits/rejected": -2.9375, "logps/chosen": -506.0, "logps/rejected": -772.0, "loss": 0.2515, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.46875, "rewards/margins": 2.8125, "rewards/rejected": -6.28125, "step": 2190 }, { "epoch": 0.28116812575883443, "grad_norm": 10.92037760467791, "learning_rate": 4.51675716148479e-07, "logits/chosen": -3.015625, "logits/rejected": -2.90625, "logps/chosen": -468.0, "logps/rejected": -744.0, "loss": 0.27, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.0625, "rewards/margins": 2.984375, "rewards/rejected": -6.03125, "step": 2200 }, { "epoch": 0.28244616269410183, "grad_norm": 9.082318419309797, "learning_rate": 4.510145193219577e-07, "logits/chosen": -3.15625, "logits/rejected": -3.125, "logps/chosen": -580.0, "logps/rejected": -852.0, "loss": 0.2611, "rewards/accuracies": 0.90625, "rewards/chosen": -4.15625, "rewards/margins": 2.875, "rewards/rejected": -7.03125, "step": 2210 }, { "epoch": 0.2837241996293693, "grad_norm": 9.696968012792881, "learning_rate": 4.503493206734681e-07, "logits/chosen": -3.234375, "logits/rejected": -3.171875, "logps/chosen": -470.0, "logps/rejected": -732.0, "loss": 0.2665, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.078125, "rewards/margins": 2.78125, "rewards/rejected": -5.875, "step": 2220 }, { "epoch": 0.2850022365646367, "grad_norm": 8.32597970165281, "learning_rate": 4.4968013344586723e-07, "logits/chosen": -3.234375, "logits/rejected": -3.140625, "logps/chosen": -548.0, "logps/rejected": -804.0, "loss": 0.2618, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -3.859375, "rewards/margins": 2.78125, "rewards/rejected": -6.65625, "step": 2230 }, { "epoch": 0.28628027349990415, "grad_norm": 9.205123244163184, "learning_rate": 4.4900697096141754e-07, "logits/chosen": -3.03125, "logits/rejected": -3.015625, "logps/chosen": -536.0, "logps/rejected": -780.0, "loss": 0.2936, "rewards/accuracies": 0.84375, "rewards/chosen": -3.796875, "rewards/margins": 2.40625, "rewards/rejected": -6.1875, "step": 2240 }, { "epoch": 0.28755831043517155, "grad_norm": 8.173644269044605, "learning_rate": 4.483298466215211e-07, "logits/chosen": -2.859375, "logits/rejected": -2.8125, "logps/chosen": -510.0, "logps/rejected": -760.0, "loss": 0.2481, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.484375, "rewards/margins": 2.703125, "rewards/rejected": -6.1875, "step": 2250 }, { "epoch": 0.288836347370439, "grad_norm": 13.842959897880784, "learning_rate": 4.4764877390645317e-07, "logits/chosen": -2.984375, "logits/rejected": -2.984375, "logps/chosen": -540.0, "logps/rejected": -824.0, "loss": 0.2398, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.96875, "rewards/margins": 2.90625, "rewards/rejected": -6.875, "step": 2260 }, { "epoch": 0.2901143843057064, "grad_norm": 10.298507286958628, "learning_rate": 4.469637663750939e-07, "logits/chosen": -3.0625, "logits/rejected": -3.0625, "logps/chosen": -568.0, "logps/rejected": -804.0, "loss": 0.2618, "rewards/accuracies": 0.90625, "rewards/chosen": -3.984375, "rewards/margins": 2.75, "rewards/rejected": -6.75, "step": 2270 }, { "epoch": 0.29139242124097386, "grad_norm": 11.781378177678372, "learning_rate": 4.4627483766465813e-07, "logits/chosen": -2.953125, "logits/rejected": -2.984375, "logps/chosen": -576.0, "logps/rejected": -840.0, "loss": 0.2419, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.1875, "rewards/margins": 2.828125, "rewards/rejected": -7.0, "step": 2280 }, { "epoch": 0.29267045817624127, "grad_norm": 9.975143928231438, "learning_rate": 4.4558200149042393e-07, "logits/chosen": -3.0625, "logits/rejected": -3.015625, "logps/chosen": -572.0, "logps/rejected": -848.0, "loss": 0.2498, "rewards/accuracies": 0.875, "rewards/chosen": -3.953125, "rewards/margins": 2.875, "rewards/rejected": -6.8125, "step": 2290 }, { "epoch": 0.2939484951115087, "grad_norm": 9.81920229020163, "learning_rate": 4.4488527164545976e-07, "logits/chosen": -3.0625, "logits/rejected": -3.0625, "logps/chosen": -580.0, "logps/rejected": -880.0, "loss": 0.2174, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.125, "rewards/margins": 3.15625, "rewards/rejected": -7.28125, "step": 2300 }, { "epoch": 0.2952265320467761, "grad_norm": 8.085865328061837, "learning_rate": 4.4418466200034974e-07, "logits/chosen": -3.265625, "logits/rejected": -3.140625, "logps/chosen": -592.0, "logps/rejected": -888.0, "loss": 0.2624, "rewards/accuracies": 0.875, "rewards/chosen": -4.46875, "rewards/margins": 3.140625, "rewards/rejected": -7.625, "step": 2310 }, { "epoch": 0.2965045689820436, "grad_norm": 10.263278935222155, "learning_rate": 4.4348018650291764e-07, "logits/chosen": -3.125, "logits/rejected": -2.828125, "logps/chosen": -496.0, "logps/rejected": -756.0, "loss": 0.2542, "rewards/accuracies": 0.90625, "rewards/chosen": -3.265625, "rewards/margins": 2.9375, "rewards/rejected": -6.21875, "step": 2320 }, { "epoch": 0.297782605917311, "grad_norm": 8.200991204991244, "learning_rate": 4.427718591779489e-07, "logits/chosen": -3.1875, "logits/rejected": -2.984375, "logps/chosen": -484.0, "logps/rejected": -780.0, "loss": 0.2513, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.375, "rewards/margins": 2.890625, "rewards/rejected": -6.25, "step": 2330 }, { "epoch": 0.29906064285257844, "grad_norm": 9.045676093918619, "learning_rate": 4.4205969412691167e-07, "logits/chosen": -3.09375, "logits/rejected": -2.984375, "logps/chosen": -568.0, "logps/rejected": -824.0, "loss": 0.2433, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -4.0625, "rewards/margins": 2.765625, "rewards/rejected": -6.8125, "step": 2340 }, { "epoch": 0.30033867978784584, "grad_norm": 9.307362421836325, "learning_rate": 4.4134370552767617e-07, "logits/chosen": -3.25, "logits/rejected": -3.125, "logps/chosen": -592.0, "logps/rejected": -884.0, "loss": 0.2511, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.21875, "rewards/margins": 3.078125, "rewards/rejected": -7.28125, "step": 2350 }, { "epoch": 0.3016167167231133, "grad_norm": 10.633510285123654, "learning_rate": 4.406239076342322e-07, "logits/chosen": -3.125, "logits/rejected": -3.03125, "logps/chosen": -536.0, "logps/rejected": -800.0, "loss": 0.2616, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.734375, "rewards/margins": 2.890625, "rewards/rejected": -6.625, "step": 2360 }, { "epoch": 0.3028947536583807, "grad_norm": 7.178556035609511, "learning_rate": 4.399003147764053e-07, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -508.0, "logps/rejected": -760.0, "loss": 0.2394, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.40625, "rewards/margins": 2.5, "rewards/rejected": -5.90625, "step": 2370 }, { "epoch": 0.30417279059364816, "grad_norm": 8.025880286415548, "learning_rate": 4.3917294135957185e-07, "logits/chosen": -2.984375, "logits/rejected": -2.953125, "logps/chosen": -502.0, "logps/rejected": -736.0, "loss": 0.2697, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.515625, "rewards/margins": 2.609375, "rewards/rejected": -6.125, "step": 2380 }, { "epoch": 0.30545082752891556, "grad_norm": 9.150701110291598, "learning_rate": 4.3844180186437205e-07, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -556.0, "logps/rejected": -796.0, "loss": 0.252, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -3.953125, "rewards/margins": 2.640625, "rewards/rejected": -6.59375, "step": 2390 }, { "epoch": 0.306728864464183, "grad_norm": 9.06253052565922, "learning_rate": 4.3770691084642153e-07, "logits/chosen": -3.140625, "logits/rejected": -3.0, "logps/chosen": -532.0, "logps/rejected": -840.0, "loss": 0.2478, "rewards/accuracies": 0.90625, "rewards/chosen": -3.828125, "rewards/margins": 3.25, "rewards/rejected": -7.0625, "step": 2400 }, { "epoch": 0.3080069013994504, "grad_norm": 8.361504252843975, "learning_rate": 4.3696828293602185e-07, "logits/chosen": -3.125, "logits/rejected": -3.15625, "logps/chosen": -544.0, "logps/rejected": -796.0, "loss": 0.2583, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.71875, "rewards/margins": 2.984375, "rewards/rejected": -6.6875, "step": 2410 }, { "epoch": 0.3092849383347179, "grad_norm": 9.674493266614986, "learning_rate": 4.3622593283786895e-07, "logits/chosen": -3.046875, "logits/rejected": -2.953125, "logps/chosen": -524.0, "logps/rejected": -816.0, "loss": 0.235, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.6875, "rewards/margins": 3.140625, "rewards/rejected": -6.84375, "step": 2420 }, { "epoch": 0.3105629752699853, "grad_norm": 8.99066458935145, "learning_rate": 4.354798753307606e-07, "logits/chosen": -2.796875, "logits/rejected": -2.6875, "logps/chosen": -540.0, "logps/rejected": -832.0, "loss": 0.2254, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -3.796875, "rewards/margins": 3.171875, "rewards/rejected": -6.96875, "step": 2430 }, { "epoch": 0.31184101220525273, "grad_norm": 10.245096178761992, "learning_rate": 4.3473012526730216e-07, "logits/chosen": -2.890625, "logits/rejected": -2.71875, "logps/chosen": -564.0, "logps/rejected": -828.0, "loss": 0.258, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -3.953125, "rewards/margins": 2.75, "rewards/rejected": -6.6875, "step": 2440 }, { "epoch": 0.31311904914052013, "grad_norm": 8.095300746315656, "learning_rate": 4.339766975736109e-07, "logits/chosen": -2.71875, "logits/rejected": -2.65625, "logps/chosen": -600.0, "logps/rejected": -904.0, "loss": 0.2535, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.5625, "rewards/margins": 2.78125, "rewards/rejected": -7.34375, "step": 2450 }, { "epoch": 0.3143970860757876, "grad_norm": 9.33334356795753, "learning_rate": 4.332196072490185e-07, "logits/chosen": -2.90625, "logits/rejected": -2.734375, "logps/chosen": -608.0, "logps/rejected": -884.0, "loss": 0.247, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.71875, "rewards/margins": 2.875, "rewards/rejected": -7.59375, "step": 2460 }, { "epoch": 0.315675123011055, "grad_norm": 9.68096887344856, "learning_rate": 4.324588693657733e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -604.0, "logps/rejected": -912.0, "loss": 0.2589, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.375, "rewards/margins": 3.25, "rewards/rejected": -7.625, "step": 2470 }, { "epoch": 0.31695315994632245, "grad_norm": 8.74974145611664, "learning_rate": 4.3169449906873925e-07, "logits/chosen": -3.015625, "logits/rejected": -2.921875, "logps/chosen": -592.0, "logps/rejected": -912.0, "loss": 0.2438, "rewards/accuracies": 0.875, "rewards/chosen": -4.21875, "rewards/margins": 3.25, "rewards/rejected": -7.46875, "step": 2480 }, { "epoch": 0.31823119688158985, "grad_norm": 6.971622374452891, "learning_rate": 4.309265115750949e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -612.0, "logps/rejected": -884.0, "loss": 0.2366, "rewards/accuracies": 0.875, "rewards/chosen": -4.5625, "rewards/margins": 2.984375, "rewards/rejected": -7.53125, "step": 2490 }, { "epoch": 0.3195092338168573, "grad_norm": 11.17024488529161, "learning_rate": 4.301549221740305e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -584.0, "logps/rejected": -868.0, "loss": 0.2874, "rewards/accuracies": 0.90625, "rewards/chosen": -4.1875, "rewards/margins": 2.90625, "rewards/rejected": -7.125, "step": 2500 }, { "epoch": 0.3207872707521247, "grad_norm": 10.477310622785978, "learning_rate": 4.293797462264436e-07, "logits/chosen": -2.9375, "logits/rejected": -2.84375, "logps/chosen": -548.0, "logps/rejected": -836.0, "loss": 0.2387, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.859375, "rewards/margins": 2.84375, "rewards/rejected": -6.71875, "step": 2510 }, { "epoch": 0.32206530768739217, "grad_norm": 10.024656692443616, "learning_rate": 4.286009991646329e-07, "logits/chosen": -3.046875, "logits/rejected": -2.984375, "logps/chosen": -580.0, "logps/rejected": -856.0, "loss": 0.253, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.09375, "rewards/margins": 3.140625, "rewards/rejected": -7.25, "step": 2520 }, { "epoch": 0.32334334462265957, "grad_norm": 8.009802554248187, "learning_rate": 4.2781869649199153e-07, "logits/chosen": -3.203125, "logits/rejected": -2.953125, "logps/chosen": -580.0, "logps/rejected": -864.0, "loss": 0.2492, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.25, "rewards/margins": 2.96875, "rewards/rejected": -7.1875, "step": 2530 }, { "epoch": 0.324621381557927, "grad_norm": 8.945237247295104, "learning_rate": 4.2703285378269815e-07, "logits/chosen": -3.015625, "logits/rejected": -2.75, "logps/chosen": -588.0, "logps/rejected": -852.0, "loss": 0.2462, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.25, "rewards/margins": 2.9375, "rewards/rejected": -7.1875, "step": 2540 }, { "epoch": 0.3258994184931944, "grad_norm": 10.551216216391607, "learning_rate": 4.262434866814067e-07, "logits/chosen": -3.140625, "logits/rejected": -2.96875, "logps/chosen": -632.0, "logps/rejected": -920.0, "loss": 0.2183, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.75, "rewards/margins": 3.0, "rewards/rejected": -7.75, "step": 2550 }, { "epoch": 0.3271774554284619, "grad_norm": 8.688942019994915, "learning_rate": 4.254506109029353e-07, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -628.0, "logps/rejected": -892.0, "loss": 0.2568, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.46875, "rewards/margins": 3.09375, "rewards/rejected": -7.5625, "step": 2560 }, { "epoch": 0.3284554923637293, "grad_norm": 6.564420476695321, "learning_rate": 4.2465424223195327e-07, "logits/chosen": -3.140625, "logits/rejected": -2.953125, "logps/chosen": -588.0, "logps/rejected": -868.0, "loss": 0.2343, "rewards/accuracies": 0.90625, "rewards/chosen": -4.34375, "rewards/margins": 2.953125, "rewards/rejected": -7.3125, "step": 2570 }, { "epoch": 0.32973352929899674, "grad_norm": 9.43421095366614, "learning_rate": 4.238543965226668e-07, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -624.0, "logps/rejected": -904.0, "loss": 0.2473, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.65625, "rewards/margins": 3.015625, "rewards/rejected": -7.6875, "step": 2580 }, { "epoch": 0.33101156623426414, "grad_norm": 13.150142319397695, "learning_rate": 4.230510896985035e-07, "logits/chosen": -2.953125, "logits/rejected": -2.921875, "logps/chosen": -688.0, "logps/rejected": -960.0, "loss": 0.2324, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.15625, "rewards/margins": 3.03125, "rewards/rejected": -8.1875, "step": 2590 }, { "epoch": 0.3322896031695316, "grad_norm": 9.050124476173716, "learning_rate": 4.2224433775179506e-07, "logits/chosen": -3.140625, "logits/rejected": -2.796875, "logps/chosen": -636.0, "logps/rejected": -904.0, "loss": 0.268, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.71875, "rewards/margins": 2.859375, "rewards/rejected": -7.59375, "step": 2600 }, { "epoch": 0.333567640104799, "grad_norm": 8.614894031818137, "learning_rate": 4.2143415674345937e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -608.0, "logps/rejected": -872.0, "loss": 0.2273, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.5, "rewards/margins": 2.703125, "rewards/rejected": -7.1875, "step": 2610 }, { "epoch": 0.33484567704006646, "grad_norm": 8.111143191701691, "learning_rate": 4.2062056280268033e-07, "logits/chosen": -3.0625, "logits/rejected": -2.765625, "logps/chosen": -652.0, "logps/rejected": -964.0, "loss": 0.2433, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.15625, "rewards/margins": 3.125, "rewards/rejected": -8.3125, "step": 2620 }, { "epoch": 0.33612371397533386, "grad_norm": 11.117676038308154, "learning_rate": 4.198035721265869e-07, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -560.0, "logps/rejected": -868.0, "loss": 0.2518, "rewards/accuracies": 0.90625, "rewards/chosen": -4.125, "rewards/margins": 3.09375, "rewards/rejected": -7.21875, "step": 2630 }, { "epoch": 0.3374017509106013, "grad_norm": 9.258130845121494, "learning_rate": 4.1898320097993085e-07, "logits/chosen": -3.0, "logits/rejected": -2.921875, "logps/chosen": -616.0, "logps/rejected": -880.0, "loss": 0.2846, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.46875, "rewards/margins": 2.859375, "rewards/rejected": -7.34375, "step": 2640 }, { "epoch": 0.3386797878458687, "grad_norm": 9.194807847235614, "learning_rate": 4.181594656947625e-07, "logits/chosen": -2.921875, "logits/rejected": -2.828125, "logps/chosen": -528.0, "logps/rejected": -796.0, "loss": 0.2202, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.75, "rewards/margins": 2.75, "rewards/rejected": -6.5, "step": 2650 }, { "epoch": 0.3399578247811362, "grad_norm": 9.12609259279053, "learning_rate": 4.173323826701062e-07, "logits/chosen": -3.03125, "logits/rejected": -2.890625, "logps/chosen": -596.0, "logps/rejected": -900.0, "loss": 0.235, "rewards/accuracies": 0.875, "rewards/chosen": -4.3125, "rewards/margins": 3.140625, "rewards/rejected": -7.46875, "step": 2660 }, { "epoch": 0.3412358617164036, "grad_norm": 8.558030417957948, "learning_rate": 4.165019683716332e-07, "logits/chosen": -2.859375, "logits/rejected": -2.75, "logps/chosen": -576.0, "logps/rejected": -836.0, "loss": 0.2471, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.15625, "rewards/margins": 2.734375, "rewards/rejected": -6.90625, "step": 2670 }, { "epoch": 0.34251389865167103, "grad_norm": 7.570195174697268, "learning_rate": 4.1566823933133444e-07, "logits/chosen": -3.046875, "logits/rejected": -2.921875, "logps/chosen": -536.0, "logps/rejected": -852.0, "loss": 0.2322, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.8125, "rewards/margins": 3.328125, "rewards/rejected": -7.125, "step": 2680 }, { "epoch": 0.34379193558693844, "grad_norm": 12.163376095462333, "learning_rate": 4.148312121471908e-07, "logits/chosen": -3.015625, "logits/rejected": -2.9375, "logps/chosen": -572.0, "logps/rejected": -924.0, "loss": 0.2238, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.125, "rewards/margins": 3.734375, "rewards/rejected": -7.875, "step": 2690 }, { "epoch": 0.3450699725222059, "grad_norm": 8.261895776179635, "learning_rate": 4.1399090348284336e-07, "logits/chosen": -3.125, "logits/rejected": -2.890625, "logps/chosen": -520.0, "logps/rejected": -812.0, "loss": 0.249, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.65625, "rewards/margins": 3.0625, "rewards/rejected": -6.71875, "step": 2700 }, { "epoch": 0.3463480094574733, "grad_norm": 7.401313646780075, "learning_rate": 4.1314733006726116e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -516.0, "logps/rejected": -784.0, "loss": 0.2496, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.59375, "rewards/margins": 2.953125, "rewards/rejected": -6.5625, "step": 2710 }, { "epoch": 0.34762604639274075, "grad_norm": 8.802407402306804, "learning_rate": 4.1230050869440825e-07, "logits/chosen": -2.953125, "logits/rejected": -2.796875, "logps/chosen": -588.0, "logps/rejected": -836.0, "loss": 0.2272, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.0625, "rewards/margins": 2.8125, "rewards/rejected": -6.875, "step": 2720 }, { "epoch": 0.34890408332800815, "grad_norm": 8.551492444117791, "learning_rate": 4.114504562229096e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -524.0, "logps/rejected": -820.0, "loss": 0.2462, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -3.734375, "rewards/margins": 3.15625, "rewards/rejected": -6.875, "step": 2730 }, { "epoch": 0.3501821202632756, "grad_norm": 13.237697656186098, "learning_rate": 4.105971895757151e-07, "logits/chosen": -2.890625, "logits/rejected": -2.71875, "logps/chosen": -528.0, "logps/rejected": -788.0, "loss": 0.262, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -3.90625, "rewards/margins": 2.703125, "rewards/rejected": -6.625, "step": 2740 }, { "epoch": 0.351460157198543, "grad_norm": 9.794093445613948, "learning_rate": 4.0974072573976295e-07, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -510.0, "logps/rejected": -812.0, "loss": 0.2488, "rewards/accuracies": 0.90625, "rewards/chosen": -3.609375, "rewards/margins": 3.046875, "rewards/rejected": -6.65625, "step": 2750 }, { "epoch": 0.35273819413381047, "grad_norm": 9.723773756216174, "learning_rate": 4.088810817656414e-07, "logits/chosen": -2.875, "logits/rejected": -2.75, "logps/chosen": -536.0, "logps/rejected": -808.0, "loss": 0.2479, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.671875, "rewards/margins": 2.984375, "rewards/rejected": -6.65625, "step": 2760 }, { "epoch": 0.35401623106907787, "grad_norm": 8.723909251690763, "learning_rate": 4.0801827476724923e-07, "logits/chosen": -2.890625, "logits/rejected": -2.796875, "logps/chosen": -548.0, "logps/rejected": -816.0, "loss": 0.2428, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.875, "rewards/margins": 2.859375, "rewards/rejected": -6.75, "step": 2770 }, { "epoch": 0.3552942680043453, "grad_norm": 8.462907156480545, "learning_rate": 4.071523219214551e-07, "logits/chosen": -2.953125, "logits/rejected": -2.734375, "logps/chosen": -580.0, "logps/rejected": -892.0, "loss": 0.2365, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.21875, "rewards/margins": 3.03125, "rewards/rejected": -7.25, "step": 2780 }, { "epoch": 0.3565723049396127, "grad_norm": 9.665834891128702, "learning_rate": 4.062832404677556e-07, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -560.0, "logps/rejected": -824.0, "loss": 0.2314, "rewards/accuracies": 0.90625, "rewards/chosen": -3.921875, "rewards/margins": 2.90625, "rewards/rejected": -6.8125, "step": 2790 }, { "epoch": 0.3578503418748802, "grad_norm": 8.989459813679218, "learning_rate": 4.054110477079321e-07, "logits/chosen": -2.953125, "logits/rejected": -2.8125, "logps/chosen": -548.0, "logps/rejected": -828.0, "loss": 0.2434, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.6875, "rewards/margins": 3.0625, "rewards/rejected": -6.75, "step": 2800 }, { "epoch": 0.3591283788101476, "grad_norm": 8.431501557282072, "learning_rate": 4.0453576100570606e-07, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -460.0, "logps/rejected": -728.0, "loss": 0.2438, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -3.125, "rewards/margins": 2.703125, "rewards/rejected": -5.84375, "step": 2810 }, { "epoch": 0.36040641574541504, "grad_norm": 15.875947653262912, "learning_rate": 4.036573977863938e-07, "logits/chosen": -2.9375, "logits/rejected": -2.765625, "logps/chosen": -576.0, "logps/rejected": -868.0, "loss": 0.2314, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -3.96875, "rewards/margins": 3.25, "rewards/rejected": -7.21875, "step": 2820 }, { "epoch": 0.36168445268068244, "grad_norm": 11.089723641989362, "learning_rate": 4.027759755365591e-07, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -588.0, "logps/rejected": -900.0, "loss": 0.2521, "rewards/accuracies": 0.90625, "rewards/chosen": -4.28125, "rewards/margins": 3.296875, "rewards/rejected": -7.5625, "step": 2830 }, { "epoch": 0.3629624896159499, "grad_norm": 7.979671614832609, "learning_rate": 4.018915118036653e-07, "logits/chosen": -2.859375, "logits/rejected": -2.75, "logps/chosen": -512.0, "logps/rejected": -808.0, "loss": 0.2409, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.46875, "rewards/margins": 3.15625, "rewards/rejected": -6.625, "step": 2840 }, { "epoch": 0.3642405265512173, "grad_norm": 7.47703355045046, "learning_rate": 4.0100402419572617e-07, "logits/chosen": -2.78125, "logits/rejected": -2.6875, "logps/chosen": -524.0, "logps/rejected": -840.0, "loss": 0.2303, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -3.46875, "rewards/margins": 3.1875, "rewards/rejected": -6.65625, "step": 2850 }, { "epoch": 0.36551856348648476, "grad_norm": 6.881027994969947, "learning_rate": 4.0011353038095497e-07, "logits/chosen": -2.96875, "logits/rejected": -2.859375, "logps/chosen": -560.0, "logps/rejected": -896.0, "loss": 0.2217, "rewards/accuracies": 0.9375, "rewards/chosen": -4.0, "rewards/margins": 3.375, "rewards/rejected": -7.375, "step": 2860 }, { "epoch": 0.36679660042175216, "grad_norm": 8.387929456445715, "learning_rate": 3.992200480874128e-07, "logits/chosen": -3.0, "logits/rejected": -2.71875, "logps/chosen": -584.0, "logps/rejected": -952.0, "loss": 0.2132, "rewards/accuracies": 0.90625, "rewards/chosen": -4.5625, "rewards/margins": 3.46875, "rewards/rejected": -8.0, "step": 2870 }, { "epoch": 0.3680746373570196, "grad_norm": 11.868719055234116, "learning_rate": 3.983235951026562e-07, "logits/chosen": -2.828125, "logits/rejected": -2.65625, "logps/chosen": -568.0, "logps/rejected": -852.0, "loss": 0.2397, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.21875, "rewards/margins": 2.921875, "rewards/rejected": -7.125, "step": 2880 }, { "epoch": 0.369352674292287, "grad_norm": 6.6165920564533565, "learning_rate": 3.974241892733821e-07, "logits/chosen": -2.90625, "logits/rejected": -2.71875, "logps/chosen": -556.0, "logps/rejected": -876.0, "loss": 0.2016, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.125, "rewards/margins": 3.171875, "rewards/rejected": -7.28125, "step": 2890 }, { "epoch": 0.3706307112275545, "grad_norm": 9.666254997669043, "learning_rate": 3.965218485050733e-07, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -560.0, "logps/rejected": -892.0, "loss": 0.2361, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.09375, "rewards/margins": 3.328125, "rewards/rejected": -7.40625, "step": 2900 }, { "epoch": 0.3719087481628219, "grad_norm": 8.197217719753612, "learning_rate": 3.9561659076164166e-07, "logits/chosen": -3.015625, "logits/rejected": -2.859375, "logps/chosen": -516.0, "logps/rejected": -868.0, "loss": 0.2378, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -3.421875, "rewards/margins": 3.78125, "rewards/rejected": -7.21875, "step": 2910 }, { "epoch": 0.37318678509808934, "grad_norm": 7.426775845977119, "learning_rate": 3.947084340650706e-07, "logits/chosen": -3.0, "logits/rejected": -2.703125, "logps/chosen": -560.0, "logps/rejected": -932.0, "loss": 0.2383, "rewards/accuracies": 0.9375, "rewards/chosen": -4.15625, "rewards/margins": 3.625, "rewards/rejected": -7.78125, "step": 2920 }, { "epoch": 0.37446482203335674, "grad_norm": 6.536302581696631, "learning_rate": 3.937973964950561e-07, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -540.0, "logps/rejected": -836.0, "loss": 0.2321, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.84375, "rewards/margins": 3.140625, "rewards/rejected": -6.96875, "step": 2930 }, { "epoch": 0.3757428589686242, "grad_norm": 12.669430664625045, "learning_rate": 3.928834961886472e-07, "logits/chosen": -3.046875, "logits/rejected": -3.03125, "logps/chosen": -502.0, "logps/rejected": -788.0, "loss": 0.2303, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.3125, "rewards/margins": 3.15625, "rewards/rejected": -6.46875, "step": 2940 }, { "epoch": 0.3770208959038916, "grad_norm": 10.03412970439382, "learning_rate": 3.919667513398843e-07, "logits/chosen": -3.171875, "logits/rejected": -2.9375, "logps/chosen": -500.0, "logps/rejected": -844.0, "loss": 0.2145, "rewards/accuracies": 0.90625, "rewards/chosen": -3.421875, "rewards/margins": 3.5, "rewards/rejected": -6.90625, "step": 2950 }, { "epoch": 0.37829893283915905, "grad_norm": 7.385896415249684, "learning_rate": 3.910471801994377e-07, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -592.0, "logps/rejected": -912.0, "loss": 0.2349, "rewards/accuracies": 0.9375, "rewards/chosen": -4.15625, "rewards/margins": 3.578125, "rewards/rejected": -7.75, "step": 2960 }, { "epoch": 0.37957696977442645, "grad_norm": 9.839499943864618, "learning_rate": 3.901248010742435e-07, "logits/chosen": -2.9375, "logits/rejected": -2.828125, "logps/chosen": -568.0, "logps/rejected": -880.0, "loss": 0.2274, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.28125, "rewards/margins": 3.140625, "rewards/rejected": -7.40625, "step": 2970 }, { "epoch": 0.3808550067096939, "grad_norm": 10.643804579819335, "learning_rate": 3.891996323271396e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -532.0, "logps/rejected": -844.0, "loss": 0.2548, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -3.734375, "rewards/margins": 3.25, "rewards/rejected": -7.0, "step": 2980 }, { "epoch": 0.3821330436449613, "grad_norm": 6.260351646142021, "learning_rate": 3.8827169237650023e-07, "logits/chosen": -3.09375, "logits/rejected": -2.875, "logps/chosen": -524.0, "logps/rejected": -812.0, "loss": 0.2137, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.6875, "rewards/margins": 3.0, "rewards/rejected": -6.6875, "step": 2990 }, { "epoch": 0.38341108058022877, "grad_norm": 9.26864693273647, "learning_rate": 3.87340999695869e-07, "logits/chosen": -3.046875, "logits/rejected": -2.890625, "logps/chosen": -536.0, "logps/rejected": -880.0, "loss": 0.2277, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -3.921875, "rewards/margins": 3.375, "rewards/rejected": -7.28125, "step": 3000 }, { "epoch": 0.38468911751549617, "grad_norm": 11.798346785124874, "learning_rate": 3.8640757281359104e-07, "logits/chosen": -2.921875, "logits/rejected": -2.796875, "logps/chosen": -524.0, "logps/rejected": -804.0, "loss": 0.2409, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -3.609375, "rewards/margins": 2.984375, "rewards/rejected": -6.59375, "step": 3010 }, { "epoch": 0.3859671544507636, "grad_norm": 9.13276020421161, "learning_rate": 3.8547143031244454e-07, "logits/chosen": -2.796875, "logits/rejected": -2.59375, "logps/chosen": -560.0, "logps/rejected": -880.0, "loss": 0.2363, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.0, "rewards/margins": 3.234375, "rewards/rejected": -7.25, "step": 3020 }, { "epoch": 0.38724519138603103, "grad_norm": 11.67031797954755, "learning_rate": 3.845325908292704e-07, "logits/chosen": -2.765625, "logits/rejected": -2.609375, "logps/chosen": -584.0, "logps/rejected": -920.0, "loss": 0.2292, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.15625, "rewards/margins": 3.40625, "rewards/rejected": -7.5625, "step": 3030 }, { "epoch": 0.3885232283212985, "grad_norm": 10.639147098006879, "learning_rate": 3.835910730546013e-07, "logits/chosen": -2.875, "logits/rejected": -2.828125, "logps/chosen": -568.0, "logps/rejected": -828.0, "loss": 0.245, "rewards/accuracies": 0.831250011920929, "rewards/chosen": -3.984375, "rewards/margins": 2.921875, "rewards/rejected": -6.875, "step": 3040 }, { "epoch": 0.3898012652565659, "grad_norm": 8.44223529175865, "learning_rate": 3.8264689573229e-07, "logits/chosen": -2.84375, "logits/rejected": -2.71875, "logps/chosen": -552.0, "logps/rejected": -880.0, "loss": 0.236, "rewards/accuracies": 0.875, "rewards/chosen": -3.875, "rewards/margins": 3.421875, "rewards/rejected": -7.28125, "step": 3050 }, { "epoch": 0.39107930219183334, "grad_norm": 8.488987566619226, "learning_rate": 3.8170007765913563e-07, "logits/chosen": -2.84375, "logits/rejected": -2.734375, "logps/chosen": -600.0, "logps/rejected": -888.0, "loss": 0.2073, "rewards/accuracies": 0.90625, "rewards/chosen": -4.21875, "rewards/margins": 3.28125, "rewards/rejected": -7.5, "step": 3060 }, { "epoch": 0.39235733912710075, "grad_norm": 8.533803292750349, "learning_rate": 3.8075063768450977e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -660.0, "logps/rejected": -996.0, "loss": 0.2315, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.96875, "rewards/margins": 3.421875, "rewards/rejected": -8.375, "step": 3070 }, { "epoch": 0.3936353760623682, "grad_norm": 13.47605711618702, "learning_rate": 3.79798594709981e-07, "logits/chosen": -3.046875, "logits/rejected": -2.9375, "logps/chosen": -592.0, "logps/rejected": -932.0, "loss": 0.2244, "rewards/accuracies": 0.90625, "rewards/chosen": -4.375, "rewards/margins": 3.53125, "rewards/rejected": -7.90625, "step": 3080 }, { "epoch": 0.3949134129976356, "grad_norm": 9.088959049871905, "learning_rate": 3.78843967688939e-07, "logits/chosen": -3.03125, "logits/rejected": -2.6875, "logps/chosen": -592.0, "logps/rejected": -892.0, "loss": 0.2113, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.34375, "rewards/margins": 3.140625, "rewards/rejected": -7.5, "step": 3090 }, { "epoch": 0.39619144993290306, "grad_norm": 6.990268151404954, "learning_rate": 3.7788677562621676e-07, "logits/chosen": -2.890625, "logits/rejected": -2.796875, "logps/chosen": -612.0, "logps/rejected": -892.0, "loss": 0.2411, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.5625, "rewards/margins": 2.96875, "rewards/rejected": -7.5, "step": 3100 }, { "epoch": 0.39746948686817046, "grad_norm": 10.21876475077512, "learning_rate": 3.769270375777126e-07, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -592.0, "logps/rejected": -920.0, "loss": 0.2438, "rewards/accuracies": 0.8374999761581421, "rewards/chosen": -4.4375, "rewards/margins": 3.1875, "rewards/rejected": -7.625, "step": 3110 }, { "epoch": 0.3987475238034379, "grad_norm": 8.94493406660772, "learning_rate": 3.7596477265001053e-07, "logits/chosen": -2.9375, "logits/rejected": -2.84375, "logps/chosen": -548.0, "logps/rejected": -856.0, "loss": 0.2207, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -3.984375, "rewards/margins": 3.171875, "rewards/rejected": -7.15625, "step": 3120 }, { "epoch": 0.4000255607387053, "grad_norm": 8.81703733566685, "learning_rate": 3.75e-07, "logits/chosen": -2.90625, "logits/rejected": -2.796875, "logps/chosen": -596.0, "logps/rejected": -912.0, "loss": 0.2356, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.3125, "rewards/margins": 3.359375, "rewards/rejected": -7.6875, "step": 3130 }, { "epoch": 0.4013035976739728, "grad_norm": 9.976485922500268, "learning_rate": 3.740327388344945e-07, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -568.0, "logps/rejected": -912.0, "loss": 0.2087, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.15625, "rewards/margins": 3.453125, "rewards/rejected": -7.59375, "step": 3140 }, { "epoch": 0.4025816346092402, "grad_norm": 6.8661146681276835, "learning_rate": 3.7306300840984927e-07, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -612.0, "logps/rejected": -1008.0, "loss": 0.2409, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.375, "rewards/margins": 4.03125, "rewards/rejected": -8.375, "step": 3150 }, { "epoch": 0.40385967154450764, "grad_norm": 10.705188172801805, "learning_rate": 3.720908280315777e-07, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -584.0, "logps/rejected": -892.0, "loss": 0.2195, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.25, "rewards/margins": 3.203125, "rewards/rejected": -7.46875, "step": 3160 }, { "epoch": 0.40513770847977504, "grad_norm": 7.260979308776354, "learning_rate": 3.711162170539673e-07, "logits/chosen": -2.828125, "logits/rejected": -2.671875, "logps/chosen": -624.0, "logps/rejected": -948.0, "loss": 0.2157, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.625, "rewards/margins": 3.453125, "rewards/rejected": -8.0625, "step": 3170 }, { "epoch": 0.4064157454150425, "grad_norm": 13.386154459966885, "learning_rate": 3.701391948796945e-07, "logits/chosen": -2.84375, "logits/rejected": -2.796875, "logps/chosen": -628.0, "logps/rejected": -888.0, "loss": 0.2287, "rewards/accuracies": 0.90625, "rewards/chosen": -4.59375, "rewards/margins": 2.90625, "rewards/rejected": -7.5, "step": 3180 }, { "epoch": 0.4076937823503099, "grad_norm": 10.71858268901349, "learning_rate": 3.6915978095943745e-07, "logits/chosen": -2.921875, "logits/rejected": -2.671875, "logps/chosen": -576.0, "logps/rejected": -884.0, "loss": 0.2193, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.3125, "rewards/margins": 3.15625, "rewards/rejected": -7.46875, "step": 3190 }, { "epoch": 0.40897181928557735, "grad_norm": 7.112292506270725, "learning_rate": 3.6817799479149007e-07, "logits/chosen": -2.921875, "logits/rejected": -2.6875, "logps/chosen": -600.0, "logps/rejected": -896.0, "loss": 0.2479, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.5, "rewards/margins": 3.09375, "rewards/rejected": -7.59375, "step": 3200 }, { "epoch": 0.41024985622084476, "grad_norm": 12.88518280843065, "learning_rate": 3.6719385592137306e-07, "logits/chosen": -2.84375, "logits/rejected": -2.59375, "logps/chosen": -576.0, "logps/rejected": -872.0, "loss": 0.2387, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.21875, "rewards/margins": 3.125, "rewards/rejected": -7.34375, "step": 3210 }, { "epoch": 0.4115278931561122, "grad_norm": 10.14123838967259, "learning_rate": 3.662073839414452e-07, "logits/chosen": -2.859375, "logits/rejected": -2.640625, "logps/chosen": -568.0, "logps/rejected": -876.0, "loss": 0.2255, "rewards/accuracies": 0.90625, "rewards/chosen": -4.0625, "rewards/margins": 3.203125, "rewards/rejected": -7.28125, "step": 3220 }, { "epoch": 0.4128059300913796, "grad_norm": 9.792291066078825, "learning_rate": 3.6521859849051274e-07, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -608.0, "logps/rejected": -904.0, "loss": 0.2431, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.53125, "rewards/margins": 3.109375, "rewards/rejected": -7.65625, "step": 3230 }, { "epoch": 0.41408396702664707, "grad_norm": 8.93660540981452, "learning_rate": 3.6422751925343906e-07, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -588.0, "logps/rejected": -900.0, "loss": 0.216, "rewards/accuracies": 0.875, "rewards/chosen": -4.21875, "rewards/margins": 3.265625, "rewards/rejected": -7.5, "step": 3240 }, { "epoch": 0.41536200396191447, "grad_norm": 9.207018470233994, "learning_rate": 3.6323416596075244e-07, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -612.0, "logps/rejected": -896.0, "loss": 0.2373, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.40625, "rewards/margins": 2.953125, "rewards/rejected": -7.34375, "step": 3250 }, { "epoch": 0.41664004089718193, "grad_norm": 8.798998468661477, "learning_rate": 3.622385583882535e-07, "logits/chosen": -2.875, "logits/rejected": -2.671875, "logps/chosen": -544.0, "logps/rejected": -840.0, "loss": 0.2528, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.0, "rewards/margins": 3.125, "rewards/rejected": -7.125, "step": 3260 }, { "epoch": 0.41791807783244933, "grad_norm": 7.39070684420073, "learning_rate": 3.6124071635662107e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -568.0, "logps/rejected": -856.0, "loss": 0.2159, "rewards/accuracies": 0.875, "rewards/chosen": -4.0625, "rewards/margins": 3.109375, "rewards/rejected": -7.1875, "step": 3270 }, { "epoch": 0.4191961147677168, "grad_norm": 10.572409208940522, "learning_rate": 3.6024065973101803e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -560.0, "logps/rejected": -860.0, "loss": 0.2188, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.0625, "rewards/margins": 3.25, "rewards/rejected": -7.3125, "step": 3280 }, { "epoch": 0.4204741517029842, "grad_norm": 11.489752500259677, "learning_rate": 3.5923840842069583e-07, "logits/chosen": -3.03125, "logits/rejected": -2.890625, "logps/chosen": -616.0, "logps/rejected": -948.0, "loss": 0.2227, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.4375, "rewards/rejected": -8.0, "step": 3290 }, { "epoch": 0.42175218863825165, "grad_norm": 8.808785722348407, "learning_rate": 3.5823398237859746e-07, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -620.0, "logps/rejected": -916.0, "loss": 0.2303, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.78125, "rewards/margins": 2.953125, "rewards/rejected": -7.71875, "step": 3300 }, { "epoch": 0.42303022557351905, "grad_norm": 8.69574601913517, "learning_rate": 3.572274016009613e-07, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -596.0, "logps/rejected": -964.0, "loss": 0.2233, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.34375, "rewards/margins": 3.484375, "rewards/rejected": -7.8125, "step": 3310 }, { "epoch": 0.4243082625087865, "grad_norm": 9.178583055571378, "learning_rate": 3.562186861269223e-07, "logits/chosen": -3.09375, "logits/rejected": -2.84375, "logps/chosen": -620.0, "logps/rejected": -936.0, "loss": 0.2118, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.5625, "rewards/margins": 3.390625, "rewards/rejected": -7.9375, "step": 3320 }, { "epoch": 0.4255862994440539, "grad_norm": 11.47875876274727, "learning_rate": 3.5520785603811296e-07, "logits/chosen": -2.890625, "logits/rejected": -2.65625, "logps/chosen": -592.0, "logps/rejected": -904.0, "loss": 0.2562, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.5, "rewards/margins": 3.078125, "rewards/rejected": -7.5625, "step": 3330 }, { "epoch": 0.42686433637932136, "grad_norm": 9.725339945230063, "learning_rate": 3.541949314582641e-07, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -600.0, "logps/rejected": -924.0, "loss": 0.2375, "rewards/accuracies": 0.875, "rewards/chosen": -4.4375, "rewards/margins": 3.25, "rewards/rejected": -7.6875, "step": 3340 }, { "epoch": 0.42814237331458876, "grad_norm": 8.970902307435473, "learning_rate": 3.5317993255280383e-07, "logits/chosen": -2.84375, "logits/rejected": -2.703125, "logps/chosen": -668.0, "logps/rejected": -984.0, "loss": 0.2017, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.9375, "rewards/margins": 3.5, "rewards/rejected": -8.4375, "step": 3350 }, { "epoch": 0.4294204102498562, "grad_norm": 8.382183001221023, "learning_rate": 3.5216287952845614e-07, "logits/chosen": -3.015625, "logits/rejected": -2.625, "logps/chosen": -680.0, "logps/rejected": -1040.0, "loss": 0.1958, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.21875, "rewards/margins": 3.625, "rewards/rejected": -8.8125, "step": 3360 }, { "epoch": 0.4306984471851236, "grad_norm": 9.189979249063406, "learning_rate": 3.511437926328387e-07, "logits/chosen": -3.125, "logits/rejected": -2.765625, "logps/chosen": -616.0, "logps/rejected": -984.0, "loss": 0.246, "rewards/accuracies": 0.90625, "rewards/chosen": -4.65625, "rewards/margins": 3.546875, "rewards/rejected": -8.1875, "step": 3370 }, { "epoch": 0.4319764841203911, "grad_norm": 8.228711642134167, "learning_rate": 3.501226921540598e-07, "logits/chosen": -3.171875, "logits/rejected": -2.84375, "logps/chosen": -580.0, "logps/rejected": -932.0, "loss": 0.2335, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.3125, "rewards/margins": 3.515625, "rewards/rejected": -7.84375, "step": 3380 }, { "epoch": 0.4332545210556585, "grad_norm": 9.21386830259457, "learning_rate": 3.4909959842031436e-07, "logits/chosen": -2.859375, "logits/rejected": -2.890625, "logps/chosen": -616.0, "logps/rejected": -884.0, "loss": 0.2189, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -4.65625, "rewards/margins": 2.84375, "rewards/rejected": -7.5, "step": 3390 }, { "epoch": 0.43453255799092594, "grad_norm": 6.248955823376214, "learning_rate": 3.480745317994793e-07, "logits/chosen": -3.015625, "logits/rejected": -2.890625, "logps/chosen": -624.0, "logps/rejected": -940.0, "loss": 0.2178, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.375, "rewards/rejected": -7.9375, "step": 3400 }, { "epoch": 0.43581059492619334, "grad_norm": 9.020550855639287, "learning_rate": 3.4704751269870774e-07, "logits/chosen": -3.078125, "logits/rejected": -2.90625, "logps/chosen": -612.0, "logps/rejected": -924.0, "loss": 0.1958, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.625, "rewards/margins": 3.375, "rewards/rejected": -8.0, "step": 3410 }, { "epoch": 0.4370886318614608, "grad_norm": 10.664082807354959, "learning_rate": 3.460185615640234e-07, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -596.0, "logps/rejected": -956.0, "loss": 0.2194, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.5, "rewards/margins": 3.671875, "rewards/rejected": -8.1875, "step": 3420 }, { "epoch": 0.4383666687967282, "grad_norm": 10.56454935423742, "learning_rate": 3.4498769887991316e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -600.0, "logps/rejected": -952.0, "loss": 0.2377, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.46875, "rewards/margins": 3.484375, "rewards/rejected": -7.9375, "step": 3430 }, { "epoch": 0.43964470573199566, "grad_norm": 8.491123471406155, "learning_rate": 3.4395494516891885e-07, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -632.0, "logps/rejected": -904.0, "loss": 0.2178, "rewards/accuracies": 0.875, "rewards/chosen": -4.875, "rewards/margins": 2.875, "rewards/rejected": -7.75, "step": 3440 }, { "epoch": 0.44092274266726306, "grad_norm": 9.201408062848294, "learning_rate": 3.4292032099122956e-07, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -592.0, "logps/rejected": -928.0, "loss": 0.2088, "rewards/accuracies": 0.90625, "rewards/chosen": -4.53125, "rewards/margins": 3.390625, "rewards/rejected": -7.9375, "step": 3450 }, { "epoch": 0.4422007796025305, "grad_norm": 7.860836704868255, "learning_rate": 3.4188384694427166e-07, "logits/chosen": -3.0, "logits/rejected": -2.921875, "logps/chosen": -628.0, "logps/rejected": -908.0, "loss": 0.2384, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.84375, "rewards/margins": 2.921875, "rewards/rejected": -7.78125, "step": 3460 }, { "epoch": 0.4434788165377979, "grad_norm": 13.747547646724954, "learning_rate": 3.4084554366229884e-07, "logits/chosen": -3.03125, "logits/rejected": -2.890625, "logps/chosen": -652.0, "logps/rejected": -900.0, "loss": 0.2205, "rewards/accuracies": 0.90625, "rewards/chosen": -4.65625, "rewards/margins": 3.078125, "rewards/rejected": -7.71875, "step": 3470 }, { "epoch": 0.44475685347306537, "grad_norm": 8.848263859657068, "learning_rate": 3.398054318159816e-07, "logits/chosen": -3.046875, "logits/rejected": -2.890625, "logps/chosen": -668.0, "logps/rejected": -992.0, "loss": 0.2116, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.96875, "rewards/margins": 3.53125, "rewards/rejected": -8.5, "step": 3480 }, { "epoch": 0.4460348904083328, "grad_norm": 10.597332854509538, "learning_rate": 3.387635321119955e-07, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -620.0, "logps/rejected": -948.0, "loss": 0.228, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.71875, "rewards/margins": 3.359375, "rewards/rejected": -8.0625, "step": 3490 }, { "epoch": 0.44731292734360023, "grad_norm": 9.674325942945405, "learning_rate": 3.3771986529260914e-07, "logits/chosen": -2.9375, "logits/rejected": -2.671875, "logps/chosen": -640.0, "logps/rejected": -996.0, "loss": 0.2123, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.03125, "rewards/margins": 3.4375, "rewards/rejected": -8.5, "step": 3500 }, { "epoch": 0.44859096427886763, "grad_norm": 12.193772222585018, "learning_rate": 3.3667445213527075e-07, "logits/chosen": -3.125, "logits/rejected": -2.765625, "logps/chosen": -656.0, "logps/rejected": -992.0, "loss": 0.2273, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.0625, "rewards/margins": 3.421875, "rewards/rejected": -8.5, "step": 3510 }, { "epoch": 0.4498690012141351, "grad_norm": 11.935477705606276, "learning_rate": 3.356273134521951e-07, "logits/chosen": -3.1875, "logits/rejected": -3.03125, "logps/chosen": -648.0, "logps/rejected": -972.0, "loss": 0.2136, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.5625, "rewards/rejected": -8.3125, "step": 3520 }, { "epoch": 0.4511470381494025, "grad_norm": 10.16354356392599, "learning_rate": 3.3457847008994895e-07, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -624.0, "logps/rejected": -960.0, "loss": 0.2291, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.75, "rewards/margins": 3.484375, "rewards/rejected": -8.25, "step": 3530 }, { "epoch": 0.45242507508466995, "grad_norm": 9.401934005328973, "learning_rate": 3.3352794292903584e-07, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -644.0, "logps/rejected": -944.0, "loss": 0.2383, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.875, "rewards/margins": 3.0625, "rewards/rejected": -7.9375, "step": 3540 }, { "epoch": 0.45370311201993735, "grad_norm": 12.304466333106431, "learning_rate": 3.324757528834809e-07, "logits/chosen": -2.9375, "logits/rejected": -2.890625, "logps/chosen": -676.0, "logps/rejected": -968.0, "loss": 0.2339, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.03125, "rewards/margins": 3.296875, "rewards/rejected": -8.3125, "step": 3550 }, { "epoch": 0.4549811489552048, "grad_norm": 10.335666794847032, "learning_rate": 3.314219209004139e-07, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -628.0, "logps/rejected": -952.0, "loss": 0.2134, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.84375, "rewards/margins": 3.421875, "rewards/rejected": -8.25, "step": 3560 }, { "epoch": 0.4562591858904722, "grad_norm": 8.135726139180283, "learning_rate": 3.303664679596526e-07, "logits/chosen": -3.09375, "logits/rejected": -2.96875, "logps/chosen": -632.0, "logps/rejected": -968.0, "loss": 0.2045, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.453125, "rewards/rejected": -8.25, "step": 3570 }, { "epoch": 0.45753722282573966, "grad_norm": 10.36951976978052, "learning_rate": 3.293094150732849e-07, "logits/chosen": -3.078125, "logits/rejected": -2.890625, "logps/chosen": -652.0, "logps/rejected": -1020.0, "loss": 0.1959, "rewards/accuracies": 0.90625, "rewards/chosen": -5.0, "rewards/margins": 3.609375, "rewards/rejected": -8.625, "step": 3580 }, { "epoch": 0.45881525976100707, "grad_norm": 12.805385346829228, "learning_rate": 3.2825078328525076e-07, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -660.0, "logps/rejected": -964.0, "loss": 0.2227, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.9375, "rewards/margins": 3.328125, "rewards/rejected": -8.25, "step": 3590 }, { "epoch": 0.4600932966962745, "grad_norm": 8.737872062348318, "learning_rate": 3.271905936709231e-07, "logits/chosen": -3.109375, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -1020.0, "loss": 0.2004, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.34375, "rewards/margins": 3.34375, "rewards/rejected": -8.6875, "step": 3600 }, { "epoch": 0.4613713336315419, "grad_norm": 5.90450642703934, "learning_rate": 3.261288673366881e-07, "logits/chosen": -3.171875, "logits/rejected": -2.890625, "logps/chosen": -632.0, "logps/rejected": -936.0, "loss": 0.2222, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.328125, "rewards/rejected": -7.96875, "step": 3610 }, { "epoch": 0.4626493705668094, "grad_norm": 10.917396256677707, "learning_rate": 3.250656254195252e-07, "logits/chosen": -3.125, "logits/rejected": -2.921875, "logps/chosen": -600.0, "logps/rejected": -912.0, "loss": 0.2324, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.4375, "rewards/margins": 3.21875, "rewards/rejected": -7.65625, "step": 3620 }, { "epoch": 0.4639274075020768, "grad_norm": 8.521570669595315, "learning_rate": 3.240008890865864e-07, "logits/chosen": -3.203125, "logits/rejected": -3.078125, "logps/chosen": -632.0, "logps/rejected": -952.0, "loss": 0.2087, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.375, "rewards/rejected": -8.1875, "step": 3630 }, { "epoch": 0.46520544443734424, "grad_norm": 8.202418931720207, "learning_rate": 3.2293467953477454e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -968.0, "loss": 0.2127, "rewards/accuracies": 0.84375, "rewards/chosen": -5.1875, "rewards/margins": 2.828125, "rewards/rejected": -8.0, "step": 3640 }, { "epoch": 0.46648348137261164, "grad_norm": 6.5224761862396585, "learning_rate": 3.218670179903216e-07, "logits/chosen": -3.0, "logits/rejected": -2.828125, "logps/chosen": -656.0, "logps/rejected": -984.0, "loss": 0.2125, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.125, "rewards/margins": 3.578125, "rewards/rejected": -8.6875, "step": 3650 }, { "epoch": 0.4677615183078791, "grad_norm": 11.98921345415937, "learning_rate": 3.207979257083658e-07, "logits/chosen": -3.0625, "logits/rejected": -2.859375, "logps/chosen": -624.0, "logps/rejected": -1008.0, "loss": 0.2271, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.5625, "rewards/margins": 3.8125, "rewards/rejected": -8.375, "step": 3660 }, { "epoch": 0.4690395552431465, "grad_norm": 9.018316199665886, "learning_rate": 3.19727423972529e-07, "logits/chosen": -3.15625, "logits/rejected": -2.890625, "logps/chosen": -632.0, "logps/rejected": -988.0, "loss": 0.2005, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.546875, "rewards/rejected": -8.3125, "step": 3670 }, { "epoch": 0.47031759217841396, "grad_norm": 7.973706582819973, "learning_rate": 3.186555340944923e-07, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -588.0, "logps/rejected": -924.0, "loss": 0.2111, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.53125, "rewards/margins": 3.5625, "rewards/rejected": -8.0625, "step": 3680 }, { "epoch": 0.47159562911368136, "grad_norm": 8.551400923945243, "learning_rate": 3.175822774135724e-07, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -640.0, "logps/rejected": -956.0, "loss": 0.2167, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.65625, "rewards/margins": 3.375, "rewards/rejected": -8.0625, "step": 3690 }, { "epoch": 0.4728736660489488, "grad_norm": 9.50674868300808, "learning_rate": 3.165076752962962e-07, "logits/chosen": -2.984375, "logits/rejected": -2.828125, "logps/chosen": -592.0, "logps/rejected": -944.0, "loss": 0.1914, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.3125, "rewards/margins": 3.734375, "rewards/rejected": -8.0625, "step": 3700 }, { "epoch": 0.4741517029842162, "grad_norm": 8.941779392618068, "learning_rate": 3.15431749135976e-07, "logits/chosen": -3.0625, "logits/rejected": -2.90625, "logps/chosen": -576.0, "logps/rejected": -896.0, "loss": 0.2065, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.28125, "rewards/margins": 3.28125, "rewards/rejected": -7.5625, "step": 3710 }, { "epoch": 0.4754297399194837, "grad_norm": 13.024991648002365, "learning_rate": 3.1435452035228306e-07, "logits/chosen": -2.953125, "logits/rejected": -2.890625, "logps/chosen": -604.0, "logps/rejected": -888.0, "loss": 0.2345, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.46875, "rewards/margins": 3.125, "rewards/rejected": -7.59375, "step": 3720 }, { "epoch": 0.4767077768547511, "grad_norm": 9.638702132762546, "learning_rate": 3.132760103908216e-07, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -652.0, "logps/rejected": -956.0, "loss": 0.221, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.4375, "rewards/rejected": -8.25, "step": 3730 }, { "epoch": 0.47798581379001853, "grad_norm": 10.2285096933021, "learning_rate": 3.1219624072270166e-07, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -976.0, "loss": 0.2144, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.6875, "rewards/margins": 3.5, "rewards/rejected": -8.1875, "step": 3740 }, { "epoch": 0.47926385072528593, "grad_norm": 9.481196208270266, "learning_rate": 3.111152328441115e-07, "logits/chosen": -2.875, "logits/rejected": -2.59375, "logps/chosen": -616.0, "logps/rejected": -980.0, "loss": 0.2314, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.625, "rewards/rejected": -8.25, "step": 3750 }, { "epoch": 0.4805418876605534, "grad_norm": 11.176541898064906, "learning_rate": 3.100330082758901e-07, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -628.0, "logps/rejected": -940.0, "loss": 0.193, "rewards/accuracies": 0.90625, "rewards/chosen": -4.71875, "rewards/margins": 3.125, "rewards/rejected": -7.8125, "step": 3760 }, { "epoch": 0.4818199245958208, "grad_norm": 10.520661709895942, "learning_rate": 3.089495885630983e-07, "logits/chosen": -2.984375, "logits/rejected": -2.671875, "logps/chosen": -620.0, "logps/rejected": -928.0, "loss": 0.2282, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.5, "rewards/margins": 3.25, "rewards/rejected": -7.75, "step": 3770 }, { "epoch": 0.48309796153108825, "grad_norm": 12.274708641803729, "learning_rate": 3.0786499527459005e-07, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -672.0, "logps/rejected": -972.0, "loss": 0.2214, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.0625, "rewards/margins": 3.171875, "rewards/rejected": -8.25, "step": 3780 }, { "epoch": 0.48437599846635565, "grad_norm": 12.872720663919305, "learning_rate": 3.0677925000258285e-07, "logits/chosen": -3.015625, "logits/rejected": -2.703125, "logps/chosen": -672.0, "logps/rejected": -1020.0, "loss": 0.2301, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.1875, "rewards/margins": 3.4375, "rewards/rejected": -8.625, "step": 3790 }, { "epoch": 0.4856540354016231, "grad_norm": 9.708450732405078, "learning_rate": 3.056923743622283e-07, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -564.0, "logps/rejected": -904.0, "loss": 0.2137, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.21875, "rewards/margins": 3.4375, "rewards/rejected": -7.65625, "step": 3800 }, { "epoch": 0.4869320723368905, "grad_norm": 8.19563346716996, "learning_rate": 3.0460438999118144e-07, "logits/chosen": -3.140625, "logits/rejected": -2.921875, "logps/chosen": -616.0, "logps/rejected": -928.0, "loss": 0.215, "rewards/accuracies": 0.90625, "rewards/chosen": -4.5625, "rewards/margins": 3.421875, "rewards/rejected": -7.96875, "step": 3810 }, { "epoch": 0.48821010927215797, "grad_norm": 8.106052628533215, "learning_rate": 3.035153185491698e-07, "logits/chosen": -3.015625, "logits/rejected": -2.984375, "logps/chosen": -632.0, "logps/rejected": -976.0, "loss": 0.2089, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.65625, "rewards/margins": 3.703125, "rewards/rejected": -8.375, "step": 3820 }, { "epoch": 0.48948814620742537, "grad_norm": 13.685997047539583, "learning_rate": 3.024251817175628e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -636.0, "logps/rejected": -960.0, "loss": 0.2047, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.6875, "rewards/margins": 3.515625, "rewards/rejected": -8.1875, "step": 3830 }, { "epoch": 0.4907661831426928, "grad_norm": 8.488651222369613, "learning_rate": 3.0133400119893947e-07, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -608.0, "logps/rejected": -976.0, "loss": 0.2136, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.5625, "rewards/margins": 3.984375, "rewards/rejected": -8.5, "step": 3840 }, { "epoch": 0.4920442200779602, "grad_norm": 12.500560067856066, "learning_rate": 3.0024179871665706e-07, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -684.0, "logps/rejected": -996.0, "loss": 0.2082, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.3125, "rewards/margins": 3.34375, "rewards/rejected": -8.625, "step": 3850 }, { "epoch": 0.4933222570132277, "grad_norm": 9.56821539066649, "learning_rate": 2.9914859601441774e-07, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -644.0, "logps/rejected": -988.0, "loss": 0.2254, "rewards/accuracies": 0.90625, "rewards/chosen": -4.8125, "rewards/margins": 3.703125, "rewards/rejected": -8.5, "step": 3860 }, { "epoch": 0.4946002939484951, "grad_norm": 11.642104475986267, "learning_rate": 2.980544148558365e-07, "logits/chosen": -3.078125, "logits/rejected": -2.78125, "logps/chosen": -644.0, "logps/rejected": -1020.0, "loss": 0.2271, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.90625, "rewards/margins": 3.859375, "rewards/rejected": -8.75, "step": 3870 }, { "epoch": 0.49587833088376254, "grad_norm": 11.975382621959218, "learning_rate": 2.9695927702400767e-07, "logits/chosen": -2.90625, "logits/rejected": -2.65625, "logps/chosen": -648.0, "logps/rejected": -1016.0, "loss": 0.216, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.875, "rewards/margins": 3.765625, "rewards/rejected": -8.625, "step": 3880 }, { "epoch": 0.49715636781902994, "grad_norm": 8.674175640451184, "learning_rate": 2.958632043210706e-07, "logits/chosen": -2.734375, "logits/rejected": -2.5, "logps/chosen": -628.0, "logps/rejected": -956.0, "loss": 0.2073, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.40625, "rewards/rejected": -8.0, "step": 3890 }, { "epoch": 0.4984344047542974, "grad_norm": 13.531335502879033, "learning_rate": 2.9476621856777686e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -652.0, "logps/rejected": -980.0, "loss": 0.2201, "rewards/accuracies": 0.90625, "rewards/chosen": -4.96875, "rewards/margins": 3.421875, "rewards/rejected": -8.375, "step": 3900 }, { "epoch": 0.4997124416895648, "grad_norm": 9.753557812476503, "learning_rate": 2.9366834160305466e-07, "logits/chosen": -2.8125, "logits/rejected": -2.59375, "logps/chosen": -664.0, "logps/rejected": -1004.0, "loss": 0.1996, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.90625, "rewards/margins": 3.625, "rewards/rejected": -8.5625, "step": 3910 }, { "epoch": 0.5009904786248323, "grad_norm": 8.623771874932421, "learning_rate": 2.925695952835746e-07, "logits/chosen": -2.8125, "logits/rejected": -2.71875, "logps/chosen": -668.0, "logps/rejected": -1032.0, "loss": 0.1957, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.78125, "rewards/rejected": -8.625, "step": 3920 }, { "epoch": 0.5022685155600997, "grad_norm": 8.619272047417937, "learning_rate": 2.9147000148331483e-07, "logits/chosen": -2.78125, "logits/rejected": -2.59375, "logps/chosen": -672.0, "logps/rejected": -988.0, "loss": 0.1965, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.09375, "rewards/margins": 3.265625, "rewards/rejected": -8.3125, "step": 3930 }, { "epoch": 0.5035465524953671, "grad_norm": 9.340092455147365, "learning_rate": 2.903695820931249e-07, "logits/chosen": -2.859375, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1020.0, "loss": 0.2062, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.40625, "rewards/margins": 3.34375, "rewards/rejected": -8.75, "step": 3940 }, { "epoch": 0.5048245894306346, "grad_norm": 10.45509856513219, "learning_rate": 2.8926835902029043e-07, "logits/chosen": -2.953125, "logits/rejected": -2.734375, "logps/chosen": -624.0, "logps/rejected": -964.0, "loss": 0.2042, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.71875, "rewards/margins": 3.5, "rewards/rejected": -8.25, "step": 3950 }, { "epoch": 0.506102626365902, "grad_norm": 10.115987074829942, "learning_rate": 2.8816635418809693e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -616.0, "logps/rejected": -996.0, "loss": 0.2186, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.6875, "rewards/margins": 3.828125, "rewards/rejected": -8.5, "step": 3960 }, { "epoch": 0.5073806633011694, "grad_norm": 9.859559908143657, "learning_rate": 2.870635895353931e-07, "logits/chosen": -2.859375, "logits/rejected": -2.78125, "logps/chosen": -600.0, "logps/rejected": -900.0, "loss": 0.1997, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.46875, "rewards/margins": 3.1875, "rewards/rejected": -7.65625, "step": 3970 }, { "epoch": 0.5086587002364368, "grad_norm": 10.87319879727593, "learning_rate": 2.8596008701615456e-07, "logits/chosen": -2.90625, "logits/rejected": -2.71875, "logps/chosen": -608.0, "logps/rejected": -952.0, "loss": 0.2066, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.625, "rewards/margins": 3.578125, "rewards/rejected": -8.1875, "step": 3980 }, { "epoch": 0.5099367371717043, "grad_norm": 7.8658320285041405, "learning_rate": 2.8485586859904596e-07, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -596.0, "logps/rejected": -952.0, "loss": 0.2377, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.53125, "rewards/margins": 3.546875, "rewards/rejected": -8.0625, "step": 3990 }, { "epoch": 0.5112147741069717, "grad_norm": 8.505547152770252, "learning_rate": 2.8375095626698464e-07, "logits/chosen": -2.875, "logits/rejected": -2.75, "logps/chosen": -592.0, "logps/rejected": -896.0, "loss": 0.2167, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.40625, "rewards/margins": 3.296875, "rewards/rejected": -7.71875, "step": 4000 }, { "epoch": 0.5124928110422391, "grad_norm": 11.91791149654105, "learning_rate": 2.826453720167021e-07, "logits/chosen": -2.890625, "logits/rejected": -2.71875, "logps/chosen": -640.0, "logps/rejected": -960.0, "loss": 0.2182, "rewards/accuracies": 0.90625, "rewards/chosen": -4.78125, "rewards/margins": 3.28125, "rewards/rejected": -8.0625, "step": 4010 }, { "epoch": 0.5137708479775065, "grad_norm": 9.968567086380052, "learning_rate": 2.815391378583069e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -660.0, "logps/rejected": -976.0, "loss": 0.2166, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.0625, "rewards/margins": 3.359375, "rewards/rejected": -8.4375, "step": 4020 }, { "epoch": 0.515048884912774, "grad_norm": 9.35733243554509, "learning_rate": 2.8043227581484556e-07, "logits/chosen": -3.015625, "logits/rejected": -2.90625, "logps/chosen": -652.0, "logps/rejected": -984.0, "loss": 0.2074, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.609375, "rewards/rejected": -8.375, "step": 4030 }, { "epoch": 0.5163269218480414, "grad_norm": 12.215635212530254, "learning_rate": 2.79324807921865e-07, "logits/chosen": -3.046875, "logits/rejected": -2.953125, "logps/chosen": -632.0, "logps/rejected": -944.0, "loss": 0.227, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.71875, "rewards/margins": 3.25, "rewards/rejected": -7.96875, "step": 4040 }, { "epoch": 0.5176049587833088, "grad_norm": 13.845726328872498, "learning_rate": 2.7821675622697317e-07, "logits/chosen": -3.03125, "logits/rejected": -2.953125, "logps/chosen": -684.0, "logps/rejected": -1056.0, "loss": 0.2227, "rewards/accuracies": 0.90625, "rewards/chosen": -5.125, "rewards/margins": 3.796875, "rewards/rejected": -8.875, "step": 4050 }, { "epoch": 0.5188829957185762, "grad_norm": 11.50962600886563, "learning_rate": 2.7710814278940057e-07, "logits/chosen": -3.078125, "logits/rejected": -2.875, "logps/chosen": -676.0, "logps/rejected": -984.0, "loss": 0.2251, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -5.0625, "rewards/margins": 3.296875, "rewards/rejected": -8.375, "step": 4060 }, { "epoch": 0.5201610326538437, "grad_norm": 9.136976034548045, "learning_rate": 2.759989896795611e-07, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -672.0, "logps/rejected": -996.0, "loss": 0.2297, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.15625, "rewards/margins": 3.34375, "rewards/rejected": -8.5, "step": 4070 }, { "epoch": 0.5214390695891111, "grad_norm": 8.603249208409274, "learning_rate": 2.748893189786122e-07, "logits/chosen": -3.046875, "logits/rejected": -2.71875, "logps/chosen": -640.0, "logps/rejected": -956.0, "loss": 0.1937, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.296875, "rewards/rejected": -8.125, "step": 4080 }, { "epoch": 0.5227171065243785, "grad_norm": 8.84947914185045, "learning_rate": 2.7377915277801586e-07, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -636.0, "logps/rejected": -980.0, "loss": 0.1985, "rewards/accuracies": 0.9375, "rewards/chosen": -4.5625, "rewards/margins": 3.78125, "rewards/rejected": -8.375, "step": 4090 }, { "epoch": 0.5239951434596459, "grad_norm": 10.135619925874042, "learning_rate": 2.726685131790983e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -648.0, "logps/rejected": -996.0, "loss": 0.2223, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.796875, "rewards/rejected": -8.625, "step": 4100 }, { "epoch": 0.5252731803949134, "grad_norm": 8.893574056072689, "learning_rate": 2.715574222926105e-07, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -640.0, "logps/rejected": -1004.0, "loss": 0.2135, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.65625, "rewards/rejected": -8.375, "step": 4110 }, { "epoch": 0.5265512173301808, "grad_norm": 5.695805913492098, "learning_rate": 2.704459022382874e-07, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -632.0, "logps/rejected": -976.0, "loss": 0.1975, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.875, "rewards/margins": 3.515625, "rewards/rejected": -8.375, "step": 4120 }, { "epoch": 0.5278292542654482, "grad_norm": 6.965426261729773, "learning_rate": 2.6933397514440796e-07, "logits/chosen": -3.015625, "logits/rejected": -2.921875, "logps/chosen": -644.0, "logps/rejected": -1024.0, "loss": 0.2055, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.625, "rewards/margins": 4.03125, "rewards/rejected": -8.6875, "step": 4130 }, { "epoch": 0.5291072912007156, "grad_norm": 8.527321271713554, "learning_rate": 2.6822166314735483e-07, "logits/chosen": -3.15625, "logits/rejected": -2.796875, "logps/chosen": -624.0, "logps/rejected": -984.0, "loss": 0.2099, "rewards/accuracies": 0.90625, "rewards/chosen": -4.625, "rewards/margins": 3.625, "rewards/rejected": -8.25, "step": 4140 }, { "epoch": 0.5303853281359832, "grad_norm": 11.360590075652906, "learning_rate": 2.67108988391173e-07, "logits/chosen": -3.0625, "logits/rejected": -2.953125, "logps/chosen": -608.0, "logps/rejected": -936.0, "loss": 0.2048, "rewards/accuracies": 0.90625, "rewards/chosen": -4.65625, "rewards/margins": 3.3125, "rewards/rejected": -7.96875, "step": 4150 }, { "epoch": 0.5316633650712506, "grad_norm": 8.516403048337656, "learning_rate": 2.659959730271295e-07, "logits/chosen": -3.046875, "logits/rejected": -2.9375, "logps/chosen": -600.0, "logps/rejected": -952.0, "loss": 0.2067, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.65625, "rewards/rejected": -8.25, "step": 4160 }, { "epoch": 0.532941402006518, "grad_norm": 8.679204620956797, "learning_rate": 2.64882639213272e-07, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -624.0, "logps/rejected": -936.0, "loss": 0.2018, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.78125, "rewards/margins": 3.359375, "rewards/rejected": -8.125, "step": 4170 }, { "epoch": 0.5342194389417854, "grad_norm": 8.214012990646285, "learning_rate": 2.6376900911398823e-07, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -984.0, "loss": 0.208, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.78125, "rewards/margins": 3.59375, "rewards/rejected": -8.375, "step": 4180 }, { "epoch": 0.5354974758770529, "grad_norm": 6.735540115827085, "learning_rate": 2.626551048995643e-07, "logits/chosen": -2.921875, "logits/rejected": -2.84375, "logps/chosen": -628.0, "logps/rejected": -928.0, "loss": 0.2165, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.8125, "rewards/margins": 3.203125, "rewards/rejected": -8.0, "step": 4190 }, { "epoch": 0.5367755128123203, "grad_norm": 13.73261695569995, "learning_rate": 2.6154094874574326e-07, "logits/chosen": -3.09375, "logits/rejected": -2.953125, "logps/chosen": -620.0, "logps/rejected": -984.0, "loss": 0.2123, "rewards/accuracies": 0.875, "rewards/chosen": -4.6875, "rewards/margins": 3.671875, "rewards/rejected": -8.375, "step": 4200 }, { "epoch": 0.5380535497475877, "grad_norm": 8.6629838322733, "learning_rate": 2.60426562833284e-07, "logits/chosen": -3.015625, "logits/rejected": -2.859375, "logps/chosen": -592.0, "logps/rejected": -952.0, "loss": 0.2017, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.3125, "rewards/margins": 3.734375, "rewards/rejected": -8.0625, "step": 4210 }, { "epoch": 0.5393315866828551, "grad_norm": 8.083638442585006, "learning_rate": 2.593119693475193e-07, "logits/chosen": -2.90625, "logits/rejected": -2.828125, "logps/chosen": -604.0, "logps/rejected": -968.0, "loss": 0.1904, "rewards/accuracies": 0.9375, "rewards/chosen": -4.5625, "rewards/margins": 3.765625, "rewards/rejected": -8.3125, "step": 4220 }, { "epoch": 0.5406096236181226, "grad_norm": 9.85692839498437, "learning_rate": 2.5819719047791467e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -596.0, "logps/rejected": -960.0, "loss": 0.2145, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.46875, "rewards/margins": 3.515625, "rewards/rejected": -8.0, "step": 4230 }, { "epoch": 0.54188766055339, "grad_norm": 8.33356041058765, "learning_rate": 2.570822484176257e-07, "logits/chosen": -3.046875, "logits/rejected": -2.625, "logps/chosen": -572.0, "logps/rejected": -920.0, "loss": 0.1869, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.28125, "rewards/margins": 3.515625, "rewards/rejected": -7.8125, "step": 4240 }, { "epoch": 0.5431656974886574, "grad_norm": 10.85907589382582, "learning_rate": 2.559671653630574e-07, "logits/chosen": -3.0625, "logits/rejected": -2.859375, "logps/chosen": -612.0, "logps/rejected": -988.0, "loss": 0.1961, "rewards/accuracies": 0.9375, "rewards/chosen": -4.65625, "rewards/margins": 3.859375, "rewards/rejected": -8.5, "step": 4250 }, { "epoch": 0.5444437344239248, "grad_norm": 12.760498280651152, "learning_rate": 2.5485196351342146e-07, "logits/chosen": -3.0625, "logits/rejected": -2.78125, "logps/chosen": -596.0, "logps/rejected": -964.0, "loss": 0.2171, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.375, "rewards/margins": 3.78125, "rewards/rejected": -8.1875, "step": 4260 }, { "epoch": 0.5457217713591923, "grad_norm": 11.083493467932884, "learning_rate": 2.537366650702944e-07, "logits/chosen": -2.96875, "logits/rejected": -2.84375, "logps/chosen": -612.0, "logps/rejected": -960.0, "loss": 0.2058, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.4375, "rewards/margins": 3.609375, "rewards/rejected": -8.0625, "step": 4270 }, { "epoch": 0.5469998082944597, "grad_norm": 9.755652576792773, "learning_rate": 2.526212922371758e-07, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -616.0, "logps/rejected": -948.0, "loss": 0.2014, "rewards/accuracies": 0.90625, "rewards/chosen": -4.625, "rewards/margins": 3.390625, "rewards/rejected": -8.0, "step": 4280 }, { "epoch": 0.5482778452297271, "grad_norm": 8.621966933478847, "learning_rate": 2.5150586721904653e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -592.0, "logps/rejected": -980.0, "loss": 0.1909, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.28125, "rewards/margins": 4.09375, "rewards/rejected": -8.375, "step": 4290 }, { "epoch": 0.5495558821649945, "grad_norm": 9.93825493293039, "learning_rate": 2.50390412221926e-07, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -588.0, "logps/rejected": -952.0, "loss": 0.2184, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.4375, "rewards/margins": 3.734375, "rewards/rejected": -8.125, "step": 4300 }, { "epoch": 0.550833919100262, "grad_norm": 6.4926059090968735, "learning_rate": 2.492749494524305e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -592.0, "logps/rejected": -936.0, "loss": 0.2104, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.34375, "rewards/margins": 3.5, "rewards/rejected": -7.875, "step": 4310 }, { "epoch": 0.5521119560355294, "grad_norm": 8.178900908424412, "learning_rate": 2.481595011173312e-07, "logits/chosen": -2.96875, "logits/rejected": -2.71875, "logps/chosen": -568.0, "logps/rejected": -924.0, "loss": 0.2101, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.25, "rewards/margins": 3.515625, "rewards/rejected": -7.75, "step": 4320 }, { "epoch": 0.5533899929707968, "grad_norm": 8.902539400477867, "learning_rate": 2.470440894231118e-07, "logits/chosen": -2.84375, "logits/rejected": -2.65625, "logps/chosen": -636.0, "logps/rejected": -940.0, "loss": 0.2214, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.203125, "rewards/rejected": -7.96875, "step": 4330 }, { "epoch": 0.5546680299060642, "grad_norm": 10.017993004725309, "learning_rate": 2.4592873657552665e-07, "logits/chosen": -3.015625, "logits/rejected": -2.9375, "logps/chosen": -624.0, "logps/rejected": -948.0, "loss": 0.2036, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.59375, "rewards/margins": 3.40625, "rewards/rejected": -8.0, "step": 4340 }, { "epoch": 0.5559460668413317, "grad_norm": 9.086190763346574, "learning_rate": 2.4481346477915834e-07, "logits/chosen": -3.125, "logits/rejected": -2.9375, "logps/chosen": -580.0, "logps/rejected": -924.0, "loss": 0.1972, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.21875, "rewards/margins": 3.5625, "rewards/rejected": -7.78125, "step": 4350 }, { "epoch": 0.5572241037765991, "grad_norm": 8.170239096649071, "learning_rate": 2.436982962369761e-07, "logits/chosen": -3.125, "logits/rejected": -2.9375, "logps/chosen": -636.0, "logps/rejected": -960.0, "loss": 0.1941, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.6875, "rewards/margins": 3.6875, "rewards/rejected": -8.375, "step": 4360 }, { "epoch": 0.5585021407118665, "grad_norm": 9.91347550249988, "learning_rate": 2.4258325314989364e-07, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -672.0, "logps/rejected": -1056.0, "loss": 0.1968, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.21875, "rewards/margins": 3.75, "rewards/rejected": -8.9375, "step": 4370 }, { "epoch": 0.559780177647134, "grad_norm": 12.649043344414686, "learning_rate": 2.4146835771632674e-07, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -668.0, "logps/rejected": -980.0, "loss": 0.2218, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -5.125, "rewards/margins": 3.359375, "rewards/rejected": -8.5, "step": 4380 }, { "epoch": 0.5610582145824015, "grad_norm": 10.059811667819446, "learning_rate": 2.4035363213175206e-07, "logits/chosen": -2.765625, "logits/rejected": -2.5625, "logps/chosen": -632.0, "logps/rejected": -944.0, "loss": 0.2293, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.6875, "rewards/margins": 3.484375, "rewards/rejected": -8.1875, "step": 4390 }, { "epoch": 0.5623362515176689, "grad_norm": 7.752963384126441, "learning_rate": 2.3923909858826473e-07, "logits/chosen": -2.921875, "logits/rejected": -2.75, "logps/chosen": -588.0, "logps/rejected": -888.0, "loss": 0.2013, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.34375, "rewards/margins": 3.25, "rewards/rejected": -7.59375, "step": 4400 }, { "epoch": 0.5636142884529363, "grad_norm": 8.121775627856461, "learning_rate": 2.3812477927413648e-07, "logits/chosen": -2.84375, "logits/rejected": -2.78125, "logps/chosen": -640.0, "logps/rejected": -944.0, "loss": 0.1966, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.90625, "rewards/margins": 3.328125, "rewards/rejected": -8.25, "step": 4410 }, { "epoch": 0.5648923253882037, "grad_norm": 10.889500920755431, "learning_rate": 2.370106963733746e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -692.0, "logps/rejected": -1040.0, "loss": 0.2137, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.25, "rewards/margins": 3.421875, "rewards/rejected": -8.6875, "step": 4420 }, { "epoch": 0.5661703623234712, "grad_norm": 7.787520343491865, "learning_rate": 2.358968720652794e-07, "logits/chosen": -3.03125, "logits/rejected": -2.875, "logps/chosen": -624.0, "logps/rejected": -972.0, "loss": 0.2206, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.71875, "rewards/margins": 3.59375, "rewards/rejected": -8.3125, "step": 4430 }, { "epoch": 0.5674483992587386, "grad_norm": 7.334229409175379, "learning_rate": 2.3478332852400337e-07, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -600.0, "logps/rejected": -964.0, "loss": 0.1986, "rewards/accuracies": 0.9375, "rewards/chosen": -4.59375, "rewards/margins": 3.453125, "rewards/rejected": -8.0625, "step": 4440 }, { "epoch": 0.568726436194006, "grad_norm": 9.306378482430759, "learning_rate": 2.3367008791810936e-07, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -672.0, "logps/rejected": -968.0, "loss": 0.2612, "rewards/accuracies": 0.90625, "rewards/chosen": -4.9375, "rewards/margins": 3.3125, "rewards/rejected": -8.25, "step": 4450 }, { "epoch": 0.5700044731292734, "grad_norm": 7.851220791242363, "learning_rate": 2.3255717241012922e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -608.0, "logps/rejected": -976.0, "loss": 0.1944, "rewards/accuracies": 0.9375, "rewards/chosen": -4.3125, "rewards/margins": 3.890625, "rewards/rejected": -8.1875, "step": 4460 }, { "epoch": 0.5712825100645409, "grad_norm": 8.633601777057988, "learning_rate": 2.314446041561231e-07, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -616.0, "logps/rejected": -948.0, "loss": 0.1993, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.6875, "rewards/margins": 3.390625, "rewards/rejected": -8.0625, "step": 4470 }, { "epoch": 0.5725605469998083, "grad_norm": 10.040452596074198, "learning_rate": 2.303324053052374e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -636.0, "logps/rejected": -1008.0, "loss": 0.2164, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.75, "rewards/margins": 3.953125, "rewards/rejected": -8.6875, "step": 4480 }, { "epoch": 0.5738385839350757, "grad_norm": 9.442768793474356, "learning_rate": 2.29220597999265e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -648.0, "logps/rejected": -992.0, "loss": 0.1952, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.5, "rewards/rejected": -8.3125, "step": 4490 }, { "epoch": 0.5751166208703431, "grad_norm": 10.471888355465538, "learning_rate": 2.281092043722034e-07, "logits/chosen": -2.984375, "logits/rejected": -2.78125, "logps/chosen": -644.0, "logps/rejected": -1004.0, "loss": 0.1983, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.734375, "rewards/rejected": -8.4375, "step": 4500 }, { "epoch": 0.5763946578056106, "grad_norm": 10.315495473189406, "learning_rate": 2.269982465498147e-07, "logits/chosen": -2.890625, "logits/rejected": -2.6875, "logps/chosen": -612.0, "logps/rejected": -1020.0, "loss": 0.191, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.71875, "rewards/margins": 3.9375, "rewards/rejected": -8.625, "step": 4510 }, { "epoch": 0.577672694740878, "grad_norm": 9.421774060820772, "learning_rate": 2.2588774664918503e-07, "logits/chosen": -2.8125, "logits/rejected": -2.6875, "logps/chosen": -628.0, "logps/rejected": -968.0, "loss": 0.2269, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.609375, "rewards/rejected": -8.4375, "step": 4520 }, { "epoch": 0.5789507316761454, "grad_norm": 10.653247108656883, "learning_rate": 2.2477772677828396e-07, "logits/chosen": -2.90625, "logits/rejected": -2.671875, "logps/chosen": -644.0, "logps/rejected": -972.0, "loss": 0.2055, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.46875, "rewards/rejected": -8.3125, "step": 4530 }, { "epoch": 0.5802287686114128, "grad_norm": 12.414812092039039, "learning_rate": 2.2366820903552478e-07, "logits/chosen": -2.828125, "logits/rejected": -2.609375, "logps/chosen": -640.0, "logps/rejected": -956.0, "loss": 0.2077, "rewards/accuracies": 0.90625, "rewards/chosen": -4.84375, "rewards/margins": 3.25, "rewards/rejected": -8.125, "step": 4540 }, { "epoch": 0.5815068055466803, "grad_norm": 15.625418407059193, "learning_rate": 2.2255921550932418e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -680.0, "logps/rejected": -1064.0, "loss": 0.1943, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.1875, "rewards/margins": 3.90625, "rewards/rejected": -9.0625, "step": 4550 }, { "epoch": 0.5827848424819477, "grad_norm": 9.748298215017456, "learning_rate": 2.214507682776627e-07, "logits/chosen": -2.90625, "logits/rejected": -2.671875, "logps/chosen": -688.0, "logps/rejected": -1072.0, "loss": 0.1975, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.28125, "rewards/margins": 3.921875, "rewards/rejected": -9.1875, "step": 4560 }, { "epoch": 0.5840628794172151, "grad_norm": 9.206249587748667, "learning_rate": 2.203428894076454e-07, "logits/chosen": -2.75, "logits/rejected": -2.609375, "logps/chosen": -648.0, "logps/rejected": -956.0, "loss": 0.2063, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.25, "rewards/rejected": -8.125, "step": 4570 }, { "epoch": 0.5853409163524825, "grad_norm": 10.742870208744552, "learning_rate": 2.1923560095506196e-07, "logits/chosen": -2.84375, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -992.0, "loss": 0.2239, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.734375, "rewards/rejected": -8.5625, "step": 4580 }, { "epoch": 0.58661895328775, "grad_norm": 9.987652609751306, "learning_rate": 2.1812892496394835e-07, "logits/chosen": -2.90625, "logits/rejected": -2.8125, "logps/chosen": -672.0, "logps/rejected": -980.0, "loss": 0.2105, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.0625, "rewards/margins": 3.40625, "rewards/rejected": -8.5, "step": 4590 }, { "epoch": 0.5878969902230174, "grad_norm": 10.530610611497671, "learning_rate": 2.170228834661472e-07, "logits/chosen": -3.140625, "logits/rejected": -2.84375, "logps/chosen": -656.0, "logps/rejected": -1064.0, "loss": 0.1949, "rewards/accuracies": 0.9375, "rewards/chosen": -4.9375, "rewards/margins": 4.28125, "rewards/rejected": -9.25, "step": 4600 }, { "epoch": 0.5891750271582848, "grad_norm": 9.50246727866108, "learning_rate": 2.1591749848086992e-07, "logits/chosen": -2.984375, "logits/rejected": -2.859375, "logps/chosen": -664.0, "logps/rejected": -1020.0, "loss": 0.2007, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.09375, "rewards/margins": 3.8125, "rewards/rejected": -8.9375, "step": 4610 }, { "epoch": 0.5904530640935522, "grad_norm": 8.38828507992591, "learning_rate": 2.148127920142577e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -992.0, "loss": 0.1928, "rewards/accuracies": 0.9375, "rewards/chosen": -4.75, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 4620 }, { "epoch": 0.5917311010288198, "grad_norm": 11.01986987749802, "learning_rate": 2.1370878605894377e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -616.0, "logps/rejected": -988.0, "loss": 0.1884, "rewards/accuracies": 0.9375, "rewards/chosen": -4.53125, "rewards/margins": 3.828125, "rewards/rejected": -8.375, "step": 4630 }, { "epoch": 0.5930091379640872, "grad_norm": 10.543139498864132, "learning_rate": 2.1260550259361576e-07, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -936.0, "loss": 0.2096, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.75, "rewards/margins": 3.09375, "rewards/rejected": -7.84375, "step": 4640 }, { "epoch": 0.5942871748993546, "grad_norm": 9.129152697037581, "learning_rate": 2.115029635825774e-07, "logits/chosen": -2.90625, "logits/rejected": -2.828125, "logps/chosen": -620.0, "logps/rejected": -944.0, "loss": 0.2175, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.75, "rewards/margins": 3.359375, "rewards/rejected": -8.125, "step": 4650 }, { "epoch": 0.595565211834622, "grad_norm": 8.38626724144314, "learning_rate": 2.104011909753122e-07, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -660.0, "logps/rejected": -972.0, "loss": 0.2239, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.4375, "rewards/rejected": -8.375, "step": 4660 }, { "epoch": 0.5968432487698895, "grad_norm": 8.949485168979885, "learning_rate": 2.093002067060457e-07, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -600.0, "logps/rejected": -936.0, "loss": 0.2056, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.46875, "rewards/margins": 3.515625, "rewards/rejected": -7.96875, "step": 4670 }, { "epoch": 0.5981212857051569, "grad_norm": 10.070633848403004, "learning_rate": 2.082000326933092e-07, "logits/chosen": -2.984375, "logits/rejected": -2.859375, "logps/chosen": -600.0, "logps/rejected": -940.0, "loss": 0.2003, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.53125, "rewards/margins": 3.578125, "rewards/rejected": -8.125, "step": 4680 }, { "epoch": 0.5993993226404243, "grad_norm": 6.916035441529172, "learning_rate": 2.0710069083950342e-07, "logits/chosen": -2.921875, "logits/rejected": -2.75, "logps/chosen": -692.0, "logps/rejected": -988.0, "loss": 0.2108, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.375, "rewards/margins": 3.21875, "rewards/rejected": -8.625, "step": 4690 }, { "epoch": 0.6006773595756917, "grad_norm": 9.474816951227877, "learning_rate": 2.060022030304621e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -664.0, "logps/rejected": -988.0, "loss": 0.2323, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.0625, "rewards/margins": 3.515625, "rewards/rejected": -8.5625, "step": 4700 }, { "epoch": 0.6019553965109592, "grad_norm": 6.119960790414473, "learning_rate": 2.0490459113501686e-07, "logits/chosen": -2.921875, "logits/rejected": -2.6875, "logps/chosen": -676.0, "logps/rejected": -984.0, "loss": 0.1866, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.125, "rewards/margins": 3.328125, "rewards/rejected": -8.4375, "step": 4710 }, { "epoch": 0.6032334334462266, "grad_norm": 8.27937601473009, "learning_rate": 2.0380787700456132e-07, "logits/chosen": -3.015625, "logits/rejected": -2.703125, "logps/chosen": -664.0, "logps/rejected": -992.0, "loss": 0.1818, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.0625, "rewards/margins": 3.53125, "rewards/rejected": -8.625, "step": 4720 }, { "epoch": 0.604511470381494, "grad_norm": 8.980086860385386, "learning_rate": 2.027120824726162e-07, "logits/chosen": -3.09375, "logits/rejected": -2.890625, "logps/chosen": -676.0, "logps/rejected": -1016.0, "loss": 0.2223, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.34375, "rewards/margins": 3.375, "rewards/rejected": -8.75, "step": 4730 }, { "epoch": 0.6057895073167614, "grad_norm": 9.660168225938056, "learning_rate": 2.0161722935439513e-07, "logits/chosen": -2.9375, "logits/rejected": -2.640625, "logps/chosen": -648.0, "logps/rejected": -1008.0, "loss": 0.1714, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0, "rewards/margins": 3.859375, "rewards/rejected": -8.875, "step": 4740 }, { "epoch": 0.6070675442520289, "grad_norm": 9.627236009510261, "learning_rate": 2.0052333944636954e-07, "logits/chosen": -3.046875, "logits/rejected": -2.875, "logps/chosen": -620.0, "logps/rejected": -920.0, "loss": 0.2116, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.203125, "rewards/rejected": -7.96875, "step": 4750 }, { "epoch": 0.6083455811872963, "grad_norm": 11.942856212349632, "learning_rate": 1.994304345258354e-07, "logits/chosen": -3.046875, "logits/rejected": -2.953125, "logps/chosen": -644.0, "logps/rejected": -956.0, "loss": 0.1951, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.5, "rewards/rejected": -8.25, "step": 4760 }, { "epoch": 0.6096236181225637, "grad_norm": 8.668637933993713, "learning_rate": 1.983385363504793e-07, "logits/chosen": -2.984375, "logits/rejected": -2.671875, "logps/chosen": -652.0, "logps/rejected": -1020.0, "loss": 0.1902, "rewards/accuracies": 0.9375, "rewards/chosen": -5.0, "rewards/margins": 3.8125, "rewards/rejected": -8.8125, "step": 4770 }, { "epoch": 0.6109016550578311, "grad_norm": 8.513667560083928, "learning_rate": 1.9724766665794541e-07, "logits/chosen": -3.0625, "logits/rejected": -2.9375, "logps/chosen": -688.0, "logps/rejected": -996.0, "loss": 0.1825, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 3.328125, "rewards/rejected": -8.625, "step": 4780 }, { "epoch": 0.6121796919930986, "grad_norm": 12.364360217821515, "learning_rate": 1.9615784716540293e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -672.0, "logps/rejected": -1088.0, "loss": 0.1966, "rewards/accuracies": 0.9375, "rewards/chosen": -5.15625, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 4790 }, { "epoch": 0.613457728928366, "grad_norm": 13.291905969721395, "learning_rate": 1.950690995691132e-07, "logits/chosen": -2.96875, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -1056.0, "loss": 0.1938, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.28125, "rewards/margins": 3.78125, "rewards/rejected": -9.0625, "step": 4800 }, { "epoch": 0.6147357658636334, "grad_norm": 12.195901066043673, "learning_rate": 1.939814455439986e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -668.0, "logps/rejected": -1032.0, "loss": 0.2181, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.21875, "rewards/margins": 3.734375, "rewards/rejected": -8.9375, "step": 4810 }, { "epoch": 0.6160138027989008, "grad_norm": 10.210694409079894, "learning_rate": 1.9289490674321017e-07, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -660.0, "logps/rejected": -1056.0, "loss": 0.1927, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.03125, "rewards/margins": 4.15625, "rewards/rejected": -9.1875, "step": 4820 }, { "epoch": 0.6172918397341683, "grad_norm": 11.462328801893005, "learning_rate": 1.9180950479769687e-07, "logits/chosen": -2.875, "logits/rejected": -2.75, "logps/chosen": -688.0, "logps/rejected": -1040.0, "loss": 0.2051, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.1875, "rewards/margins": 3.796875, "rewards/rejected": -9.0, "step": 4830 }, { "epoch": 0.6185698766694357, "grad_norm": 11.618759475704962, "learning_rate": 1.9072526131577542e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -688.0, "logps/rejected": -1064.0, "loss": 0.1983, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.1875, "rewards/margins": 3.84375, "rewards/rejected": -9.0, "step": 4840 }, { "epoch": 0.6198479136047031, "grad_norm": 8.771287614014419, "learning_rate": 1.8964219788269936e-07, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -692.0, "logps/rejected": -1032.0, "loss": 0.2049, "rewards/accuracies": 0.90625, "rewards/chosen": -5.34375, "rewards/margins": 3.609375, "rewards/rejected": -8.9375, "step": 4850 }, { "epoch": 0.6211259505399706, "grad_norm": 8.382633721719383, "learning_rate": 1.885603360602298e-07, "logits/chosen": -3.109375, "logits/rejected": -2.78125, "logps/chosen": -652.0, "logps/rejected": -1020.0, "loss": 0.1928, "rewards/accuracies": 0.9375, "rewards/chosen": -5.03125, "rewards/margins": 3.796875, "rewards/rejected": -8.8125, "step": 4860 }, { "epoch": 0.6224039874752381, "grad_norm": 11.874740935105166, "learning_rate": 1.8747969738620588e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -648.0, "logps/rejected": -1008.0, "loss": 0.213, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.90625, "rewards/margins": 3.703125, "rewards/rejected": -8.625, "step": 4870 }, { "epoch": 0.6236820244105055, "grad_norm": 10.358959890573601, "learning_rate": 1.8640030337411655e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -676.0, "logps/rejected": -1016.0, "loss": 0.2078, "rewards/accuracies": 0.875, "rewards/chosen": -5.25, "rewards/margins": 3.453125, "rewards/rejected": -8.6875, "step": 4880 }, { "epoch": 0.6249600613457729, "grad_norm": 6.897393388027196, "learning_rate": 1.8532217551267143e-07, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -652.0, "logps/rejected": -1020.0, "loss": 0.211, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.84375, "rewards/rejected": -8.625, "step": 4890 }, { "epoch": 0.6262380982810403, "grad_norm": 8.014943203049432, "learning_rate": 1.842453352653735e-07, "logits/chosen": -3.046875, "logits/rejected": -2.875, "logps/chosen": -652.0, "logps/rejected": -980.0, "loss": 0.1946, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.96875, "rewards/margins": 3.4375, "rewards/rejected": -8.375, "step": 4900 }, { "epoch": 0.6275161352163078, "grad_norm": 12.135630997047638, "learning_rate": 1.8316980407009213e-07, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -604.0, "logps/rejected": -964.0, "loss": 0.2056, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.5625, "rewards/margins": 3.671875, "rewards/rejected": -8.25, "step": 4910 }, { "epoch": 0.6287941721515752, "grad_norm": 8.388293855177318, "learning_rate": 1.820956033386355e-07, "logits/chosen": -3.03125, "logits/rejected": -2.875, "logps/chosen": -648.0, "logps/rejected": -1008.0, "loss": 0.2067, "rewards/accuracies": 0.90625, "rewards/chosen": -4.875, "rewards/margins": 3.75, "rewards/rejected": -8.625, "step": 4920 }, { "epoch": 0.6300722090868426, "grad_norm": 8.19919332061552, "learning_rate": 1.8102275445632508e-07, "logits/chosen": -2.90625, "logits/rejected": -2.65625, "logps/chosen": -640.0, "logps/rejected": -1004.0, "loss": 0.2086, "rewards/accuracies": 0.9375, "rewards/chosen": -4.8125, "rewards/margins": 3.734375, "rewards/rejected": -8.5625, "step": 4930 }, { "epoch": 0.63135024602211, "grad_norm": 11.524825983155498, "learning_rate": 1.7995127878156934e-07, "logits/chosen": -2.875, "logits/rejected": -2.78125, "logps/chosen": -696.0, "logps/rejected": -1008.0, "loss": 0.1916, "rewards/accuracies": 0.90625, "rewards/chosen": -5.1875, "rewards/margins": 3.5, "rewards/rejected": -8.6875, "step": 4940 }, { "epoch": 0.6326282829573775, "grad_norm": 11.935553547051724, "learning_rate": 1.788811976454388e-07, "logits/chosen": -2.921875, "logits/rejected": -2.890625, "logps/chosen": -688.0, "logps/rejected": -1008.0, "loss": 0.2033, "rewards/accuracies": 0.875, "rewards/chosen": -5.25, "rewards/margins": 3.4375, "rewards/rejected": -8.6875, "step": 4950 }, { "epoch": 0.6339063198926449, "grad_norm": 8.72541909787161, "learning_rate": 1.7781253235124143e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1016.0, "loss": 0.217, "rewards/accuracies": 0.875, "rewards/chosen": -5.34375, "rewards/margins": 3.421875, "rewards/rejected": -8.75, "step": 4960 }, { "epoch": 0.6351843568279123, "grad_norm": 9.174733860455342, "learning_rate": 1.7674530417409823e-07, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -636.0, "logps/rejected": -996.0, "loss": 0.1636, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.6875, "rewards/margins": 3.734375, "rewards/rejected": -8.4375, "step": 4970 }, { "epoch": 0.6364623937631797, "grad_norm": 8.931555059599443, "learning_rate": 1.7567953436052025e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -660.0, "logps/rejected": -1032.0, "loss": 0.219, "rewards/accuracies": 0.9375, "rewards/chosen": -5.03125, "rewards/margins": 3.78125, "rewards/rejected": -8.8125, "step": 4980 }, { "epoch": 0.6377404306984472, "grad_norm": 14.476630420450562, "learning_rate": 1.7461524412798503e-07, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -664.0, "logps/rejected": -1020.0, "loss": 0.2224, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0625, "rewards/margins": 3.6875, "rewards/rejected": -8.75, "step": 4990 }, { "epoch": 0.6390184676337146, "grad_norm": 9.976279367626784, "learning_rate": 1.735524546645142e-07, "logits/chosen": -3.0625, "logits/rejected": -2.90625, "logps/chosen": -648.0, "logps/rejected": -992.0, "loss": 0.1926, "rewards/accuracies": 0.9375, "rewards/chosen": -4.8125, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 5000 }, { "epoch": 0.640296504568982, "grad_norm": 10.403381451664748, "learning_rate": 1.7249118712825246e-07, "logits/chosen": -3.09375, "logits/rejected": -2.765625, "logps/chosen": -612.0, "logps/rejected": -924.0, "loss": 0.2016, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.625, "rewards/margins": 3.1875, "rewards/rejected": -7.78125, "step": 5010 }, { "epoch": 0.6415745415042494, "grad_norm": 10.392003155571185, "learning_rate": 1.7143146264704522e-07, "logits/chosen": -2.90625, "logits/rejected": -2.8125, "logps/chosen": -608.0, "logps/rejected": -920.0, "loss": 0.1941, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.21875, "rewards/rejected": -7.8125, "step": 5020 }, { "epoch": 0.6428525784395169, "grad_norm": 9.045623892838288, "learning_rate": 1.70373302318019e-07, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -612.0, "logps/rejected": -1012.0, "loss": 0.2013, "rewards/accuracies": 0.9375, "rewards/chosen": -4.625, "rewards/margins": 3.96875, "rewards/rejected": -8.5625, "step": 5030 }, { "epoch": 0.6441306153747843, "grad_norm": 10.532156784041835, "learning_rate": 1.6931672720716072e-07, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -996.0, "loss": 0.2078, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 5040 }, { "epoch": 0.6454086523100517, "grad_norm": 13.016020098769182, "learning_rate": 1.6826175834889866e-07, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -620.0, "logps/rejected": -976.0, "loss": 0.2018, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.75, "rewards/margins": 3.59375, "rewards/rejected": -8.3125, "step": 5050 }, { "epoch": 0.6466866892453191, "grad_norm": 8.620907567456209, "learning_rate": 1.6720841674568365e-07, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -628.0, "logps/rejected": -1000.0, "loss": 0.2049, "rewards/accuracies": 0.90625, "rewards/chosen": -4.6875, "rewards/margins": 3.953125, "rewards/rejected": -8.625, "step": 5060 }, { "epoch": 0.6479647261805866, "grad_norm": 12.006083637867917, "learning_rate": 1.6615672336757074e-07, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -632.0, "logps/rejected": -956.0, "loss": 0.1859, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.75, "rewards/margins": 3.578125, "rewards/rejected": -8.3125, "step": 5070 }, { "epoch": 0.649242763115854, "grad_norm": 8.58425824054044, "learning_rate": 1.6510669915180238e-07, "logits/chosen": -2.90625, "logits/rejected": -2.796875, "logps/chosen": -632.0, "logps/rejected": -972.0, "loss": 0.1866, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.65625, "rewards/margins": 3.703125, "rewards/rejected": -8.375, "step": 5080 }, { "epoch": 0.6505208000511215, "grad_norm": 10.81312666340287, "learning_rate": 1.6405836500239062e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -664.0, "logps/rejected": -1024.0, "loss": 0.1932, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.9375, "rewards/margins": 3.875, "rewards/rejected": -8.8125, "step": 5090 }, { "epoch": 0.6517988369863889, "grad_norm": 10.6110280495848, "learning_rate": 1.6301174178970162e-07, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -664.0, "logps/rejected": -1048.0, "loss": 0.1746, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.03125, "rewards/margins": 3.875, "rewards/rejected": -8.875, "step": 5100 }, { "epoch": 0.6530768739216564, "grad_norm": 9.639168092530126, "learning_rate": 1.6196685035004015e-07, "logits/chosen": -2.765625, "logits/rejected": -2.625, "logps/chosen": -664.0, "logps/rejected": -1024.0, "loss": 0.188, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.03125, "rewards/margins": 3.71875, "rewards/rejected": -8.75, "step": 5110 }, { "epoch": 0.6543549108569238, "grad_norm": 9.11174679070626, "learning_rate": 1.6092371148523442e-07, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -668.0, "logps/rejected": -1008.0, "loss": 0.2018, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.125, "rewards/margins": 3.515625, "rewards/rejected": -8.625, "step": 5120 }, { "epoch": 0.6556329477921912, "grad_norm": 13.377523718184259, "learning_rate": 1.5988234596222234e-07, "logits/chosen": -2.953125, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -1040.0, "loss": 0.2055, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.78125, "rewards/margins": 4.0, "rewards/rejected": -8.8125, "step": 5130 }, { "epoch": 0.6569109847274586, "grad_norm": 7.383701674519595, "learning_rate": 1.5884277451263772e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -612.0, "logps/rejected": -968.0, "loss": 0.1936, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.53125, "rewards/margins": 3.625, "rewards/rejected": -8.1875, "step": 5140 }, { "epoch": 0.6581890216627261, "grad_norm": 8.953980108485803, "learning_rate": 1.57805017832398e-07, "logits/chosen": -2.90625, "logits/rejected": -2.765625, "logps/chosen": -608.0, "logps/rejected": -932.0, "loss": 0.1813, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.53125, "rewards/margins": 3.5625, "rewards/rejected": -8.125, "step": 5150 }, { "epoch": 0.6594670585979935, "grad_norm": 9.211324366300165, "learning_rate": 1.567690965812918e-07, "logits/chosen": -3.046875, "logits/rejected": -2.671875, "logps/chosen": -588.0, "logps/rejected": -972.0, "loss": 0.1927, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.53125, "rewards/margins": 3.84375, "rewards/rejected": -8.375, "step": 5160 }, { "epoch": 0.6607450955332609, "grad_norm": 10.95169892297747, "learning_rate": 1.5573503138256756e-07, "logits/chosen": -2.75, "logits/rejected": -2.78125, "logps/chosen": -644.0, "logps/rejected": -964.0, "loss": 0.2315, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.421875, "rewards/rejected": -8.25, "step": 5170 }, { "epoch": 0.6620231324685283, "grad_norm": 7.565071174200895, "learning_rate": 1.5470284282252367e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -604.0, "logps/rejected": -972.0, "loss": 0.2028, "rewards/accuracies": 0.90625, "rewards/chosen": -4.5625, "rewards/margins": 3.71875, "rewards/rejected": -8.25, "step": 5180 }, { "epoch": 0.6633011694037958, "grad_norm": 8.40586608624689, "learning_rate": 1.5367255145009784e-07, "logits/chosen": -2.796875, "logits/rejected": -2.6875, "logps/chosen": -636.0, "logps/rejected": -1024.0, "loss": 0.1959, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.9375, "rewards/rejected": -8.75, "step": 5190 }, { "epoch": 0.6645792063390632, "grad_norm": 8.935789239896653, "learning_rate": 1.5264417777645849e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -616.0, "logps/rejected": -1040.0, "loss": 0.1954, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.625, "rewards/margins": 4.28125, "rewards/rejected": -8.9375, "step": 5200 }, { "epoch": 0.6658572432743306, "grad_norm": 10.318660688920453, "learning_rate": 1.516177422745962e-07, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -632.0, "logps/rejected": -980.0, "loss": 0.2219, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.625, "rewards/rejected": -8.25, "step": 5210 }, { "epoch": 0.667135280209598, "grad_norm": 10.599513311823376, "learning_rate": 1.5059326537891602e-07, "logits/chosen": -2.84375, "logits/rejected": -2.640625, "logps/chosen": -644.0, "logps/rejected": -996.0, "loss": 0.2089, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.9375, "rewards/margins": 3.625, "rewards/rejected": -8.5625, "step": 5220 }, { "epoch": 0.6684133171448655, "grad_norm": 7.096941650990519, "learning_rate": 1.4957076748483121e-07, "logits/chosen": -2.875, "logits/rejected": -2.578125, "logps/chosen": -652.0, "logps/rejected": -1032.0, "loss": 0.1783, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -4.875, "rewards/margins": 3.890625, "rewards/rejected": -8.75, "step": 5230 }, { "epoch": 0.6696913540801329, "grad_norm": 9.134873030915466, "learning_rate": 1.4855026894835636e-07, "logits/chosen": -2.765625, "logits/rejected": -2.609375, "logps/chosen": -644.0, "logps/rejected": -980.0, "loss": 0.2057, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.9375, "rewards/margins": 3.328125, "rewards/rejected": -8.25, "step": 5240 }, { "epoch": 0.6709693910154003, "grad_norm": 9.184838305556584, "learning_rate": 1.4753179008570304e-07, "logits/chosen": -2.84375, "logits/rejected": -2.609375, "logps/chosen": -632.0, "logps/rejected": -968.0, "loss": 0.1984, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.453125, "rewards/rejected": -8.25, "step": 5250 }, { "epoch": 0.6722474279506677, "grad_norm": 10.491068837200642, "learning_rate": 1.4651535117287462e-07, "logits/chosen": -2.921875, "logits/rejected": -2.796875, "logps/chosen": -668.0, "logps/rejected": -1032.0, "loss": 0.1957, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.8125, "rewards/rejected": -8.6875, "step": 5260 }, { "epoch": 0.6735254648859352, "grad_norm": 13.654106335755474, "learning_rate": 1.4550097244526294e-07, "logits/chosen": -3.015625, "logits/rejected": -2.640625, "logps/chosen": -612.0, "logps/rejected": -976.0, "loss": 0.2077, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.625, "rewards/margins": 3.65625, "rewards/rejected": -8.25, "step": 5270 }, { "epoch": 0.6748035018212026, "grad_norm": 7.2043046993448305, "learning_rate": 1.444886740972455e-07, "logits/chosen": -2.78125, "logits/rejected": -2.609375, "logps/chosen": -620.0, "logps/rejected": -952.0, "loss": 0.2154, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.53125, "rewards/margins": 3.640625, "rewards/rejected": -8.1875, "step": 5280 }, { "epoch": 0.67608153875647, "grad_norm": 10.381553344000324, "learning_rate": 1.4347847628178315e-07, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -616.0, "logps/rejected": -932.0, "loss": 0.1997, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.53125, "rewards/margins": 3.390625, "rewards/rejected": -7.9375, "step": 5290 }, { "epoch": 0.6773595756917374, "grad_norm": 8.206259981152979, "learning_rate": 1.4247039911001952e-07, "logits/chosen": -2.828125, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -952.0, "loss": 0.2015, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.234375, "rewards/rejected": -8.0625, "step": 5300 }, { "epoch": 0.678637612627005, "grad_norm": 9.733010610608309, "learning_rate": 1.4146446265087966e-07, "logits/chosen": -3.078125, "logits/rejected": -2.65625, "logps/chosen": -648.0, "logps/rejected": -1012.0, "loss": 0.1698, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.90625, "rewards/margins": 3.796875, "rewards/rejected": -8.6875, "step": 5310 }, { "epoch": 0.6799156495622724, "grad_norm": 7.818977341913373, "learning_rate": 1.4046068693067142e-07, "logits/chosen": -2.875, "logits/rejected": -2.65625, "logps/chosen": -644.0, "logps/rejected": -988.0, "loss": 0.1922, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.8125, "rewards/margins": 3.640625, "rewards/rejected": -8.4375, "step": 5320 }, { "epoch": 0.6811936864975398, "grad_norm": 11.569940395290942, "learning_rate": 1.3945909193268618e-07, "logits/chosen": -2.828125, "logits/rejected": -2.65625, "logps/chosen": -684.0, "logps/rejected": -1040.0, "loss": 0.2018, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.1875, "rewards/margins": 3.9375, "rewards/rejected": -9.125, "step": 5330 }, { "epoch": 0.6824717234328072, "grad_norm": 7.915367886076844, "learning_rate": 1.3845969759680125e-07, "logits/chosen": -2.890625, "logits/rejected": -2.734375, "logps/chosen": -696.0, "logps/rejected": -1040.0, "loss": 0.19, "rewards/accuracies": 0.9375, "rewards/chosen": -5.34375, "rewards/margins": 3.640625, "rewards/rejected": -9.0, "step": 5340 }, { "epoch": 0.6837497603680747, "grad_norm": 10.40013069384589, "learning_rate": 1.374625238190832e-07, "logits/chosen": -2.8125, "logits/rejected": -2.65625, "logps/chosen": -624.0, "logps/rejected": -984.0, "loss": 0.1884, "rewards/accuracies": 0.90625, "rewards/chosen": -4.78125, "rewards/margins": 3.609375, "rewards/rejected": -8.375, "step": 5350 }, { "epoch": 0.6850277973033421, "grad_norm": 8.748911256197829, "learning_rate": 1.36467590451391e-07, "logits/chosen": -2.84375, "logits/rejected": -2.59375, "logps/chosen": -668.0, "logps/rejected": -1008.0, "loss": 0.2055, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.90625, "rewards/margins": 3.65625, "rewards/rejected": -8.5625, "step": 5360 }, { "epoch": 0.6863058342386095, "grad_norm": 10.775546531691665, "learning_rate": 1.354749173009815e-07, "logits/chosen": -2.84375, "logits/rejected": -2.625, "logps/chosen": -664.0, "logps/rejected": -1004.0, "loss": 0.2074, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.0625, "rewards/margins": 3.4375, "rewards/rejected": -8.5, "step": 5370 }, { "epoch": 0.6875838711738769, "grad_norm": 10.000688484948581, "learning_rate": 1.3448452413011487e-07, "logits/chosen": -2.890625, "logits/rejected": -2.765625, "logps/chosen": -684.0, "logps/rejected": -1064.0, "loss": 0.1916, "rewards/accuracies": 0.90625, "rewards/chosen": -5.28125, "rewards/margins": 3.875, "rewards/rejected": -9.1875, "step": 5380 }, { "epoch": 0.6888619081091444, "grad_norm": 8.846111947475448, "learning_rate": 1.3349643065566117e-07, "logits/chosen": -2.828125, "logits/rejected": -2.609375, "logps/chosen": -648.0, "logps/rejected": -980.0, "loss": 0.2055, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.875, "rewards/margins": 3.28125, "rewards/rejected": -8.125, "step": 5390 }, { "epoch": 0.6901399450444118, "grad_norm": 6.691267912708803, "learning_rate": 1.325106565487077e-07, "logits/chosen": -2.890625, "logits/rejected": -2.640625, "logps/chosen": -616.0, "logps/rejected": -992.0, "loss": 0.169, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.4375, "rewards/margins": 4.0625, "rewards/rejected": -8.5, "step": 5400 }, { "epoch": 0.6914179819796792, "grad_norm": 10.58421750198331, "learning_rate": 1.3152722143416748e-07, "logits/chosen": -2.875, "logits/rejected": -2.640625, "logps/chosen": -628.0, "logps/rejected": -972.0, "loss": 0.2012, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.78125, "rewards/rejected": -8.4375, "step": 5410 }, { "epoch": 0.6926960189149466, "grad_norm": 10.823577363873508, "learning_rate": 1.305461448903889e-07, "logits/chosen": -2.875, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -948.0, "loss": 0.1942, "rewards/accuracies": 0.90625, "rewards/chosen": -4.8125, "rewards/margins": 3.328125, "rewards/rejected": -8.125, "step": 5420 }, { "epoch": 0.6939740558502141, "grad_norm": 10.703964915891877, "learning_rate": 1.2956744644876527e-07, "logits/chosen": -2.859375, "logits/rejected": -2.78125, "logps/chosen": -632.0, "logps/rejected": -976.0, "loss": 0.1891, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.6875, "rewards/margins": 3.734375, "rewards/rejected": -8.4375, "step": 5430 }, { "epoch": 0.6952520927854815, "grad_norm": 11.074215783268322, "learning_rate": 1.2859114559334643e-07, "logits/chosen": -2.921875, "logits/rejected": -2.625, "logps/chosen": -624.0, "logps/rejected": -972.0, "loss": 0.2077, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.609375, "rewards/rejected": -8.1875, "step": 5440 }, { "epoch": 0.6965301297207489, "grad_norm": 9.906928540767801, "learning_rate": 1.27617261760451e-07, "logits/chosen": -2.9375, "logits/rejected": -2.421875, "logps/chosen": -596.0, "logps/rejected": -996.0, "loss": 0.175, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.5, "rewards/margins": 4.0, "rewards/rejected": -8.5, "step": 5450 }, { "epoch": 0.6978081666560163, "grad_norm": 7.989791957700388, "learning_rate": 1.26645814338279e-07, "logits/chosen": -2.9375, "logits/rejected": -2.859375, "logps/chosen": -596.0, "logps/rejected": -944.0, "loss": 0.2011, "rewards/accuracies": 0.875, "rewards/chosen": -4.5, "rewards/margins": 3.625, "rewards/rejected": -8.125, "step": 5460 }, { "epoch": 0.6990862035912838, "grad_norm": 9.439843209419791, "learning_rate": 1.2567682266652603e-07, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -588.0, "logps/rejected": -956.0, "loss": 0.1897, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.34375, "rewards/margins": 3.875, "rewards/rejected": -8.25, "step": 5470 }, { "epoch": 0.7003642405265512, "grad_norm": 12.246737461104036, "learning_rate": 1.2471030603599848e-07, "logits/chosen": -2.875, "logits/rejected": -2.609375, "logps/chosen": -608.0, "logps/rejected": -984.0, "loss": 0.2083, "rewards/accuracies": 0.90625, "rewards/chosen": -4.5, "rewards/margins": 3.953125, "rewards/rejected": -8.4375, "step": 5480 }, { "epoch": 0.7016422774618186, "grad_norm": 10.904656119606363, "learning_rate": 1.2374628368822914e-07, "logits/chosen": -2.84375, "logits/rejected": -2.65625, "logps/chosen": -596.0, "logps/rejected": -948.0, "loss": 0.1955, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.4375, "rewards/margins": 3.734375, "rewards/rejected": -8.1875, "step": 5490 }, { "epoch": 0.702920314397086, "grad_norm": 10.223805884896036, "learning_rate": 1.2278477481509462e-07, "logits/chosen": -2.921875, "logits/rejected": -2.765625, "logps/chosen": -592.0, "logps/rejected": -936.0, "loss": 0.1922, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.375, "rewards/margins": 3.5625, "rewards/rejected": -7.9375, "step": 5500 }, { "epoch": 0.7041983513323535, "grad_norm": 10.18405106135057, "learning_rate": 1.2182579855843247e-07, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -592.0, "logps/rejected": -940.0, "loss": 0.1946, "rewards/accuracies": 0.9375, "rewards/chosen": -4.59375, "rewards/margins": 3.515625, "rewards/rejected": -8.125, "step": 5510 }, { "epoch": 0.7054763882676209, "grad_norm": 9.516374791611842, "learning_rate": 1.2086937400966129e-07, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -624.0, "logps/rejected": -992.0, "loss": 0.1833, "rewards/accuracies": 0.96875, "rewards/chosen": -4.625, "rewards/margins": 3.984375, "rewards/rejected": -8.625, "step": 5520 }, { "epoch": 0.7067544252028883, "grad_norm": 9.178417627197618, "learning_rate": 1.1991552020939918e-07, "logits/chosen": -2.890625, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -988.0, "loss": 0.1939, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.71875, "rewards/margins": 3.53125, "rewards/rejected": -8.25, "step": 5530 }, { "epoch": 0.7080324621381557, "grad_norm": 7.880000574893539, "learning_rate": 1.1896425614708589e-07, "logits/chosen": -2.875, "logits/rejected": -2.765625, "logps/chosen": -620.0, "logps/rejected": -964.0, "loss": 0.1916, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.71875, "rewards/margins": 3.46875, "rewards/rejected": -8.1875, "step": 5540 }, { "epoch": 0.7093104990734233, "grad_norm": 8.442036728841282, "learning_rate": 1.1801560076060444e-07, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -628.0, "logps/rejected": -1012.0, "loss": 0.1796, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.78125, "rewards/margins": 4.03125, "rewards/rejected": -8.8125, "step": 5550 }, { "epoch": 0.7105885360086907, "grad_norm": 9.96646161178662, "learning_rate": 1.170695729359037e-07, "logits/chosen": -2.765625, "logits/rejected": -2.734375, "logps/chosen": -668.0, "logps/rejected": -968.0, "loss": 0.1884, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.0, "rewards/margins": 3.265625, "rewards/rejected": -8.25, "step": 5560 }, { "epoch": 0.711866572943958, "grad_norm": 9.055717430232225, "learning_rate": 1.1612619150662303e-07, "logits/chosen": -2.859375, "logits/rejected": -2.703125, "logps/chosen": -648.0, "logps/rejected": -1032.0, "loss": 0.1913, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.84375, "rewards/margins": 4.125, "rewards/rejected": -9.0, "step": 5570 }, { "epoch": 0.7131446098792255, "grad_norm": 10.28494479666354, "learning_rate": 1.151854752537168e-07, "logits/chosen": -2.96875, "logits/rejected": -2.640625, "logps/chosen": -648.0, "logps/rejected": -1016.0, "loss": 0.2001, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.03125, "rewards/margins": 3.484375, "rewards/rejected": -8.5, "step": 5580 }, { "epoch": 0.714422646814493, "grad_norm": 32.22289570587893, "learning_rate": 1.1424744290508078e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -664.0, "logps/rejected": -1000.0, "loss": 0.189, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.9375, "rewards/margins": 3.71875, "rewards/rejected": -8.625, "step": 5590 }, { "epoch": 0.7157006837497604, "grad_norm": 12.269299820154668, "learning_rate": 1.1331211313517925e-07, "logits/chosen": -2.90625, "logits/rejected": -2.5625, "logps/chosen": -640.0, "logps/rejected": -1024.0, "loss": 0.2084, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 4.03125, "rewards/rejected": -8.8125, "step": 5600 }, { "epoch": 0.7169787206850278, "grad_norm": 9.260295940193947, "learning_rate": 1.1237950456467308e-07, "logits/chosen": -2.859375, "logits/rejected": -2.65625, "logps/chosen": -668.0, "logps/rejected": -1032.0, "loss": 0.19, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.9375, "rewards/margins": 3.6875, "rewards/rejected": -8.625, "step": 5610 }, { "epoch": 0.7182567576202952, "grad_norm": 10.398934473010499, "learning_rate": 1.1144963576004954e-07, "logits/chosen": -2.9375, "logits/rejected": -2.671875, "logps/chosen": -644.0, "logps/rejected": -996.0, "loss": 0.2061, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.875, "rewards/margins": 3.765625, "rewards/rejected": -8.625, "step": 5620 }, { "epoch": 0.7195347945555627, "grad_norm": 12.87874977657243, "learning_rate": 1.105225252332519e-07, "logits/chosen": -2.859375, "logits/rejected": -2.625, "logps/chosen": -660.0, "logps/rejected": -1040.0, "loss": 0.2031, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.96875, "rewards/margins": 3.984375, "rewards/rejected": -8.9375, "step": 5630 }, { "epoch": 0.7208128314908301, "grad_norm": 10.678499925600923, "learning_rate": 1.0959819144131144e-07, "logits/chosen": -2.78125, "logits/rejected": -2.578125, "logps/chosen": -624.0, "logps/rejected": -1012.0, "loss": 0.188, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -4.6875, "rewards/margins": 4.0, "rewards/rejected": -8.6875, "step": 5640 }, { "epoch": 0.7220908684260975, "grad_norm": 10.020846983046951, "learning_rate": 1.0867665278597999e-07, "logits/chosen": -2.921875, "logits/rejected": -2.59375, "logps/chosen": -672.0, "logps/rejected": -1020.0, "loss": 0.1866, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.03125, "rewards/margins": 3.796875, "rewards/rejected": -8.8125, "step": 5650 }, { "epoch": 0.7233689053613649, "grad_norm": 11.621911848428802, "learning_rate": 1.0775792761336324e-07, "logits/chosen": -2.84375, "logits/rejected": -2.671875, "logps/chosen": -636.0, "logps/rejected": -976.0, "loss": 0.1806, "rewards/accuracies": 0.90625, "rewards/chosen": -4.71875, "rewards/margins": 3.6875, "rewards/rejected": -8.4375, "step": 5660 }, { "epoch": 0.7246469422966324, "grad_norm": 10.137810011465284, "learning_rate": 1.0684203421355587e-07, "logits/chosen": -2.9375, "logits/rejected": -2.765625, "logps/chosen": -664.0, "logps/rejected": -1072.0, "loss": 0.1977, "rewards/accuracies": 0.9375, "rewards/chosen": -4.84375, "rewards/margins": 4.1875, "rewards/rejected": -9.0, "step": 5670 }, { "epoch": 0.7259249792318998, "grad_norm": 7.480549183495816, "learning_rate": 1.0592899082027715e-07, "logits/chosen": -2.8125, "logits/rejected": -2.53125, "logps/chosen": -612.0, "logps/rejected": -968.0, "loss": 0.1947, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.5625, "rewards/margins": 3.765625, "rewards/rejected": -8.3125, "step": 5680 }, { "epoch": 0.7272030161671672, "grad_norm": 8.95863895797164, "learning_rate": 1.050188156105081e-07, "logits/chosen": -2.8125, "logits/rejected": -2.640625, "logps/chosen": -656.0, "logps/rejected": -996.0, "loss": 0.2002, "rewards/accuracies": 0.90625, "rewards/chosen": -4.875, "rewards/margins": 3.5625, "rewards/rejected": -8.4375, "step": 5690 }, { "epoch": 0.7284810531024346, "grad_norm": 10.004556249685614, "learning_rate": 1.0411152670412975e-07, "logits/chosen": -2.859375, "logits/rejected": -2.5625, "logps/chosen": -620.0, "logps/rejected": -1012.0, "loss": 0.1985, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.6875, "rewards/margins": 3.8125, "rewards/rejected": -8.5, "step": 5700 }, { "epoch": 0.7297590900377021, "grad_norm": 9.942783279630579, "learning_rate": 1.03207142163562e-07, "logits/chosen": -2.828125, "logits/rejected": -2.65625, "logps/chosen": -668.0, "logps/rejected": -992.0, "loss": 0.1943, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.0, "rewards/margins": 3.578125, "rewards/rejected": -8.5625, "step": 5710 }, { "epoch": 0.7310371269729695, "grad_norm": 11.150110790338827, "learning_rate": 1.023056799934045e-07, "logits/chosen": -2.90625, "logits/rejected": -2.640625, "logps/chosen": -660.0, "logps/rejected": -1012.0, "loss": 0.2104, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.9375, "rewards/margins": 3.828125, "rewards/rejected": -8.75, "step": 5720 }, { "epoch": 0.7323151639082369, "grad_norm": 11.60741920901857, "learning_rate": 1.0140715814007783e-07, "logits/chosen": -2.84375, "logits/rejected": -2.609375, "logps/chosen": -688.0, "logps/rejected": -992.0, "loss": 0.2133, "rewards/accuracies": 0.875, "rewards/chosen": -5.34375, "rewards/margins": 3.296875, "rewards/rejected": -8.625, "step": 5730 }, { "epoch": 0.7335932008435043, "grad_norm": 11.162758346658364, "learning_rate": 1.0051159449146643e-07, "logits/chosen": -2.796875, "logits/rejected": -2.40625, "logps/chosen": -644.0, "logps/rejected": -956.0, "loss": 0.1933, "rewards/accuracies": 0.90625, "rewards/chosen": -4.9375, "rewards/margins": 3.265625, "rewards/rejected": -8.1875, "step": 5740 }, { "epoch": 0.7348712377787718, "grad_norm": 8.587527124699529, "learning_rate": 9.961900687656242e-08, "logits/chosen": -2.921875, "logits/rejected": -2.609375, "logps/chosen": -652.0, "logps/rejected": -984.0, "loss": 0.1919, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.09375, "rewards/margins": 3.34375, "rewards/rejected": -8.4375, "step": 5750 }, { "epoch": 0.7361492747140392, "grad_norm": 13.109963319894257, "learning_rate": 9.872941306511056e-08, "logits/chosen": -2.828125, "logits/rejected": -2.515625, "logps/chosen": -620.0, "logps/rejected": -972.0, "loss": 0.1747, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.625, "rewards/margins": 3.703125, "rewards/rejected": -8.3125, "step": 5760 }, { "epoch": 0.7374273116493066, "grad_norm": 9.579077500243383, "learning_rate": 9.784283076725494e-08, "logits/chosen": -2.828125, "logits/rejected": -2.6875, "logps/chosen": -612.0, "logps/rejected": -928.0, "loss": 0.176, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.65625, "rewards/margins": 3.34375, "rewards/rejected": -8.0, "step": 5770 }, { "epoch": 0.738705348584574, "grad_norm": 10.231998859747685, "learning_rate": 9.695927763318568e-08, "logits/chosen": -2.921875, "logits/rejected": -2.546875, "logps/chosen": -636.0, "logps/rejected": -976.0, "loss": 0.2214, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.546875, "rewards/rejected": -8.375, "step": 5780 }, { "epoch": 0.7399833855198416, "grad_norm": 9.12737106467952, "learning_rate": 9.607877125278823e-08, "logits/chosen": -2.859375, "logits/rejected": -2.71875, "logps/chosen": -644.0, "logps/rejected": -952.0, "loss": 0.2009, "rewards/accuracies": 0.8500000238418579, "rewards/chosen": -4.96875, "rewards/margins": 3.125, "rewards/rejected": -8.0625, "step": 5790 }, { "epoch": 0.741261422455109, "grad_norm": 10.593757644763386, "learning_rate": 9.520132915529269e-08, "logits/chosen": -2.796875, "logits/rejected": -2.59375, "logps/chosen": -612.0, "logps/rejected": -1000.0, "loss": 0.2101, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.59375, "rewards/margins": 3.84375, "rewards/rejected": -8.4375, "step": 5800 }, { "epoch": 0.7425394593903764, "grad_norm": 9.210720279818025, "learning_rate": 9.432696880892513e-08, "logits/chosen": -2.859375, "logits/rejected": -2.609375, "logps/chosen": -628.0, "logps/rejected": -972.0, "loss": 0.1908, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.625, "rewards/margins": 3.671875, "rewards/rejected": -8.25, "step": 5810 }, { "epoch": 0.7438174963256438, "grad_norm": 7.802302327948419, "learning_rate": 9.345570762055968e-08, "logits/chosen": -2.84375, "logits/rejected": -2.625, "logps/chosen": -640.0, "logps/rejected": -952.0, "loss": 0.2068, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.296875, "rewards/rejected": -8.125, "step": 5820 }, { "epoch": 0.7450955332609113, "grad_norm": 10.07799824095782, "learning_rate": 9.2587562935372e-08, "logits/chosen": -2.890625, "logits/rejected": -2.578125, "logps/chosen": -656.0, "logps/rejected": -972.0, "loss": 0.2027, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.96875, "rewards/margins": 3.375, "rewards/rejected": -8.3125, "step": 5830 }, { "epoch": 0.7463735701961787, "grad_norm": 10.091705801781881, "learning_rate": 9.172255203649429e-08, "logits/chosen": -2.890625, "logits/rejected": -2.5625, "logps/chosen": -656.0, "logps/rejected": -1032.0, "loss": 0.1918, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.9375, "rewards/margins": 3.953125, "rewards/rejected": -8.875, "step": 5840 }, { "epoch": 0.7476516071314461, "grad_norm": 8.580309136599698, "learning_rate": 9.086069214467062e-08, "logits/chosen": -3.0, "logits/rejected": -2.71875, "logps/chosen": -672.0, "logps/rejected": -1064.0, "loss": 0.2003, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.125, "rewards/margins": 4.03125, "rewards/rejected": -9.125, "step": 5850 }, { "epoch": 0.7489296440667135, "grad_norm": 8.566545143121003, "learning_rate": 9.00020004179145e-08, "logits/chosen": -2.953125, "logits/rejected": -2.609375, "logps/chosen": -608.0, "logps/rejected": -1016.0, "loss": 0.1871, "rewards/accuracies": 0.90625, "rewards/chosen": -4.625, "rewards/margins": 4.0, "rewards/rejected": -8.625, "step": 5860 }, { "epoch": 0.750207681001981, "grad_norm": 12.927074227474392, "learning_rate": 8.914649395116752e-08, "logits/chosen": -2.921875, "logits/rejected": -2.703125, "logps/chosen": -672.0, "logps/rejected": -1008.0, "loss": 0.2015, "rewards/accuracies": 0.875, "rewards/chosen": -5.15625, "rewards/margins": 3.625, "rewards/rejected": -8.75, "step": 5870 }, { "epoch": 0.7514857179372484, "grad_norm": 10.661616736265797, "learning_rate": 8.829418977595815e-08, "logits/chosen": -2.96875, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -1008.0, "loss": 0.2289, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.75, "rewards/margins": 3.78125, "rewards/rejected": -8.5625, "step": 5880 }, { "epoch": 0.7527637548725158, "grad_norm": 10.401588278458307, "learning_rate": 8.744510486006376e-08, "logits/chosen": -3.0625, "logits/rejected": -2.78125, "logps/chosen": -648.0, "logps/rejected": -1008.0, "loss": 0.1992, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.78125, "rewards/margins": 3.59375, "rewards/rejected": -8.375, "step": 5890 }, { "epoch": 0.7540417918077832, "grad_norm": 8.981678643474089, "learning_rate": 8.659925610717208e-08, "logits/chosen": -2.984375, "logits/rejected": -2.71875, "logps/chosen": -636.0, "logps/rejected": -988.0, "loss": 0.2206, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.84375, "rewards/margins": 3.578125, "rewards/rejected": -8.4375, "step": 5900 }, { "epoch": 0.7553198287430507, "grad_norm": 9.88446374371275, "learning_rate": 8.575666035654482e-08, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -660.0, "logps/rejected": -992.0, "loss": 0.2016, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.03125, "rewards/margins": 3.59375, "rewards/rejected": -8.625, "step": 5910 }, { "epoch": 0.7565978656783181, "grad_norm": 8.026599552996204, "learning_rate": 8.491733438268281e-08, "logits/chosen": -3.09375, "logits/rejected": -2.765625, "logps/chosen": -652.0, "logps/rejected": -972.0, "loss": 0.1923, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.546875, "rewards/rejected": -8.375, "step": 5920 }, { "epoch": 0.7578759026135855, "grad_norm": 11.0399148587071, "learning_rate": 8.408129489499147e-08, "logits/chosen": -3.046875, "logits/rejected": -2.59375, "logps/chosen": -668.0, "logps/rejected": -1004.0, "loss": 0.2214, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.03125, "rewards/margins": 3.5625, "rewards/rejected": -8.5625, "step": 5930 }, { "epoch": 0.7591539395488529, "grad_norm": 9.850294858942783, "learning_rate": 8.324855853744858e-08, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -660.0, "logps/rejected": -1032.0, "loss": 0.1793, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.84375, "rewards/margins": 4.03125, "rewards/rejected": -8.875, "step": 5940 }, { "epoch": 0.7604319764841204, "grad_norm": 8.702468926444062, "learning_rate": 8.241914188827278e-08, "logits/chosen": -2.953125, "logits/rejected": -2.6875, "logps/chosen": -656.0, "logps/rejected": -1048.0, "loss": 0.1857, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.71875, "rewards/margins": 4.03125, "rewards/rejected": -8.75, "step": 5950 }, { "epoch": 0.7617100134193878, "grad_norm": 13.512408481784156, "learning_rate": 8.159306145959346e-08, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -628.0, "logps/rejected": -1032.0, "loss": 0.1995, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.953125, "rewards/rejected": -8.8125, "step": 5960 }, { "epoch": 0.7629880503546552, "grad_norm": 13.572590071392261, "learning_rate": 8.077033369712233e-08, "logits/chosen": -2.953125, "logits/rejected": -2.796875, "logps/chosen": -648.0, "logps/rejected": -1004.0, "loss": 0.207, "rewards/accuracies": 0.9375, "rewards/chosen": -4.875, "rewards/margins": 3.90625, "rewards/rejected": -8.75, "step": 5970 }, { "epoch": 0.7642660872899226, "grad_norm": 10.395036087264575, "learning_rate": 7.99509749798255e-08, "logits/chosen": -2.984375, "logits/rejected": -2.65625, "logps/chosen": -668.0, "logps/rejected": -928.0, "loss": 0.1916, "rewards/accuracies": 0.824999988079071, "rewards/chosen": -4.875, "rewards/margins": 3.0, "rewards/rejected": -7.875, "step": 5980 }, { "epoch": 0.7655441242251901, "grad_norm": 11.431047760742594, "learning_rate": 7.913500161959808e-08, "logits/chosen": -2.96875, "logits/rejected": -2.6875, "logps/chosen": -620.0, "logps/rejected": -1032.0, "loss": 0.1859, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.5625, "rewards/margins": 4.34375, "rewards/rejected": -8.9375, "step": 5990 }, { "epoch": 0.7668221611604575, "grad_norm": 12.12350658078121, "learning_rate": 7.832242986093877e-08, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -680.0, "logps/rejected": -1020.0, "loss": 0.1896, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -5.25, "rewards/margins": 3.515625, "rewards/rejected": -8.75, "step": 6000 }, { "epoch": 0.7681001980957249, "grad_norm": 12.154386070056951, "learning_rate": 7.751327588062686e-08, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -640.0, "logps/rejected": -988.0, "loss": 0.1784, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.75, "rewards/rejected": -8.5, "step": 6010 }, { "epoch": 0.7693782350309923, "grad_norm": 8.9004131803929, "learning_rate": 7.670755578740012e-08, "logits/chosen": -2.984375, "logits/rejected": -2.671875, "logps/chosen": -624.0, "logps/rejected": -1048.0, "loss": 0.1882, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.59375, "rewards/margins": 4.0625, "rewards/rejected": -8.625, "step": 6020 }, { "epoch": 0.7706562719662599, "grad_norm": 13.277525759657642, "learning_rate": 7.590528562163393e-08, "logits/chosen": -2.84375, "logits/rejected": -2.53125, "logps/chosen": -592.0, "logps/rejected": -956.0, "loss": 0.1881, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.3125, "rewards/margins": 3.859375, "rewards/rejected": -8.1875, "step": 6030 }, { "epoch": 0.7719343089015273, "grad_norm": 10.410706613714614, "learning_rate": 7.51064813550224e-08, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -592.0, "logps/rejected": -940.0, "loss": 0.1864, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.40625, "rewards/margins": 3.6875, "rewards/rejected": -8.125, "step": 6040 }, { "epoch": 0.7732123458367947, "grad_norm": 9.515020477677249, "learning_rate": 7.431115889025969e-08, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -632.0, "logps/rejected": -960.0, "loss": 0.1924, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.75, "rewards/margins": 3.53125, "rewards/rejected": -8.25, "step": 6050 }, { "epoch": 0.7744903827720621, "grad_norm": 13.072675219028618, "learning_rate": 7.35193340607242e-08, "logits/chosen": -2.796875, "logits/rejected": -2.546875, "logps/chosen": -640.0, "logps/rejected": -1000.0, "loss": 0.1824, "rewards/accuracies": 0.9375, "rewards/chosen": -4.6875, "rewards/margins": 3.875, "rewards/rejected": -8.5625, "step": 6060 }, { "epoch": 0.7757684197073296, "grad_norm": 12.782269185266843, "learning_rate": 7.27310226301627e-08, "logits/chosen": -2.828125, "logits/rejected": -2.640625, "logps/chosen": -624.0, "logps/rejected": -992.0, "loss": 0.2286, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 6070 }, { "epoch": 0.777046456642597, "grad_norm": 9.418209612849605, "learning_rate": 7.194624029237686e-08, "logits/chosen": -2.828125, "logits/rejected": -2.5, "logps/chosen": -628.0, "logps/rejected": -1008.0, "loss": 0.1923, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.6875, "rewards/rejected": -8.5625, "step": 6080 }, { "epoch": 0.7783244935778644, "grad_norm": 12.973911791232169, "learning_rate": 7.116500267091077e-08, "logits/chosen": -2.875, "logits/rejected": -2.59375, "logps/chosen": -628.0, "logps/rejected": -972.0, "loss": 0.1974, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.6875, "rewards/margins": 3.671875, "rewards/rejected": -8.375, "step": 6090 }, { "epoch": 0.7796025305131318, "grad_norm": 10.172661744116416, "learning_rate": 7.038732531873972e-08, "logits/chosen": -2.90625, "logits/rejected": -2.703125, "logps/chosen": -632.0, "logps/rejected": -992.0, "loss": 0.1876, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.71875, "rewards/rejected": -8.5, "step": 6100 }, { "epoch": 0.7808805674483993, "grad_norm": 11.421418477500511, "learning_rate": 6.961322371796105e-08, "logits/chosen": -2.796875, "logits/rejected": -2.65625, "logps/chosen": -612.0, "logps/rejected": -968.0, "loss": 0.181, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.625, "rewards/margins": 3.65625, "rewards/rejected": -8.3125, "step": 6110 }, { "epoch": 0.7821586043836667, "grad_norm": 11.428398005666034, "learning_rate": 6.884271327948524e-08, "logits/chosen": -3.046875, "logits/rejected": -2.84375, "logps/chosen": -604.0, "logps/rejected": -956.0, "loss": 0.1896, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.46875, "rewards/margins": 3.65625, "rewards/rejected": -8.125, "step": 6120 }, { "epoch": 0.7834366413189341, "grad_norm": 8.869925031073004, "learning_rate": 6.807580934272961e-08, "logits/chosen": -2.90625, "logits/rejected": -2.734375, "logps/chosen": -612.0, "logps/rejected": -980.0, "loss": 0.2008, "rewards/accuracies": 0.9375, "rewards/chosen": -4.46875, "rewards/margins": 3.765625, "rewards/rejected": -8.25, "step": 6130 }, { "epoch": 0.7847146782542015, "grad_norm": 10.839154644975004, "learning_rate": 6.731252717531288e-08, "logits/chosen": -3.046875, "logits/rejected": -2.765625, "logps/chosen": -596.0, "logps/rejected": -968.0, "loss": 0.1885, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.3125, "rewards/margins": 3.875, "rewards/rejected": -8.1875, "step": 6140 }, { "epoch": 0.785992715189469, "grad_norm": 8.44382605953575, "learning_rate": 6.655288197275103e-08, "logits/chosen": -3.046875, "logits/rejected": -2.671875, "logps/chosen": -612.0, "logps/rejected": -1016.0, "loss": 0.1932, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.59375, "rewards/margins": 4.0625, "rewards/rejected": -8.625, "step": 6150 }, { "epoch": 0.7872707521247364, "grad_norm": 11.179183535261652, "learning_rate": 6.579688885815488e-08, "logits/chosen": -3.078125, "logits/rejected": -2.75, "logps/chosen": -612.0, "logps/rejected": -988.0, "loss": 0.1932, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.4375, "rewards/margins": 3.96875, "rewards/rejected": -8.375, "step": 6160 }, { "epoch": 0.7885487890600038, "grad_norm": 11.052997512838385, "learning_rate": 6.504456288192911e-08, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -640.0, "logps/rejected": -988.0, "loss": 0.1824, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.65625, "rewards/margins": 3.75, "rewards/rejected": -8.4375, "step": 6170 }, { "epoch": 0.7898268259952712, "grad_norm": 11.267748953593356, "learning_rate": 6.429591902147238e-08, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -596.0, "logps/rejected": -1012.0, "loss": 0.1826, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.4375, "rewards/margins": 4.03125, "rewards/rejected": -8.5, "step": 6180 }, { "epoch": 0.7911048629305387, "grad_norm": 8.88100812958174, "learning_rate": 6.355097218087968e-08, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -604.0, "logps/rejected": -1000.0, "loss": 0.1853, "rewards/accuracies": 0.9375, "rewards/chosen": -4.46875, "rewards/margins": 4.28125, "rewards/rejected": -8.75, "step": 6190 }, { "epoch": 0.7923828998658061, "grad_norm": 13.932434101787049, "learning_rate": 6.280973719064486e-08, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -640.0, "logps/rejected": -992.0, "loss": 0.209, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.640625, "rewards/rejected": -8.4375, "step": 6200 }, { "epoch": 0.7936609368010735, "grad_norm": 12.849144357607967, "learning_rate": 6.207222880736618e-08, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -644.0, "logps/rejected": -968.0, "loss": 0.2326, "rewards/accuracies": 0.90625, "rewards/chosen": -4.8125, "rewards/margins": 3.53125, "rewards/rejected": -8.375, "step": 6210 }, { "epoch": 0.7949389737363409, "grad_norm": 13.733842838610288, "learning_rate": 6.133846171345189e-08, "logits/chosen": -2.921875, "logits/rejected": -2.6875, "logps/chosen": -624.0, "logps/rejected": -972.0, "loss": 0.2189, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.65625, "rewards/margins": 3.59375, "rewards/rejected": -8.25, "step": 6220 }, { "epoch": 0.7962170106716084, "grad_norm": 8.758429764508215, "learning_rate": 6.060845051682809e-08, "logits/chosen": -2.921875, "logits/rejected": -2.78125, "logps/chosen": -592.0, "logps/rejected": -928.0, "loss": 0.1997, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.40625, "rewards/margins": 3.484375, "rewards/rejected": -7.875, "step": 6230 }, { "epoch": 0.7974950476068758, "grad_norm": 10.236369223553389, "learning_rate": 5.988220975064839e-08, "logits/chosen": -2.890625, "logits/rejected": -2.671875, "logps/chosen": -632.0, "logps/rejected": -988.0, "loss": 0.1822, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.71875, "rewards/margins": 3.625, "rewards/rejected": -8.375, "step": 6240 }, { "epoch": 0.7987730845421432, "grad_norm": 8.991486561608422, "learning_rate": 5.915975387300382e-08, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -624.0, "logps/rejected": -964.0, "loss": 0.1948, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.71875, "rewards/margins": 3.515625, "rewards/rejected": -8.25, "step": 6250 }, { "epoch": 0.8000511214774106, "grad_norm": 10.298101814631558, "learning_rate": 5.844109726663568e-08, "logits/chosen": -2.96875, "logits/rejected": -2.6875, "logps/chosen": -652.0, "logps/rejected": -976.0, "loss": 0.1978, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.875, "rewards/margins": 3.46875, "rewards/rejected": -8.3125, "step": 6260 }, { "epoch": 0.8013291584126782, "grad_norm": 11.9546027222749, "learning_rate": 5.772625423864866e-08, "logits/chosen": -2.859375, "logits/rejected": -2.65625, "logps/chosen": -616.0, "logps/rejected": -936.0, "loss": 0.2097, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.59375, "rewards/margins": 3.390625, "rewards/rejected": -7.96875, "step": 6270 }, { "epoch": 0.8026071953479456, "grad_norm": 10.119304805105461, "learning_rate": 5.701523902022659e-08, "logits/chosen": -2.875, "logits/rejected": -2.609375, "logps/chosen": -656.0, "logps/rejected": -996.0, "loss": 0.1952, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.5625, "rewards/rejected": -8.375, "step": 6280 }, { "epoch": 0.803885232283213, "grad_norm": 11.201139785327099, "learning_rate": 5.6308065766348335e-08, "logits/chosen": -2.921875, "logits/rejected": -2.578125, "logps/chosen": -632.0, "logps/rejected": -992.0, "loss": 0.2013, "rewards/accuracies": 0.90625, "rewards/chosen": -4.71875, "rewards/margins": 3.75, "rewards/rejected": -8.5, "step": 6290 }, { "epoch": 0.8051632692184804, "grad_norm": 9.377393381822943, "learning_rate": 5.5604748555506706e-08, "logits/chosen": -2.96875, "logits/rejected": -2.578125, "logps/chosen": -600.0, "logps/rejected": -992.0, "loss": 0.193, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.59375, "rewards/margins": 3.9375, "rewards/rejected": -8.5625, "step": 6300 }, { "epoch": 0.8064413061537479, "grad_norm": 8.309075509709407, "learning_rate": 5.490530138942806e-08, "logits/chosen": -2.953125, "logits/rejected": -2.625, "logps/chosen": -628.0, "logps/rejected": -1008.0, "loss": 0.1863, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.78125, "rewards/margins": 3.828125, "rewards/rejected": -8.625, "step": 6310 }, { "epoch": 0.8077193430890153, "grad_norm": 10.253650621317618, "learning_rate": 5.4209738192793146e-08, "logits/chosen": -2.859375, "logits/rejected": -2.625, "logps/chosen": -656.0, "logps/rejected": -1004.0, "loss": 0.1884, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.9375, "rewards/margins": 3.640625, "rewards/rejected": -8.5625, "step": 6320 }, { "epoch": 0.8089973800242827, "grad_norm": 8.25575714700977, "learning_rate": 5.3518072812960444e-08, "logits/chosen": -3.015625, "logits/rejected": -2.6875, "logps/chosen": -636.0, "logps/rejected": -1000.0, "loss": 0.1922, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.75, "rewards/margins": 3.78125, "rewards/rejected": -8.5, "step": 6330 }, { "epoch": 0.8102754169595501, "grad_norm": 11.079645770003177, "learning_rate": 5.283031901969001e-08, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -596.0, "logps/rejected": -968.0, "loss": 0.1864, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.3125, "rewards/margins": 4.0, "rewards/rejected": -8.3125, "step": 6340 }, { "epoch": 0.8115534538948176, "grad_norm": 8.856148371837136, "learning_rate": 5.214649050486961e-08, "logits/chosen": -2.84375, "logits/rejected": -2.71875, "logps/chosen": -632.0, "logps/rejected": -960.0, "loss": 0.184, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.625, "rewards/margins": 3.453125, "rewards/rejected": -8.0625, "step": 6350 }, { "epoch": 0.812831490830085, "grad_norm": 11.181004151997012, "learning_rate": 5.146660088224209e-08, "logits/chosen": -2.96875, "logits/rejected": -2.75, "logps/chosen": -628.0, "logps/rejected": -1008.0, "loss": 0.2045, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.75, "rewards/margins": 3.765625, "rewards/rejected": -8.5, "step": 6360 }, { "epoch": 0.8141095277653524, "grad_norm": 9.153609232514398, "learning_rate": 5.079066368713425e-08, "logits/chosen": -2.96875, "logits/rejected": -2.671875, "logps/chosen": -612.0, "logps/rejected": -992.0, "loss": 0.1839, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.921875, "rewards/rejected": -8.4375, "step": 6370 }, { "epoch": 0.8153875647006198, "grad_norm": 9.535738595895644, "learning_rate": 5.011869237618768e-08, "logits/chosen": -2.875, "logits/rejected": -2.625, "logps/chosen": -596.0, "logps/rejected": -980.0, "loss": 0.1878, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.5, "rewards/margins": 3.765625, "rewards/rejected": -8.25, "step": 6380 }, { "epoch": 0.8166656016358873, "grad_norm": 8.741327828462396, "learning_rate": 4.9450700327090535e-08, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -656.0, "logps/rejected": -980.0, "loss": 0.2076, "rewards/accuracies": 0.90625, "rewards/chosen": -4.8125, "rewards/margins": 3.5625, "rewards/rejected": -8.375, "step": 6390 }, { "epoch": 0.8179436385711547, "grad_norm": 11.426080897454439, "learning_rate": 4.8786700838311184e-08, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -980.0, "loss": 0.2068, "rewards/accuracies": 0.856249988079071, "rewards/chosen": -4.9375, "rewards/margins": 3.296875, "rewards/rejected": -8.25, "step": 6400 }, { "epoch": 0.8192216755064221, "grad_norm": 11.034650603841335, "learning_rate": 4.812670712883396e-08, "logits/chosen": -2.875, "logits/rejected": -2.65625, "logps/chosen": -632.0, "logps/rejected": -988.0, "loss": 0.2023, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.875, "rewards/margins": 3.65625, "rewards/rejected": -8.5, "step": 6410 }, { "epoch": 0.8204997124416895, "grad_norm": 9.254914740173248, "learning_rate": 4.7470732337895335e-08, "logits/chosen": -2.890625, "logits/rejected": -2.625, "logps/chosen": -620.0, "logps/rejected": -992.0, "loss": 0.1828, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -4.59375, "rewards/margins": 3.71875, "rewards/rejected": -8.3125, "step": 6420 }, { "epoch": 0.821777749376957, "grad_norm": 11.445609691444933, "learning_rate": 4.681878952472276e-08, "logits/chosen": -2.875, "logits/rejected": -2.578125, "logps/chosen": -636.0, "logps/rejected": -1024.0, "loss": 0.1843, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.8125, "rewards/margins": 3.84375, "rewards/rejected": -8.625, "step": 6430 }, { "epoch": 0.8230557863122244, "grad_norm": 10.758939904944432, "learning_rate": 4.6170891668274616e-08, "logits/chosen": -2.90625, "logits/rejected": -2.65625, "logps/chosen": -616.0, "logps/rejected": -988.0, "loss": 0.1844, "rewards/accuracies": 0.9375, "rewards/chosen": -4.5625, "rewards/margins": 4.0, "rewards/rejected": -8.5625, "step": 6440 }, { "epoch": 0.8243338232474918, "grad_norm": 6.823878634242757, "learning_rate": 4.552705166698165e-08, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -652.0, "logps/rejected": -968.0, "loss": 0.198, "rewards/accuracies": 0.84375, "rewards/chosen": -5.03125, "rewards/margins": 3.140625, "rewards/rejected": -8.125, "step": 6450 }, { "epoch": 0.8256118601827592, "grad_norm": 11.462876385323876, "learning_rate": 4.488728233849054e-08, "logits/chosen": -2.859375, "logits/rejected": -2.796875, "logps/chosen": -616.0, "logps/rejected": -976.0, "loss": 0.1919, "rewards/accuracies": 0.90625, "rewards/chosen": -4.78125, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 6460 }, { "epoch": 0.8268898971180267, "grad_norm": 11.511738897771243, "learning_rate": 4.425159641940826e-08, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -636.0, "logps/rejected": -1020.0, "loss": 0.1908, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.65625, "rewards/margins": 3.921875, "rewards/rejected": -8.5625, "step": 6470 }, { "epoch": 0.8281679340532941, "grad_norm": 10.012274810790023, "learning_rate": 4.362000656504902e-08, "logits/chosen": -2.890625, "logits/rejected": -2.75, "logps/chosen": -664.0, "logps/rejected": -988.0, "loss": 0.2, "rewards/accuracies": 0.90625, "rewards/chosen": -4.96875, "rewards/margins": 3.625, "rewards/rejected": -8.5625, "step": 6480 }, { "epoch": 0.8294459709885615, "grad_norm": 6.3696111029717715, "learning_rate": 4.299252534918185e-08, "logits/chosen": -2.859375, "logits/rejected": -2.515625, "logps/chosen": -624.0, "logps/rejected": -1072.0, "loss": 0.1921, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.625, "rewards/margins": 4.5, "rewards/rejected": -9.125, "step": 6490 }, { "epoch": 0.8307240079238289, "grad_norm": 8.711459660714853, "learning_rate": 4.2369165263780595e-08, "logits/chosen": -2.890625, "logits/rejected": -2.5625, "logps/chosen": -620.0, "logps/rejected": -1032.0, "loss": 0.1839, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.53125, "rewards/margins": 4.0625, "rewards/rejected": -8.625, "step": 6500 }, { "epoch": 0.8320020448590965, "grad_norm": 8.586531255673448, "learning_rate": 4.174993871877513e-08, "logits/chosen": -2.953125, "logits/rejected": -2.609375, "logps/chosen": -640.0, "logps/rejected": -988.0, "loss": 0.1835, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.546875, "rewards/rejected": -8.375, "step": 6510 }, { "epoch": 0.8332800817943639, "grad_norm": 9.159130220138737, "learning_rate": 4.1134858041804165e-08, "logits/chosen": -3.140625, "logits/rejected": -2.71875, "logps/chosen": -680.0, "logps/rejected": -1024.0, "loss": 0.2108, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.09375, "rewards/margins": 3.640625, "rewards/rejected": -8.75, "step": 6520 }, { "epoch": 0.8345581187296313, "grad_norm": 14.110031544509095, "learning_rate": 4.05239354779702e-08, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -668.0, "logps/rejected": -1056.0, "loss": 0.1964, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.125, "rewards/margins": 3.734375, "rewards/rejected": -8.875, "step": 6530 }, { "epoch": 0.8358361556648987, "grad_norm": 13.647332815864479, "learning_rate": 3.991718318959528e-08, "logits/chosen": -2.828125, "logits/rejected": -2.734375, "logps/chosen": -648.0, "logps/rejected": -988.0, "loss": 0.1952, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.6875, "rewards/rejected": -8.5625, "step": 6540 }, { "epoch": 0.8371141926001662, "grad_norm": 7.729306259968811, "learning_rate": 3.931461325597924e-08, "logits/chosen": -3.0, "logits/rejected": -2.703125, "logps/chosen": -620.0, "logps/rejected": -1000.0, "loss": 0.1847, "rewards/accuracies": 0.90625, "rewards/chosen": -4.71875, "rewards/margins": 3.75, "rewards/rejected": -8.5, "step": 6550 }, { "epoch": 0.8383922295354336, "grad_norm": 12.245685775856511, "learning_rate": 3.871623767315896e-08, "logits/chosen": -2.921875, "logits/rejected": -2.71875, "logps/chosen": -604.0, "logps/rejected": -952.0, "loss": 0.2085, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.5625, "rewards/margins": 3.640625, "rewards/rejected": -8.25, "step": 6560 }, { "epoch": 0.839670266470701, "grad_norm": 8.700079920104228, "learning_rate": 3.812206835366977e-08, "logits/chosen": -3.0625, "logits/rejected": -2.671875, "logps/chosen": -608.0, "logps/rejected": -1000.0, "loss": 0.1891, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.59375, "rewards/margins": 4.125, "rewards/rejected": -8.75, "step": 6570 }, { "epoch": 0.8409483034059684, "grad_norm": 9.331886204678842, "learning_rate": 3.753211712630808e-08, "logits/chosen": -3.03125, "logits/rejected": -2.609375, "logps/chosen": -652.0, "logps/rejected": -980.0, "loss": 0.1941, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.625, "rewards/rejected": -8.4375, "step": 6580 }, { "epoch": 0.8422263403412359, "grad_norm": 9.558487053586102, "learning_rate": 3.6946395735896e-08, "logits/chosen": -2.96875, "logits/rejected": -2.59375, "logps/chosen": -660.0, "logps/rejected": -1004.0, "loss": 0.1877, "rewards/accuracies": 0.875, "rewards/chosen": -4.96875, "rewards/margins": 3.5625, "rewards/rejected": -8.5, "step": 6590 }, { "epoch": 0.8435043772765033, "grad_norm": 9.536711446586605, "learning_rate": 3.636491584304774e-08, "logits/chosen": -2.953125, "logits/rejected": -2.859375, "logps/chosen": -616.0, "logps/rejected": -984.0, "loss": 0.2018, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.5, "rewards/margins": 3.859375, "rewards/rejected": -8.375, "step": 6600 }, { "epoch": 0.8447824142117707, "grad_norm": 9.745832695689701, "learning_rate": 3.578768902393697e-08, "logits/chosen": -2.90625, "logits/rejected": -2.78125, "logps/chosen": -632.0, "logps/rejected": -976.0, "loss": 0.1645, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.53125, "rewards/rejected": -8.3125, "step": 6610 }, { "epoch": 0.8460604511470381, "grad_norm": 6.944861472906955, "learning_rate": 3.521472677006676e-08, "logits/chosen": -2.984375, "logits/rejected": -2.78125, "logps/chosen": -600.0, "logps/rejected": -996.0, "loss": 0.171, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.5625, "rewards/margins": 4.125, "rewards/rejected": -8.6875, "step": 6620 }, { "epoch": 0.8473384880823056, "grad_norm": 6.74838826324146, "learning_rate": 3.464604048804079e-08, "logits/chosen": -3.0, "logits/rejected": -2.796875, "logps/chosen": -636.0, "logps/rejected": -1032.0, "loss": 0.1939, "rewards/accuracies": 0.90625, "rewards/chosen": -4.78125, "rewards/margins": 3.859375, "rewards/rejected": -8.625, "step": 6630 }, { "epoch": 0.848616525017573, "grad_norm": 11.60280140004919, "learning_rate": 3.408164149933587e-08, "logits/chosen": -2.984375, "logits/rejected": -2.828125, "logps/chosen": -608.0, "logps/rejected": -968.0, "loss": 0.1966, "rewards/accuracies": 0.90625, "rewards/chosen": -4.53125, "rewards/margins": 3.859375, "rewards/rejected": -8.375, "step": 6640 }, { "epoch": 0.8498945619528404, "grad_norm": 8.055840883849848, "learning_rate": 3.352154104007721e-08, "logits/chosen": -2.953125, "logits/rejected": -2.734375, "logps/chosen": -612.0, "logps/rejected": -996.0, "loss": 0.187, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.46875, "rewards/margins": 4.09375, "rewards/rejected": -8.5625, "step": 6650 }, { "epoch": 0.8511725988881078, "grad_norm": 10.181553904665757, "learning_rate": 3.296575026081408e-08, "logits/chosen": -2.90625, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -968.0, "loss": 0.1939, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.390625, "rewards/rejected": -8.1875, "step": 6660 }, { "epoch": 0.8524506358233753, "grad_norm": 10.249516867924106, "learning_rate": 3.241428022629822e-08, "logits/chosen": -2.875, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -960.0, "loss": 0.2086, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.40625, "rewards/rejected": -8.1875, "step": 6670 }, { "epoch": 0.8537286727586427, "grad_norm": 12.054480356165014, "learning_rate": 3.186714191526346e-08, "logits/chosen": -2.953125, "logits/rejected": -2.703125, "logps/chosen": -616.0, "logps/rejected": -1004.0, "loss": 0.1844, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.5625, "rewards/margins": 3.90625, "rewards/rejected": -8.5, "step": 6680 }, { "epoch": 0.8550067096939101, "grad_norm": 6.844044080742228, "learning_rate": 3.132434622020716e-08, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -624.0, "logps/rejected": -928.0, "loss": 0.1877, "rewards/accuracies": 0.90625, "rewards/chosen": -4.6875, "rewards/margins": 3.34375, "rewards/rejected": -8.0, "step": 6690 }, { "epoch": 0.8562847466291775, "grad_norm": 11.715109071014988, "learning_rate": 3.078590394717323e-08, "logits/chosen": -2.96875, "logits/rejected": -2.84375, "logps/chosen": -636.0, "logps/rejected": -952.0, "loss": 0.1969, "rewards/accuracies": 0.9375, "rewards/chosen": -4.6875, "rewards/margins": 3.53125, "rewards/rejected": -8.25, "step": 6700 }, { "epoch": 0.857562783564445, "grad_norm": 9.302220076847602, "learning_rate": 3.0251825815537204e-08, "logits/chosen": -3.09375, "logits/rejected": -2.8125, "logps/chosen": -628.0, "logps/rejected": -988.0, "loss": 0.1986, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.6875, "rewards/rejected": -8.5, "step": 6710 }, { "epoch": 0.8588408204997124, "grad_norm": 10.93280433427468, "learning_rate": 2.972212245779271e-08, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -644.0, "logps/rejected": -988.0, "loss": 0.2024, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.5625, "rewards/rejected": -8.5, "step": 6720 }, { "epoch": 0.8601188574349798, "grad_norm": 8.427937780048167, "learning_rate": 2.9196804419339917e-08, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -644.0, "logps/rejected": -1016.0, "loss": 0.1962, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.875, "rewards/margins": 3.71875, "rewards/rejected": -8.5625, "step": 6730 }, { "epoch": 0.8613968943702472, "grad_norm": 8.597293479711446, "learning_rate": 2.867588215827538e-08, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -648.0, "logps/rejected": -1004.0, "loss": 0.2005, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.9375, "rewards/margins": 3.75, "rewards/rejected": -8.6875, "step": 6740 }, { "epoch": 0.8626749313055148, "grad_norm": 9.995814473290233, "learning_rate": 2.8159366045184124e-08, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -612.0, "logps/rejected": -960.0, "loss": 0.1853, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.75, "rewards/margins": 3.65625, "rewards/rejected": -8.375, "step": 6750 }, { "epoch": 0.8639529682407822, "grad_norm": 9.928836544697068, "learning_rate": 2.7647266362932936e-08, "logits/chosen": -2.953125, "logits/rejected": -2.8125, "logps/chosen": -640.0, "logps/rejected": -1032.0, "loss": 0.1877, "rewards/accuracies": 0.90625, "rewards/chosen": -4.6875, "rewards/margins": 4.25, "rewards/rejected": -8.9375, "step": 6760 }, { "epoch": 0.8652310051760496, "grad_norm": 10.113163869171162, "learning_rate": 2.7139593306465758e-08, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -668.0, "logps/rejected": -1032.0, "loss": 0.1788, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.03125, "rewards/margins": 3.796875, "rewards/rejected": -8.8125, "step": 6770 }, { "epoch": 0.866509042111317, "grad_norm": 10.346262544298897, "learning_rate": 2.6636356982600695e-08, "logits/chosen": -3.0, "logits/rejected": -2.71875, "logps/chosen": -644.0, "logps/rejected": -996.0, "loss": 0.1902, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.90625, "rewards/margins": 3.6875, "rewards/rejected": -8.5625, "step": 6780 }, { "epoch": 0.8677870790465845, "grad_norm": 8.842070789217034, "learning_rate": 2.6137567409828864e-08, "logits/chosen": -2.921875, "logits/rejected": -2.84375, "logps/chosen": -644.0, "logps/rejected": -984.0, "loss": 0.1692, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.703125, "rewards/rejected": -8.5625, "step": 6790 }, { "epoch": 0.8690651159818519, "grad_norm": 10.348672126261382, "learning_rate": 2.5643234518114955e-08, "logits/chosen": -2.9375, "logits/rejected": -2.671875, "logps/chosen": -632.0, "logps/rejected": -992.0, "loss": 0.1875, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.734375, "rewards/rejected": -8.5625, "step": 6800 }, { "epoch": 0.8703431529171193, "grad_norm": 14.107914181886756, "learning_rate": 2.515336814869942e-08, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -668.0, "logps/rejected": -1040.0, "loss": 0.1811, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.09375, "rewards/margins": 3.875, "rewards/rejected": -8.9375, "step": 6810 }, { "epoch": 0.8716211898523867, "grad_norm": 10.208150894155457, "learning_rate": 2.466797805390272e-08, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -616.0, "logps/rejected": -960.0, "loss": 0.1868, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.75, "rewards/margins": 3.53125, "rewards/rejected": -8.25, "step": 6820 }, { "epoch": 0.8728992267876542, "grad_norm": 9.143576947306267, "learning_rate": 2.4187073896930977e-08, "logits/chosen": -3.140625, "logits/rejected": -2.734375, "logps/chosen": -632.0, "logps/rejected": -1040.0, "loss": 0.1748, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 4.1875, "rewards/rejected": -8.9375, "step": 6830 }, { "epoch": 0.8741772637229216, "grad_norm": 9.619393237764903, "learning_rate": 2.3710665251683775e-08, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -652.0, "logps/rejected": -1000.0, "loss": 0.1968, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.578125, "rewards/rejected": -8.5, "step": 6840 }, { "epoch": 0.875455300658189, "grad_norm": 12.665374011457072, "learning_rate": 2.3238761602563496e-08, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -640.0, "logps/rejected": -1032.0, "loss": 0.1823, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 4.15625, "rewards/rejected": -8.9375, "step": 6850 }, { "epoch": 0.8767333375934564, "grad_norm": 8.602114064258634, "learning_rate": 2.2771372344286417e-08, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -636.0, "logps/rejected": -976.0, "loss": 0.2042, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.546875, "rewards/rejected": -8.375, "step": 6860 }, { "epoch": 0.8780113745287239, "grad_norm": 6.8825535391482715, "learning_rate": 2.2308506781695906e-08, "logits/chosen": -2.796875, "logits/rejected": -2.671875, "logps/chosen": -628.0, "logps/rejected": -976.0, "loss": 0.2029, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.625, "rewards/margins": 3.625, "rewards/rejected": -8.25, "step": 6870 }, { "epoch": 0.8792894114639913, "grad_norm": 11.099108660135526, "learning_rate": 2.1850174129576933e-08, "logits/chosen": -2.8125, "logits/rejected": -2.6875, "logps/chosen": -640.0, "logps/rejected": -992.0, "loss": 0.2016, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.484375, "rewards/rejected": -8.375, "step": 6880 }, { "epoch": 0.8805674483992587, "grad_norm": 12.681665287086265, "learning_rate": 2.1396383512472722e-08, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -616.0, "logps/rejected": -996.0, "loss": 0.1715, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.625, "rewards/margins": 3.859375, "rewards/rejected": -8.5, "step": 6890 }, { "epoch": 0.8818454853345261, "grad_norm": 10.890987789058547, "learning_rate": 2.09471439645032e-08, "logits/chosen": -2.9375, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -1064.0, "loss": 0.2022, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.6875, "rewards/margins": 4.375, "rewards/rejected": -9.0625, "step": 6900 }, { "epoch": 0.8831235222697936, "grad_norm": 9.926453846449595, "learning_rate": 2.0502464429184933e-08, "logits/chosen": -2.96875, "logits/rejected": -2.71875, "logps/chosen": -632.0, "logps/rejected": -1016.0, "loss": 0.1624, "rewards/accuracies": 0.9375, "rewards/chosen": -4.65625, "rewards/margins": 3.859375, "rewards/rejected": -8.5, "step": 6910 }, { "epoch": 0.884401559205061, "grad_norm": 11.278826880908559, "learning_rate": 2.006235375925322e-08, "logits/chosen": -2.9375, "logits/rejected": -2.765625, "logps/chosen": -648.0, "logps/rejected": -988.0, "loss": 0.1749, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.0, "rewards/margins": 3.59375, "rewards/rejected": -8.5625, "step": 6920 }, { "epoch": 0.8856795961403284, "grad_norm": 11.038211397843158, "learning_rate": 1.962682071648586e-08, "logits/chosen": -2.90625, "logits/rejected": -2.640625, "logps/chosen": -628.0, "logps/rejected": -1000.0, "loss": 0.2006, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.71875, "rewards/margins": 3.890625, "rewards/rejected": -8.625, "step": 6930 }, { "epoch": 0.8869576330755958, "grad_norm": 10.785169033376835, "learning_rate": 1.919587397152861e-08, "logits/chosen": -2.921875, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -1012.0, "loss": 0.1896, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.65625, "rewards/margins": 3.953125, "rewards/rejected": -8.625, "step": 6940 }, { "epoch": 0.8882356700108633, "grad_norm": 9.925891527391922, "learning_rate": 1.8769522103722763e-08, "logits/chosen": -2.96875, "logits/rejected": -2.5625, "logps/chosen": -656.0, "logps/rejected": -1048.0, "loss": 0.1796, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.0, "rewards/margins": 3.796875, "rewards/rejected": -8.8125, "step": 6950 }, { "epoch": 0.8895137069461307, "grad_norm": 8.771182460846884, "learning_rate": 1.8347773600934063e-08, "logits/chosen": -2.96875, "logits/rejected": -2.78125, "logps/chosen": -612.0, "logps/rejected": -1008.0, "loss": 0.1865, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.53125, "rewards/margins": 4.1875, "rewards/rejected": -8.75, "step": 6960 }, { "epoch": 0.8907917438813981, "grad_norm": 9.61805627828746, "learning_rate": 1.7930636859384064e-08, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -612.0, "logps/rejected": -972.0, "loss": 0.2153, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.625, "rewards/margins": 3.640625, "rewards/rejected": -8.25, "step": 6970 }, { "epoch": 0.8920697808166655, "grad_norm": 10.059738064893434, "learning_rate": 1.7518120183482755e-08, "logits/chosen": -2.90625, "logits/rejected": -2.84375, "logps/chosen": -644.0, "logps/rejected": -996.0, "loss": 0.1975, "rewards/accuracies": 0.9375, "rewards/chosen": -4.65625, "rewards/margins": 3.9375, "rewards/rejected": -8.5625, "step": 6980 }, { "epoch": 0.8933478177519331, "grad_norm": 13.222656919567447, "learning_rate": 1.7110231785663094e-08, "logits/chosen": -2.875, "logits/rejected": -2.703125, "logps/chosen": -636.0, "logps/rejected": -1064.0, "loss": 0.2009, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 4.21875, "rewards/rejected": -9.0, "step": 6990 }, { "epoch": 0.8946258546872005, "grad_norm": 11.481242847946271, "learning_rate": 1.670697978621802e-08, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -624.0, "logps/rejected": -1004.0, "loss": 0.1726, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.625, "rewards/margins": 4.0, "rewards/rejected": -8.625, "step": 7000 }, { "epoch": 0.8959038916224679, "grad_norm": 12.332320196979492, "learning_rate": 1.6308372213138244e-08, "logits/chosen": -2.78125, "logits/rejected": -2.625, "logps/chosen": -648.0, "logps/rejected": -1016.0, "loss": 0.1755, "rewards/accuracies": 0.9375, "rewards/chosen": -4.875, "rewards/margins": 3.796875, "rewards/rejected": -8.6875, "step": 7010 }, { "epoch": 0.8971819285577353, "grad_norm": 12.808012844296925, "learning_rate": 1.5914417001952785e-08, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -660.0, "logps/rejected": -1016.0, "loss": 0.2049, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.09375, "rewards/margins": 3.703125, "rewards/rejected": -8.8125, "step": 7020 }, { "epoch": 0.8984599654930028, "grad_norm": 7.820818021151389, "learning_rate": 1.5525121995570823e-08, "logits/chosen": -3.03125, "logits/rejected": -2.828125, "logps/chosen": -664.0, "logps/rejected": -1032.0, "loss": 0.1914, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.78125, "rewards/margins": 3.859375, "rewards/rejected": -8.625, "step": 7030 }, { "epoch": 0.8997380024282702, "grad_norm": 9.370835431954198, "learning_rate": 1.5140494944125588e-08, "logits/chosen": -2.890625, "logits/rejected": -2.625, "logps/chosen": -644.0, "logps/rejected": -1008.0, "loss": 0.158, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.90625, "rewards/margins": 3.765625, "rewards/rejected": -8.6875, "step": 7040 }, { "epoch": 0.9010160393635376, "grad_norm": 11.067566635775739, "learning_rate": 1.4760543504820088e-08, "logits/chosen": -2.953125, "logits/rejected": -2.640625, "logps/chosen": -660.0, "logps/rejected": -1048.0, "loss": 0.1862, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.96875, "rewards/margins": 4.09375, "rewards/rejected": -9.0625, "step": 7050 }, { "epoch": 0.902294076298805, "grad_norm": 9.623901803980623, "learning_rate": 1.438527524177463e-08, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -632.0, "logps/rejected": -1000.0, "loss": 0.1872, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.71875, "rewards/margins": 3.765625, "rewards/rejected": -8.5, "step": 7060 }, { "epoch": 0.9035721132340725, "grad_norm": 11.5687168737442, "learning_rate": 1.4014697625876426e-08, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -664.0, "logps/rejected": -1012.0, "loss": 0.1964, "rewards/accuracies": 0.90625, "rewards/chosen": -5.0625, "rewards/margins": 3.546875, "rewards/rejected": -8.625, "step": 7070 }, { "epoch": 0.9048501501693399, "grad_norm": 8.21021164784677, "learning_rate": 1.3648818034630499e-08, "logits/chosen": -2.890625, "logits/rejected": -2.765625, "logps/chosen": -624.0, "logps/rejected": -972.0, "loss": 0.2012, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.8125, "rewards/margins": 3.625, "rewards/rejected": -8.4375, "step": 7080 }, { "epoch": 0.9061281871046073, "grad_norm": 11.17367745513911, "learning_rate": 1.3287643752013162e-08, "logits/chosen": -2.859375, "logits/rejected": -2.671875, "logps/chosen": -648.0, "logps/rejected": -1000.0, "loss": 0.2041, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.0, "rewards/margins": 3.609375, "rewards/rejected": -8.625, "step": 7090 }, { "epoch": 0.9074062240398747, "grad_norm": 9.879895150119761, "learning_rate": 1.2931181968326765e-08, "logits/chosen": -2.953125, "logits/rejected": -2.703125, "logps/chosen": -628.0, "logps/rejected": -1020.0, "loss": 0.2031, "rewards/accuracies": 0.9375, "rewards/chosen": -4.78125, "rewards/margins": 3.921875, "rewards/rejected": -8.6875, "step": 7100 }, { "epoch": 0.9086842609751422, "grad_norm": 8.750603926836277, "learning_rate": 1.257943978005674e-08, "logits/chosen": -2.875, "logits/rejected": -2.703125, "logps/chosen": -640.0, "logps/rejected": -972.0, "loss": 0.2044, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.546875, "rewards/rejected": -8.4375, "step": 7110 }, { "epoch": 0.9099622979104096, "grad_norm": 9.732574600229979, "learning_rate": 1.2232424189730084e-08, "logits/chosen": -2.875, "logits/rejected": -2.703125, "logps/chosen": -644.0, "logps/rejected": -992.0, "loss": 0.1951, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.875, "rewards/margins": 3.609375, "rewards/rejected": -8.5, "step": 7120 }, { "epoch": 0.911240334845677, "grad_norm": 10.799063558489175, "learning_rate": 1.189014210577624e-08, "logits/chosen": -2.765625, "logits/rejected": -2.625, "logps/chosen": -628.0, "logps/rejected": -952.0, "loss": 0.1805, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.421875, "rewards/rejected": -8.1875, "step": 7130 }, { "epoch": 0.9125183717809444, "grad_norm": 8.342418997195931, "learning_rate": 1.1552600342389363e-08, "logits/chosen": -2.875, "logits/rejected": -2.59375, "logps/chosen": -664.0, "logps/rejected": -1032.0, "loss": 0.172, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.9375, "rewards/margins": 3.703125, "rewards/rejected": -8.625, "step": 7140 }, { "epoch": 0.9137964087162119, "grad_norm": 10.71141399194098, "learning_rate": 1.121980561939273e-08, "logits/chosen": -2.921875, "logits/rejected": -2.78125, "logps/chosen": -656.0, "logps/rejected": -1040.0, "loss": 0.1729, "rewards/accuracies": 0.9375, "rewards/chosen": -5.03125, "rewards/margins": 3.8125, "rewards/rejected": -8.8125, "step": 7150 }, { "epoch": 0.9150744456514793, "grad_norm": 9.405664113940032, "learning_rate": 1.0891764562104888e-08, "logits/chosen": -2.9375, "logits/rejected": -2.625, "logps/chosen": -656.0, "logps/rejected": -1024.0, "loss": 0.1956, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.0, "rewards/margins": 3.6875, "rewards/rejected": -8.6875, "step": 7160 }, { "epoch": 0.9163524825867467, "grad_norm": 11.026789747690092, "learning_rate": 1.0568483701207941e-08, "logits/chosen": -2.921875, "logits/rejected": -2.640625, "logps/chosen": -648.0, "logps/rejected": -972.0, "loss": 0.2024, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.9375, "rewards/margins": 3.515625, "rewards/rejected": -8.4375, "step": 7170 }, { "epoch": 0.9176305195220141, "grad_norm": 9.18756797117969, "learning_rate": 1.0249969472617304e-08, "logits/chosen": -2.84375, "logits/rejected": -2.75, "logps/chosen": -624.0, "logps/rejected": -976.0, "loss": 0.1915, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.75, "rewards/margins": 3.765625, "rewards/rejected": -8.5, "step": 7180 }, { "epoch": 0.9189085564572816, "grad_norm": 10.972057226223303, "learning_rate": 9.936228217353709e-09, "logits/chosen": -2.890625, "logits/rejected": -2.671875, "logps/chosen": -632.0, "logps/rejected": -952.0, "loss": 0.1814, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.75, "rewards/margins": 3.453125, "rewards/rejected": -8.1875, "step": 7190 }, { "epoch": 0.920186593392549, "grad_norm": 7.87780338280655, "learning_rate": 9.627266181416987e-09, "logits/chosen": -2.96875, "logits/rejected": -2.625, "logps/chosen": -636.0, "logps/rejected": -1032.0, "loss": 0.1828, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.8125, "rewards/margins": 4.0, "rewards/rejected": -8.8125, "step": 7200 }, { "epoch": 0.9214646303278164, "grad_norm": 7.847069398644123, "learning_rate": 9.323089515661603e-09, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -648.0, "logps/rejected": -988.0, "loss": 0.1631, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.8125, "rewards/margins": 3.796875, "rewards/rejected": -8.625, "step": 7210 }, { "epoch": 0.9227426672630838, "grad_norm": 9.819571950781643, "learning_rate": 9.023704275674394e-09, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -672.0, "logps/rejected": -1024.0, "loss": 0.194, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.0, "rewards/margins": 3.96875, "rewards/rejected": -8.9375, "step": 7220 }, { "epoch": 0.9240207041983514, "grad_norm": 11.758652729017486, "learning_rate": 8.729116421653748e-09, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -656.0, "logps/rejected": -1008.0, "loss": 0.1746, "rewards/accuracies": 0.90625, "rewards/chosen": -5.0625, "rewards/margins": 3.65625, "rewards/rejected": -8.75, "step": 7230 }, { "epoch": 0.9252987411336188, "grad_norm": 9.074528747089998, "learning_rate": 8.439331818291261e-09, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -656.0, "logps/rejected": -996.0, "loss": 0.2084, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.0, "rewards/margins": 3.734375, "rewards/rejected": -8.75, "step": 7240 }, { "epoch": 0.9265767780688862, "grad_norm": 10.439935658287324, "learning_rate": 8.154356234654713e-09, "logits/chosen": -2.90625, "logits/rejected": -2.671875, "logps/chosen": -660.0, "logps/rejected": -1024.0, "loss": 0.203, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.0625, "rewards/margins": 3.703125, "rewards/rejected": -8.75, "step": 7250 }, { "epoch": 0.9278548150041536, "grad_norm": 13.20542442648142, "learning_rate": 7.874195344073297e-09, "logits/chosen": -2.90625, "logits/rejected": -2.515625, "logps/chosen": -664.0, "logps/rejected": -1024.0, "loss": 0.1838, "rewards/accuracies": 0.9375, "rewards/chosen": -5.0, "rewards/margins": 3.671875, "rewards/rejected": -8.6875, "step": 7260 }, { "epoch": 0.9291328519394211, "grad_norm": 11.62024702139443, "learning_rate": 7.598854724024828e-09, "logits/chosen": -2.953125, "logits/rejected": -2.65625, "logps/chosen": -644.0, "logps/rejected": -1008.0, "loss": 0.185, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.90625, "rewards/margins": 3.78125, "rewards/rejected": -8.6875, "step": 7270 }, { "epoch": 0.9304108888746885, "grad_norm": 10.099788710806347, "learning_rate": 7.328339856024407e-09, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -624.0, "logps/rejected": -1008.0, "loss": 0.1723, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.6875, "rewards/margins": 3.984375, "rewards/rejected": -8.6875, "step": 7280 }, { "epoch": 0.9316889258099559, "grad_norm": 9.897081640900788, "learning_rate": 7.062656125515654e-09, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -632.0, "logps/rejected": -1024.0, "loss": 0.1759, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.8125, "rewards/margins": 4.125, "rewards/rejected": -8.9375, "step": 7290 }, { "epoch": 0.9329669627452233, "grad_norm": 10.589111836344392, "learning_rate": 6.801808821763178e-09, "logits/chosen": -2.84375, "logits/rejected": -2.59375, "logps/chosen": -632.0, "logps/rejected": -1040.0, "loss": 0.2106, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.6875, "rewards/margins": 4.09375, "rewards/rejected": -8.75, "step": 7300 }, { "epoch": 0.9342449996804908, "grad_norm": 10.724801842725016, "learning_rate": 6.545803137747385e-09, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -632.0, "logps/rejected": -1016.0, "loss": 0.2011, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.84375, "rewards/margins": 3.90625, "rewards/rejected": -8.75, "step": 7310 }, { "epoch": 0.9355230366157582, "grad_norm": 9.611442651274807, "learning_rate": 6.294644170061258e-09, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -636.0, "logps/rejected": -988.0, "loss": 0.1943, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.9375, "rewards/margins": 3.640625, "rewards/rejected": -8.5625, "step": 7320 }, { "epoch": 0.9368010735510256, "grad_norm": 11.751486176205045, "learning_rate": 6.0483369188085995e-09, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -632.0, "logps/rejected": -1012.0, "loss": 0.1801, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.84375, "rewards/rejected": -8.75, "step": 7330 }, { "epoch": 0.938079110486293, "grad_norm": 11.512912312322182, "learning_rate": 5.806886287504725e-09, "logits/chosen": -2.90625, "logits/rejected": -2.625, "logps/chosen": -652.0, "logps/rejected": -1024.0, "loss": 0.1873, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.875, "rewards/margins": 3.703125, "rewards/rejected": -8.625, "step": 7340 }, { "epoch": 0.9393571474215605, "grad_norm": 7.192083175792538, "learning_rate": 5.570297082978709e-09, "logits/chosen": -2.890625, "logits/rejected": -2.65625, "logps/chosen": -620.0, "logps/rejected": -1032.0, "loss": 0.1647, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.5625, "rewards/margins": 4.1875, "rewards/rejected": -8.75, "step": 7350 }, { "epoch": 0.9406351843568279, "grad_norm": 10.843092841652517, "learning_rate": 5.3385740152777634e-09, "logits/chosen": -2.953125, "logits/rejected": -2.6875, "logps/chosen": -676.0, "logps/rejected": -1048.0, "loss": 0.2038, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0, "rewards/margins": 4.0, "rewards/rejected": -9.0, "step": 7360 }, { "epoch": 0.9419132212920953, "grad_norm": 8.467409445214516, "learning_rate": 5.111721697573457e-09, "logits/chosen": -2.890625, "logits/rejected": -2.765625, "logps/chosen": -640.0, "logps/rejected": -1040.0, "loss": 0.1804, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.6875, "rewards/margins": 4.25, "rewards/rejected": -8.9375, "step": 7370 }, { "epoch": 0.9431912582273627, "grad_norm": 6.715800157562806, "learning_rate": 4.889744646069837e-09, "logits/chosen": -2.875, "logits/rejected": -2.78125, "logps/chosen": -644.0, "logps/rejected": -1040.0, "loss": 0.1889, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.03125, "rewards/margins": 3.96875, "rewards/rejected": -9.0, "step": 7380 }, { "epoch": 0.9444692951626302, "grad_norm": 8.400806549616826, "learning_rate": 4.672647279913622e-09, "logits/chosen": -2.921875, "logits/rejected": -2.71875, "logps/chosen": -628.0, "logps/rejected": -1016.0, "loss": 0.1971, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.84375, "rewards/margins": 3.8125, "rewards/rejected": -8.625, "step": 7390 }, { "epoch": 0.9457473320978976, "grad_norm": 13.972001950715883, "learning_rate": 4.4604339211059824e-09, "logits/chosen": -2.90625, "logits/rejected": -2.78125, "logps/chosen": -652.0, "logps/rejected": -1040.0, "loss": 0.1976, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.78125, "rewards/margins": 4.0625, "rewards/rejected": -8.875, "step": 7400 }, { "epoch": 0.947025369033165, "grad_norm": 9.677769647797021, "learning_rate": 4.253108794416871e-09, "logits/chosen": -2.96875, "logits/rejected": -2.71875, "logps/chosen": -640.0, "logps/rejected": -1080.0, "loss": 0.1928, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.78125, "rewards/margins": 4.40625, "rewards/rejected": -9.1875, "step": 7410 }, { "epoch": 0.9483034059684324, "grad_norm": 9.345026975785517, "learning_rate": 4.050676027300554e-09, "logits/chosen": -2.953125, "logits/rejected": -2.640625, "logps/chosen": -680.0, "logps/rejected": -1032.0, "loss": 0.2038, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.21875, "rewards/margins": 3.59375, "rewards/rejected": -8.8125, "step": 7420 }, { "epoch": 0.9495814429037, "grad_norm": 9.695675333701805, "learning_rate": 3.853139649813625e-09, "logits/chosen": -2.859375, "logits/rejected": -2.5, "logps/chosen": -644.0, "logps/rejected": -1064.0, "loss": 0.1957, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.875, "rewards/margins": 4.15625, "rewards/rejected": -9.0625, "step": 7430 }, { "epoch": 0.9508594798389673, "grad_norm": 7.770057515915393, "learning_rate": 3.660503594534875e-09, "logits/chosen": -2.96875, "logits/rejected": -2.671875, "logps/chosen": -652.0, "logps/rejected": -1040.0, "loss": 0.1867, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.96875, "rewards/margins": 3.671875, "rewards/rejected": -8.6875, "step": 7440 }, { "epoch": 0.9521375167742347, "grad_norm": 12.504038666390727, "learning_rate": 3.4727716964866595e-09, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -652.0, "logps/rejected": -996.0, "loss": 0.1886, "rewards/accuracies": 0.90625, "rewards/chosen": -4.875, "rewards/margins": 3.5, "rewards/rejected": -8.375, "step": 7450 }, { "epoch": 0.9534155537095022, "grad_norm": 9.181970318686727, "learning_rate": 3.2899476930589863e-09, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -656.0, "logps/rejected": -992.0, "loss": 0.1931, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.0, "rewards/margins": 3.59375, "rewards/rejected": -8.625, "step": 7460 }, { "epoch": 0.9546935906447697, "grad_norm": 11.183563173203545, "learning_rate": 3.112035223934689e-09, "logits/chosen": -2.921875, "logits/rejected": -2.625, "logps/chosen": -620.0, "logps/rejected": -972.0, "loss": 0.1982, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.84375, "rewards/margins": 3.59375, "rewards/rejected": -8.4375, "step": 7470 }, { "epoch": 0.9559716275800371, "grad_norm": 10.636414438455274, "learning_rate": 2.939037831017288e-09, "logits/chosen": -2.9375, "logits/rejected": -2.765625, "logps/chosen": -636.0, "logps/rejected": -1020.0, "loss": 0.1819, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.8125, "rewards/margins": 3.921875, "rewards/rejected": -8.75, "step": 7480 }, { "epoch": 0.9572496645153045, "grad_norm": 11.633662618889964, "learning_rate": 2.7709589583603254e-09, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -652.0, "logps/rejected": -1032.0, "loss": 0.1904, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.96875, "rewards/margins": 3.953125, "rewards/rejected": -8.9375, "step": 7490 }, { "epoch": 0.9585277014505719, "grad_norm": 12.38372710586855, "learning_rate": 2.6078019520988093e-09, "logits/chosen": -2.875, "logits/rejected": -2.796875, "logps/chosen": -660.0, "logps/rejected": -984.0, "loss": 0.196, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.0625, "rewards/margins": 3.484375, "rewards/rejected": -8.5625, "step": 7500 }, { "epoch": 0.9598057383858394, "grad_norm": 12.644481438195353, "learning_rate": 2.449570060382711e-09, "logits/chosen": -2.765625, "logits/rejected": -2.71875, "logps/chosen": -672.0, "logps/rejected": -1040.0, "loss": 0.186, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0, "rewards/margins": 4.0625, "rewards/rejected": -9.0625, "step": 7510 }, { "epoch": 0.9610837753211068, "grad_norm": 9.432942475563783, "learning_rate": 2.2962664333120994e-09, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -620.0, "logps/rejected": -1008.0, "loss": 0.1963, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.71875, "rewards/margins": 3.96875, "rewards/rejected": -8.6875, "step": 7520 }, { "epoch": 0.9623618122563742, "grad_norm": 11.928069118877714, "learning_rate": 2.1478941228745828e-09, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -656.0, "logps/rejected": -1016.0, "loss": 0.2113, "rewards/accuracies": 0.90625, "rewards/chosen": -5.0, "rewards/margins": 3.59375, "rewards/rejected": -8.625, "step": 7530 }, { "epoch": 0.9636398491916416, "grad_norm": 7.886668745442734, "learning_rate": 2.0044560828845205e-09, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -652.0, "logps/rejected": -1072.0, "loss": 0.186, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.875, "rewards/margins": 4.375, "rewards/rejected": -9.25, "step": 7540 }, { "epoch": 0.9649178861269091, "grad_norm": 9.34734142715555, "learning_rate": 1.8659551689241547e-09, "logits/chosen": -3.0, "logits/rejected": -2.6875, "logps/chosen": -640.0, "logps/rejected": -1016.0, "loss": 0.2107, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.84375, "rewards/rejected": -8.6875, "step": 7550 }, { "epoch": 0.9661959230621765, "grad_norm": 9.611208535032741, "learning_rate": 1.7323941382869078e-09, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -684.0, "logps/rejected": -1008.0, "loss": 0.1923, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.0625, "rewards/margins": 3.71875, "rewards/rejected": -8.75, "step": 7560 }, { "epoch": 0.9674739599974439, "grad_norm": 9.527127269024511, "learning_rate": 1.603775649922312e-09, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -632.0, "logps/rejected": -1020.0, "loss": 0.1666, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.5625, "rewards/margins": 4.15625, "rewards/rejected": -8.75, "step": 7570 }, { "epoch": 0.9687519969327113, "grad_norm": 9.291770961378111, "learning_rate": 1.4801022643831652e-09, "logits/chosen": -2.859375, "logits/rejected": -2.65625, "logps/chosen": -680.0, "logps/rejected": -1024.0, "loss": 0.2, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0625, "rewards/margins": 3.75, "rewards/rejected": -8.8125, "step": 7580 }, { "epoch": 0.9700300338679788, "grad_norm": 12.925961238581618, "learning_rate": 1.3613764437745978e-09, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -640.0, "logps/rejected": -1008.0, "loss": 0.1738, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 3.84375, "rewards/rejected": -8.5625, "step": 7590 }, { "epoch": 0.9713080708032462, "grad_norm": 10.408031687484408, "learning_rate": 1.2476005517049192e-09, "logits/chosen": -2.90625, "logits/rejected": -2.765625, "logps/chosen": -640.0, "logps/rejected": -984.0, "loss": 0.185, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.78125, "rewards/margins": 3.625, "rewards/rejected": -8.4375, "step": 7600 }, { "epoch": 0.9725861077385136, "grad_norm": 9.808605265046817, "learning_rate": 1.138776853238682e-09, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -660.0, "logps/rejected": -1008.0, "loss": 0.1914, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.59375, "rewards/rejected": -8.4375, "step": 7610 }, { "epoch": 0.973864144673781, "grad_norm": 7.618616317950994, "learning_rate": 1.0349075148516074e-09, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -640.0, "logps/rejected": -1004.0, "loss": 0.1606, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.78125, "rewards/rejected": -8.625, "step": 7620 }, { "epoch": 0.9751421816090485, "grad_norm": 11.794438615303235, "learning_rate": 9.359946043873412e-10, "logits/chosen": -2.8125, "logits/rejected": -2.65625, "logps/chosen": -656.0, "logps/rejected": -976.0, "loss": 0.1954, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.75, "rewards/margins": 3.71875, "rewards/rejected": -8.5, "step": 7630 }, { "epoch": 0.9764202185443159, "grad_norm": 12.563234469246394, "learning_rate": 8.420400910163494e-10, "logits/chosen": -3.171875, "logits/rejected": -2.71875, "logps/chosen": -652.0, "logps/rejected": -1024.0, "loss": 0.1973, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.96875, "rewards/margins": 3.703125, "rewards/rejected": -8.625, "step": 7640 }, { "epoch": 0.9776982554795833, "grad_norm": 16.042841291187038, "learning_rate": 7.530458451967814e-10, "logits/chosen": -3.0, "logits/rejected": -2.6875, "logps/chosen": -636.0, "logps/rejected": -1024.0, "loss": 0.1815, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.875, "rewards/margins": 3.890625, "rewards/rejected": -8.75, "step": 7650 }, { "epoch": 0.9789762924148507, "grad_norm": 13.296730357123543, "learning_rate": 6.69013638637056e-10, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -632.0, "logps/rejected": -1020.0, "loss": 0.1816, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.6875, "rewards/margins": 4.125, "rewards/rejected": -8.8125, "step": 7660 }, { "epoch": 0.9802543293501182, "grad_norm": 12.112711580204818, "learning_rate": 5.899451442607784e-10, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -632.0, "logps/rejected": -1040.0, "loss": 0.1893, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.71875, "rewards/margins": 4.125, "rewards/rejected": -8.875, "step": 7670 }, { "epoch": 0.9815323662853856, "grad_norm": 14.646352722458984, "learning_rate": 5.158419361733501e-10, "logits/chosen": -2.9375, "logits/rejected": -2.703125, "logps/chosen": -684.0, "logps/rejected": -1040.0, "loss": 0.1982, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.125, "rewards/margins": 3.734375, "rewards/rejected": -8.875, "step": 7680 }, { "epoch": 0.982810403220653, "grad_norm": 13.260198930878687, "learning_rate": 4.4670548963063327e-10, "logits/chosen": -2.875, "logits/rejected": -2.609375, "logps/chosen": -648.0, "logps/rejected": -992.0, "loss": 0.1835, "rewards/accuracies": 0.90625, "rewards/chosen": -4.9375, "rewards/margins": 3.546875, "rewards/rejected": -8.5, "step": 7690 }, { "epoch": 0.9840884401559205, "grad_norm": 8.788325394613999, "learning_rate": 3.82537181009529e-10, "logits/chosen": -3.015625, "logits/rejected": -2.546875, "logps/chosen": -656.0, "logps/rejected": -1088.0, "loss": 0.1928, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.84375, "rewards/margins": 4.625, "rewards/rejected": -9.4375, "step": 7700 }, { "epoch": 0.985366477091188, "grad_norm": 12.794711894897244, "learning_rate": 3.233382877806945e-10, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -676.0, "logps/rejected": -1032.0, "loss": 0.2011, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.03125, "rewards/margins": 3.765625, "rewards/rejected": -8.8125, "step": 7710 }, { "epoch": 0.9866445140264554, "grad_norm": 11.753555535364036, "learning_rate": 2.6910998848306276e-10, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -656.0, "logps/rejected": -1000.0, "loss": 0.1921, "rewards/accuracies": 0.90625, "rewards/chosen": -4.96875, "rewards/margins": 3.625, "rewards/rejected": -8.625, "step": 7720 }, { "epoch": 0.9879225509617228, "grad_norm": 14.594250427636789, "learning_rate": 2.1985336270033405e-10, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -644.0, "logps/rejected": -1024.0, "loss": 0.1975, "rewards/accuracies": 0.90625, "rewards/chosen": -4.90625, "rewards/margins": 3.953125, "rewards/rejected": -8.875, "step": 7730 }, { "epoch": 0.9892005878969902, "grad_norm": 12.77554394664042, "learning_rate": 1.7556939103952063e-10, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -620.0, "logps/rejected": -956.0, "loss": 0.1936, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.65625, "rewards/margins": 3.59375, "rewards/rejected": -8.25, "step": 7740 }, { "epoch": 0.9904786248322577, "grad_norm": 10.519067298735793, "learning_rate": 1.3625895511137907e-10, "logits/chosen": -2.890625, "logits/rejected": -2.640625, "logps/chosen": -640.0, "logps/rejected": -1004.0, "loss": 0.1752, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.75, "rewards/margins": 3.953125, "rewards/rejected": -8.6875, "step": 7750 }, { "epoch": 0.9917566617675251, "grad_norm": 9.931366042203358, "learning_rate": 1.0192283751297992e-10, "logits/chosen": -3.015625, "logits/rejected": -2.640625, "logps/chosen": -620.0, "logps/rejected": -1000.0, "loss": 0.1821, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.75, "rewards/margins": 3.828125, "rewards/rejected": -8.5625, "step": 7760 }, { "epoch": 0.9930346987027925, "grad_norm": 9.246116106532282, "learning_rate": 7.256172181199783e-11, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -656.0, "logps/rejected": -1032.0, "loss": 0.1969, "rewards/accuracies": 0.90625, "rewards/chosen": -4.9375, "rewards/margins": 4.0, "rewards/rejected": -8.9375, "step": 7770 }, { "epoch": 0.9943127356380599, "grad_norm": 7.26117544128573, "learning_rate": 4.8176192533139206e-11, "logits/chosen": -2.859375, "logits/rejected": -2.765625, "logps/chosen": -632.0, "logps/rejected": -1016.0, "loss": 0.194, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.78125, "rewards/rejected": -8.625, "step": 7780 }, { "epoch": 0.9955907725733274, "grad_norm": 12.332218978053229, "learning_rate": 2.876673514659589e-11, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -628.0, "logps/rejected": -972.0, "loss": 0.1812, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.671875, "rewards/rejected": -8.5, "step": 7790 }, { "epoch": 0.9968688095085948, "grad_norm": 9.768254436897418, "learning_rate": 1.4333736058219636e-11, "logits/chosen": -2.90625, "logits/rejected": -2.65625, "logps/chosen": -648.0, "logps/rejected": -1004.0, "loss": 0.1937, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -4.90625, "rewards/margins": 3.640625, "rewards/rejected": -8.5625, "step": 7800 }, { "epoch": 0.9981468464438622, "grad_norm": 12.753093983260724, "learning_rate": 4.8774826019448715e-12, "logits/chosen": -2.890625, "logits/rejected": -2.640625, "logps/chosen": -656.0, "logps/rejected": -1072.0, "loss": 0.1805, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 4.1875, "rewards/rejected": -8.9375, "step": 7810 }, { "epoch": 0.9994248833791296, "grad_norm": 9.444618426186262, "learning_rate": 3.981630340710307e-13, "logits/chosen": -2.890625, "logits/rejected": -2.640625, "logps/chosen": -632.0, "logps/rejected": -1012.0, "loss": 0.1828, "rewards/accuracies": 0.96875, "rewards/chosen": -4.84375, "rewards/margins": 4.03125, "rewards/rejected": -8.875, "step": 7820 }, { "epoch": 1.0007668221611605, "grad_norm": 9.793739897002116, "learning_rate": 2.93088577631297e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -692.0, "logps/rejected": -1048.0, "loss": 0.1666, "rewards/accuracies": 0.9166666865348816, "rewards/chosen": -5.25, "rewards/margins": 3.8125, "rewards/rejected": -9.0625, "step": 7830 }, { "epoch": 1.002044859096428, "grad_norm": 9.002669484334344, "learning_rate": 2.925391243611856e-07, "logits/chosen": -3.125, "logits/rejected": -2.875, "logps/chosen": -608.0, "logps/rejected": -1032.0, "loss": 0.1877, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.6875, "rewards/margins": 4.1875, "rewards/rejected": -8.875, "step": 7840 }, { "epoch": 1.0033228960316953, "grad_norm": 7.3436812439478585, "learning_rate": 2.9198945940233674e-07, "logits/chosen": -3.015625, "logits/rejected": -2.90625, "logps/chosen": -648.0, "logps/rejected": -992.0, "loss": 0.1676, "rewards/accuracies": 0.9375, "rewards/chosen": -4.96875, "rewards/margins": 3.578125, "rewards/rejected": -8.5, "step": 7850 }, { "epoch": 1.0046009329669627, "grad_norm": 10.110088255433219, "learning_rate": 2.914395854900648e-07, "logits/chosen": -3.09375, "logits/rejected": -2.8125, "logps/chosen": -632.0, "logps/rejected": -1032.0, "loss": 0.1521, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.71875, "rewards/margins": 4.03125, "rewards/rejected": -8.75, "step": 7860 }, { "epoch": 1.0058789699022301, "grad_norm": 10.536172769097313, "learning_rate": 2.908895053607241e-07, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -672.0, "logps/rejected": -1136.0, "loss": 0.1689, "rewards/accuracies": 0.9375, "rewards/chosen": -4.90625, "rewards/margins": 4.84375, "rewards/rejected": -9.75, "step": 7870 }, { "epoch": 1.0071570068374975, "grad_norm": 11.272134541218382, "learning_rate": 2.903392217516952e-07, "logits/chosen": -2.953125, "logits/rejected": -2.8125, "logps/chosen": -660.0, "logps/rejected": -1032.0, "loss": 0.1912, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.90625, "rewards/margins": 3.828125, "rewards/rejected": -8.75, "step": 7880 }, { "epoch": 1.008435043772765, "grad_norm": 10.96979479582042, "learning_rate": 2.89788737401371e-07, "logits/chosen": -2.796875, "logits/rejected": -2.640625, "logps/chosen": -632.0, "logps/rejected": -1012.0, "loss": 0.1673, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.75, "rewards/margins": 4.03125, "rewards/rejected": -8.75, "step": 7890 }, { "epoch": 1.0097130807080326, "grad_norm": 9.981606177601087, "learning_rate": 2.8923805504914363e-07, "logits/chosen": -2.96875, "logits/rejected": -2.65625, "logps/chosen": -596.0, "logps/rejected": -1000.0, "loss": 0.1789, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.5, "rewards/margins": 4.0, "rewards/rejected": -8.5, "step": 7900 }, { "epoch": 1.0109911176433, "grad_norm": 11.993598319058572, "learning_rate": 2.886871774353903e-07, "logits/chosen": -2.921875, "logits/rejected": -2.828125, "logps/chosen": -624.0, "logps/rejected": -1000.0, "loss": 0.182, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.75, "rewards/margins": 3.890625, "rewards/rejected": -8.625, "step": 7910 }, { "epoch": 1.0122691545785674, "grad_norm": 13.605038367767165, "learning_rate": 2.8813610730146003e-07, "logits/chosen": -3.0625, "logits/rejected": -2.953125, "logps/chosen": -668.0, "logps/rejected": -1048.0, "loss": 0.2072, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.96875, "rewards/margins": 4.0625, "rewards/rejected": -9.0, "step": 7920 }, { "epoch": 1.0135471915138348, "grad_norm": 11.172416814175007, "learning_rate": 2.8758484738966e-07, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -568.0, "logps/rejected": -904.0, "loss": 0.1721, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.03125, "rewards/margins": 3.6875, "rewards/rejected": -7.6875, "step": 7930 }, { "epoch": 1.0148252284491022, "grad_norm": 9.814062210000673, "learning_rate": 2.870334004432416e-07, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -636.0, "logps/rejected": -996.0, "loss": 0.1868, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.71875, "rewards/margins": 3.859375, "rewards/rejected": -8.5625, "step": 7940 }, { "epoch": 1.0161032653843696, "grad_norm": 13.3322442728074, "learning_rate": 2.86481769206387e-07, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -592.0, "logps/rejected": -960.0, "loss": 0.1819, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.375, "rewards/margins": 3.828125, "rewards/rejected": -8.25, "step": 7950 }, { "epoch": 1.017381302319637, "grad_norm": 8.583508636588496, "learning_rate": 2.859299564241955e-07, "logits/chosen": -2.984375, "logits/rejected": -2.90625, "logps/chosen": -600.0, "logps/rejected": -920.0, "loss": 0.1972, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -4.4375, "rewards/margins": 3.296875, "rewards/rejected": -7.75, "step": 7960 }, { "epoch": 1.0186593392549044, "grad_norm": 9.09390192789269, "learning_rate": 2.853779648426697e-07, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -572.0, "logps/rejected": -984.0, "loss": 0.1918, "rewards/accuracies": 0.9375, "rewards/chosen": -4.1875, "rewards/margins": 4.0, "rewards/rejected": -8.1875, "step": 7970 }, { "epoch": 1.019937376190172, "grad_norm": 9.317681243705513, "learning_rate": 2.8482579720870234e-07, "logits/chosen": -2.984375, "logits/rejected": -2.6875, "logps/chosen": -636.0, "logps/rejected": -1024.0, "loss": 0.1859, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.9375, "rewards/rejected": -8.8125, "step": 7980 }, { "epoch": 1.0212154131254394, "grad_norm": 9.655269321006157, "learning_rate": 2.842734562700618e-07, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -612.0, "logps/rejected": -1040.0, "loss": 0.184, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.53125, "rewards/margins": 4.4375, "rewards/rejected": -9.0, "step": 7990 }, { "epoch": 1.0224934500607068, "grad_norm": 9.513634678774741, "learning_rate": 2.837209447753793e-07, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -572.0, "logps/rejected": -960.0, "loss": 0.1792, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -3.9375, "rewards/margins": 4.15625, "rewards/rejected": -8.125, "step": 8000 }, { "epoch": 1.0237714869959742, "grad_norm": 10.067010169627945, "learning_rate": 2.8316826547413423e-07, "logits/chosen": -3.046875, "logits/rejected": -2.703125, "logps/chosen": -620.0, "logps/rejected": -1008.0, "loss": 0.1777, "rewards/accuracies": 0.90625, "rewards/chosen": -4.59375, "rewards/margins": 3.859375, "rewards/rejected": -8.4375, "step": 8010 }, { "epoch": 1.0250495239312416, "grad_norm": 10.700118461764552, "learning_rate": 2.826154211166417e-07, "logits/chosen": -2.90625, "logits/rejected": -2.921875, "logps/chosen": -596.0, "logps/rejected": -972.0, "loss": 0.1872, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.25, "rewards/margins": 4.0625, "rewards/rejected": -8.3125, "step": 8020 }, { "epoch": 1.026327560866509, "grad_norm": 11.209752195570033, "learning_rate": 2.820624144540377e-07, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -612.0, "logps/rejected": -1032.0, "loss": 0.1639, "rewards/accuracies": 0.981249988079071, "rewards/chosen": -4.53125, "rewards/margins": 4.21875, "rewards/rejected": -8.75, "step": 8030 }, { "epoch": 1.0276055978017764, "grad_norm": 6.347408079079281, "learning_rate": 2.8150924823826625e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -680.0, "logps/rejected": -1088.0, "loss": 0.1655, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.09375, "rewards/rejected": -9.4375, "step": 8040 }, { "epoch": 1.0288836347370438, "grad_norm": 12.357991396104282, "learning_rate": 2.809559252220652e-07, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -644.0, "logps/rejected": -1136.0, "loss": 0.1649, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.9375, "rewards/margins": 4.90625, "rewards/rejected": -9.875, "step": 8050 }, { "epoch": 1.0301616716723114, "grad_norm": 9.8048668569966, "learning_rate": 2.8040244815895254e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -672.0, "logps/rejected": -1040.0, "loss": 0.1731, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.09375, "rewards/margins": 3.96875, "rewards/rejected": -9.0625, "step": 8060 }, { "epoch": 1.0314397086075788, "grad_norm": 7.393877631897585, "learning_rate": 2.798488198032131e-07, "logits/chosen": -3.125, "logits/rejected": -2.875, "logps/chosen": -656.0, "logps/rejected": -1120.0, "loss": 0.1781, "rewards/accuracies": 0.9375, "rewards/chosen": -4.9375, "rewards/margins": 4.65625, "rewards/rejected": -9.5625, "step": 8070 }, { "epoch": 1.0327177455428462, "grad_norm": 10.729316226388566, "learning_rate": 2.792950429098844e-07, "logits/chosen": -2.953125, "logits/rejected": -2.703125, "logps/chosen": -640.0, "logps/rejected": -1020.0, "loss": 0.1625, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.78125, "rewards/rejected": -8.625, "step": 8080 }, { "epoch": 1.0339957824781136, "grad_norm": 7.516989134163046, "learning_rate": 2.7874112023474333e-07, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -696.0, "logps/rejected": -1072.0, "loss": 0.17, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.375, "rewards/margins": 3.96875, "rewards/rejected": -9.3125, "step": 8090 }, { "epoch": 1.035273819413381, "grad_norm": 6.442338927759854, "learning_rate": 2.781870545342921e-07, "logits/chosen": -3.015625, "logits/rejected": -3.015625, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1694, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.5625, "rewards/margins": 4.28125, "rewards/rejected": -9.875, "step": 8100 }, { "epoch": 1.0365518563486484, "grad_norm": 11.18028552973371, "learning_rate": 2.7763284856574483e-07, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -672.0, "logps/rejected": -1112.0, "loss": 0.1788, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.15625, "rewards/margins": 4.5, "rewards/rejected": -9.6875, "step": 8110 }, { "epoch": 1.0378298932839158, "grad_norm": 10.878882803166348, "learning_rate": 2.7707850508701323e-07, "logits/chosen": -3.046875, "logits/rejected": -2.734375, "logps/chosen": -680.0, "logps/rejected": -1056.0, "loss": 0.1829, "rewards/accuracies": 0.90625, "rewards/chosen": -5.25, "rewards/margins": 3.875, "rewards/rejected": -9.125, "step": 8120 }, { "epoch": 1.0391079302191832, "grad_norm": 9.574475689760854, "learning_rate": 2.7652402685669387e-07, "logits/chosen": -3.046875, "logits/rejected": -2.875, "logps/chosen": -656.0, "logps/rejected": -1004.0, "loss": 0.1775, "rewards/accuracies": 0.9375, "rewards/chosen": -4.9375, "rewards/margins": 3.6875, "rewards/rejected": -8.625, "step": 8130 }, { "epoch": 1.0403859671544509, "grad_norm": 11.199987026614055, "learning_rate": 2.7596941663405345e-07, "logits/chosen": -3.125, "logits/rejected": -2.875, "logps/chosen": -704.0, "logps/rejected": -1056.0, "loss": 0.1726, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.59375, "rewards/margins": 3.625, "rewards/rejected": -9.25, "step": 8140 }, { "epoch": 1.0416640040897183, "grad_norm": 8.36200035927947, "learning_rate": 2.7541467717901573e-07, "logits/chosen": -3.125, "logits/rejected": -2.890625, "logps/chosen": -688.0, "logps/rejected": -1088.0, "loss": 0.1811, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.375, "rewards/margins": 4.21875, "rewards/rejected": -9.5625, "step": 8150 }, { "epoch": 1.0429420410249857, "grad_norm": 8.854793907312779, "learning_rate": 2.7485981125214764e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -624.0, "logps/rejected": -984.0, "loss": 0.1955, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.6875, "rewards/rejected": -8.4375, "step": 8160 }, { "epoch": 1.044220077960253, "grad_norm": 7.9267747325867095, "learning_rate": 2.743048216146452e-07, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -628.0, "logps/rejected": -1020.0, "loss": 0.1542, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.75, "rewards/margins": 3.890625, "rewards/rejected": -8.625, "step": 8170 }, { "epoch": 1.0454981148955205, "grad_norm": 10.39920919594746, "learning_rate": 2.737497110283202e-07, "logits/chosen": -3.1875, "logits/rejected": -2.921875, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1638, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.3125, "rewards/rejected": -9.8125, "step": 8180 }, { "epoch": 1.0467761518307879, "grad_norm": 9.61248623593858, "learning_rate": 2.731944822555864e-07, "logits/chosen": -3.125, "logits/rejected": -2.71875, "logps/chosen": -716.0, "logps/rejected": -1184.0, "loss": 0.1757, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.625, "rewards/margins": 4.53125, "rewards/rejected": -10.1875, "step": 8190 }, { "epoch": 1.0480541887660553, "grad_norm": 12.611978808856993, "learning_rate": 2.726391380594456e-07, "logits/chosen": -3.140625, "logits/rejected": -2.84375, "logps/chosen": -692.0, "logps/rejected": -1096.0, "loss": 0.1872, "rewards/accuracies": 0.90625, "rewards/chosen": -5.15625, "rewards/margins": 4.1875, "rewards/rejected": -9.375, "step": 8200 }, { "epoch": 1.0493322257013227, "grad_norm": 5.8579043608799015, "learning_rate": 2.720836812034741e-07, "logits/chosen": -3.03125, "logits/rejected": -2.859375, "logps/chosen": -688.0, "logps/rejected": -1056.0, "loss": 0.1775, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.25, "rewards/margins": 3.984375, "rewards/rejected": -9.25, "step": 8210 }, { "epoch": 1.0506102626365903, "grad_norm": 11.77064222897185, "learning_rate": 2.715281144518085e-07, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -708.0, "logps/rejected": -1080.0, "loss": 0.1777, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.03125, "rewards/rejected": -9.3125, "step": 8220 }, { "epoch": 1.0518882995718577, "grad_norm": 9.77725061181982, "learning_rate": 2.709724405691327e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -660.0, "logps/rejected": -1024.0, "loss": 0.175, "rewards/accuracies": 0.90625, "rewards/chosen": -5.0625, "rewards/margins": 3.703125, "rewards/rejected": -8.75, "step": 8230 }, { "epoch": 1.053166336507125, "grad_norm": 8.648896874296353, "learning_rate": 2.7041666232066346e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -632.0, "logps/rejected": -984.0, "loss": 0.176, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.78125, "rewards/margins": 3.65625, "rewards/rejected": -8.4375, "step": 8240 }, { "epoch": 1.0544443734423925, "grad_norm": 9.555378197262757, "learning_rate": 2.69860782472137e-07, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -660.0, "logps/rejected": -1072.0, "loss": 0.1941, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.15625, "rewards/margins": 4.09375, "rewards/rejected": -9.25, "step": 8250 }, { "epoch": 1.05572241037766, "grad_norm": 8.749517199368952, "learning_rate": 2.6930480378979513e-07, "logits/chosen": -2.96875, "logits/rejected": -2.75, "logps/chosen": -624.0, "logps/rejected": -1012.0, "loss": 0.1828, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.71875, "rewards/margins": 3.984375, "rewards/rejected": -8.6875, "step": 8260 }, { "epoch": 1.0570004473129273, "grad_norm": 9.905567957439871, "learning_rate": 2.687487290403715e-07, "logits/chosen": -2.84375, "logits/rejected": -2.625, "logps/chosen": -644.0, "logps/rejected": -1020.0, "loss": 0.1736, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.8125, "rewards/margins": 3.84375, "rewards/rejected": -8.625, "step": 8270 }, { "epoch": 1.0582784842481947, "grad_norm": 10.75370904435311, "learning_rate": 2.6819256099107756e-07, "logits/chosen": -2.984375, "logits/rejected": -2.828125, "logps/chosen": -640.0, "logps/rejected": -1048.0, "loss": 0.1674, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.6875, "rewards/margins": 4.21875, "rewards/rejected": -8.9375, "step": 8280 }, { "epoch": 1.059556521183462, "grad_norm": 10.255362189541565, "learning_rate": 2.6763630240958946e-07, "logits/chosen": -3.03125, "logits/rejected": -2.65625, "logps/chosen": -604.0, "logps/rejected": -980.0, "loss": 0.1728, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.53125, "rewards/margins": 4.0, "rewards/rejected": -8.5, "step": 8290 }, { "epoch": 1.0608345581187297, "grad_norm": 10.246388181727587, "learning_rate": 2.6707995606403366e-07, "logits/chosen": -2.828125, "logits/rejected": -2.71875, "logps/chosen": -636.0, "logps/rejected": -1040.0, "loss": 0.1617, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.90625, "rewards/margins": 4.0625, "rewards/rejected": -8.9375, "step": 8300 }, { "epoch": 1.0621125950539971, "grad_norm": 12.929394573191676, "learning_rate": 2.665235247229734e-07, "logits/chosen": -2.96875, "logits/rejected": -2.65625, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.1687, "rewards/accuracies": 0.9375, "rewards/chosen": -5.21875, "rewards/margins": 4.3125, "rewards/rejected": -9.5625, "step": 8310 }, { "epoch": 1.0633906319892645, "grad_norm": 10.050061531106172, "learning_rate": 2.659670111553949e-07, "logits/chosen": -2.9375, "logits/rejected": -2.640625, "logps/chosen": -680.0, "logps/rejected": -1088.0, "loss": 0.1782, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.125, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 8320 }, { "epoch": 1.064668668924532, "grad_norm": 11.590725397037096, "learning_rate": 2.654104181306935e-07, "logits/chosen": -2.84375, "logits/rejected": -2.546875, "logps/chosen": -672.0, "logps/rejected": -1048.0, "loss": 0.1947, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 3.625, "rewards/rejected": -8.75, "step": 8330 }, { "epoch": 1.0659467058597993, "grad_norm": 10.884747764045423, "learning_rate": 2.6485374841865994e-07, "logits/chosen": -2.921875, "logits/rejected": -2.703125, "logps/chosen": -672.0, "logps/rejected": -1072.0, "loss": 0.1643, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.09375, "rewards/rejected": -9.4375, "step": 8340 }, { "epoch": 1.0672247427950667, "grad_norm": 9.412234086973548, "learning_rate": 2.6429700478946667e-07, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -700.0, "logps/rejected": -1064.0, "loss": 0.168, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 3.890625, "rewards/rejected": -9.375, "step": 8350 }, { "epoch": 1.0685027797303341, "grad_norm": 9.866005395581741, "learning_rate": 2.6374019001365396e-07, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -664.0, "logps/rejected": -1064.0, "loss": 0.1699, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.09375, "rewards/margins": 4.03125, "rewards/rejected": -9.125, "step": 8360 }, { "epoch": 1.0697808166656015, "grad_norm": 12.839280711409044, "learning_rate": 2.631833068621161e-07, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -632.0, "logps/rejected": -1008.0, "loss": 0.185, "rewards/accuracies": 0.90625, "rewards/chosen": -4.9375, "rewards/margins": 3.671875, "rewards/rejected": -8.625, "step": 8370 }, { "epoch": 1.0710588536008692, "grad_norm": 9.543783056734359, "learning_rate": 2.6262635810608763e-07, "logits/chosen": -2.96875, "logits/rejected": -2.6875, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1583, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.25, "rewards/rejected": -9.625, "step": 8380 }, { "epoch": 1.0723368905361366, "grad_norm": 8.913154707551678, "learning_rate": 2.620693465171295e-07, "logits/chosen": -2.9375, "logits/rejected": -2.765625, "logps/chosen": -648.0, "logps/rejected": -1032.0, "loss": 0.1592, "rewards/accuracies": 0.9375, "rewards/chosen": -4.875, "rewards/margins": 4.0625, "rewards/rejected": -8.9375, "step": 8390 }, { "epoch": 1.073614927471404, "grad_norm": 12.493619267004417, "learning_rate": 2.6151227486711547e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -636.0, "logps/rejected": -1032.0, "loss": 0.1827, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.875, "rewards/margins": 4.125, "rewards/rejected": -9.0, "step": 8400 }, { "epoch": 1.0748929644066714, "grad_norm": 11.245710975859144, "learning_rate": 2.6095514592821814e-07, "logits/chosen": -3.140625, "logits/rejected": -2.796875, "logps/chosen": -684.0, "logps/rejected": -1056.0, "loss": 0.1703, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.28125, "rewards/margins": 3.953125, "rewards/rejected": -9.25, "step": 8410 }, { "epoch": 1.0761710013419388, "grad_norm": 10.870414471136572, "learning_rate": 2.603979624728952e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -628.0, "logps/rejected": -1000.0, "loss": 0.1635, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.71875, "rewards/margins": 3.6875, "rewards/rejected": -8.375, "step": 8420 }, { "epoch": 1.0774490382772062, "grad_norm": 11.81784499074986, "learning_rate": 2.598407272738756e-07, "logits/chosen": -3.015625, "logits/rejected": -2.859375, "logps/chosen": -664.0, "logps/rejected": -1040.0, "loss": 0.1618, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.90625, "rewards/margins": 4.15625, "rewards/rejected": -9.0625, "step": 8430 }, { "epoch": 1.0787270752124736, "grad_norm": 8.920813120865647, "learning_rate": 2.592834431041457e-07, "logits/chosen": -2.953125, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1088.0, "loss": 0.1677, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.0625, "rewards/rejected": -9.5, "step": 8440 }, { "epoch": 1.080005112147741, "grad_norm": 9.651821827804635, "learning_rate": 2.5872611273693573e-07, "logits/chosen": -2.96875, "logits/rejected": -2.5625, "logps/chosen": -684.0, "logps/rejected": -1096.0, "loss": 0.1783, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.34375, "rewards/margins": 4.15625, "rewards/rejected": -9.5, "step": 8450 }, { "epoch": 1.0812831490830086, "grad_norm": 7.782821307879692, "learning_rate": 2.581687389457058e-07, "logits/chosen": -2.8125, "logits/rejected": -2.5, "logps/chosen": -696.0, "logps/rejected": -1040.0, "loss": 0.1506, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.34375, "rewards/margins": 3.8125, "rewards/rejected": -9.125, "step": 8460 }, { "epoch": 1.082561186018276, "grad_norm": 11.993079437671195, "learning_rate": 2.576113245041319e-07, "logits/chosen": -2.96875, "logits/rejected": -2.671875, "logps/chosen": -708.0, "logps/rejected": -1088.0, "loss": 0.1714, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.125, "rewards/rejected": -9.625, "step": 8470 }, { "epoch": 1.0838392229535434, "grad_norm": 9.027091864147854, "learning_rate": 2.5705387218609267e-07, "logits/chosen": -2.921875, "logits/rejected": -2.59375, "logps/chosen": -704.0, "logps/rejected": -1072.0, "loss": 0.1875, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.3125, "rewards/margins": 3.671875, "rewards/rejected": -9.0, "step": 8480 }, { "epoch": 1.0851172598888108, "grad_norm": 11.172202529858502, "learning_rate": 2.5649638476565485e-07, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -644.0, "logps/rejected": -1032.0, "loss": 0.1774, "rewards/accuracies": 0.9375, "rewards/chosen": -4.90625, "rewards/margins": 3.984375, "rewards/rejected": -8.875, "step": 8490 }, { "epoch": 1.0863952968240782, "grad_norm": 8.366326928331228, "learning_rate": 2.5593886501705994e-07, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -676.0, "logps/rejected": -1040.0, "loss": 0.1814, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.25, "rewards/margins": 3.890625, "rewards/rejected": -9.125, "step": 8500 }, { "epoch": 1.0876733337593456, "grad_norm": 15.49608599918426, "learning_rate": 2.553813157147107e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -700.0, "logps/rejected": -1056.0, "loss": 0.1688, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.3125, "rewards/margins": 3.703125, "rewards/rejected": -9.0, "step": 8510 }, { "epoch": 1.088951370694613, "grad_norm": 13.194232623449643, "learning_rate": 2.548237396331564e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -668.0, "logps/rejected": -1080.0, "loss": 0.1812, "rewards/accuracies": 0.9375, "rewards/chosen": -5.0, "rewards/margins": 4.28125, "rewards/rejected": -9.25, "step": 8520 }, { "epoch": 1.0902294076298804, "grad_norm": 10.570147031293036, "learning_rate": 2.5426613954708e-07, "logits/chosen": -2.96875, "logits/rejected": -2.75, "logps/chosen": -636.0, "logps/rejected": -996.0, "loss": 0.1892, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.84375, "rewards/margins": 3.78125, "rewards/rejected": -8.625, "step": 8530 }, { "epoch": 1.091507444565148, "grad_norm": 12.792969719649745, "learning_rate": 2.5370851823128376e-07, "logits/chosen": -2.796875, "logits/rejected": -2.5625, "logps/chosen": -692.0, "logps/rejected": -1088.0, "loss": 0.1781, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 3.96875, "rewards/rejected": -9.125, "step": 8540 }, { "epoch": 1.0927854815004154, "grad_norm": 8.155866731584963, "learning_rate": 2.5315087846067546e-07, "logits/chosen": -2.875, "logits/rejected": -2.734375, "logps/chosen": -660.0, "logps/rejected": -1096.0, "loss": 0.1661, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.125, "rewards/margins": 4.3125, "rewards/rejected": -9.4375, "step": 8550 }, { "epoch": 1.0940635184356828, "grad_norm": 11.43299208228307, "learning_rate": 2.5259322301025493e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -656.0, "logps/rejected": -1072.0, "loss": 0.1785, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.8125, "rewards/margins": 4.34375, "rewards/rejected": -9.125, "step": 8560 }, { "epoch": 1.0953415553709502, "grad_norm": 11.581246848208384, "learning_rate": 2.520355546550999e-07, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -616.0, "logps/rejected": -1020.0, "loss": 0.1762, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.75, "rewards/margins": 4.125, "rewards/rejected": -8.875, "step": 8570 }, { "epoch": 1.0966195923062176, "grad_norm": 7.5531191014349375, "learning_rate": 2.5147787617035235e-07, "logits/chosen": -2.984375, "logits/rejected": -2.78125, "logps/chosen": -588.0, "logps/rejected": -996.0, "loss": 0.1516, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.375, "rewards/margins": 4.28125, "rewards/rejected": -8.6875, "step": 8580 }, { "epoch": 1.097897629241485, "grad_norm": 12.22511911705573, "learning_rate": 2.5092019033120484e-07, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -632.0, "logps/rejected": -1056.0, "loss": 0.1901, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.71875, "rewards/margins": 4.46875, "rewards/rejected": -9.1875, "step": 8590 }, { "epoch": 1.0991756661767524, "grad_norm": 10.172246205013995, "learning_rate": 2.5036249991288614e-07, "logits/chosen": -3.09375, "logits/rejected": -2.90625, "logps/chosen": -660.0, "logps/rejected": -1064.0, "loss": 0.1892, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.90625, "rewards/margins": 4.1875, "rewards/rejected": -9.0625, "step": 8600 }, { "epoch": 1.1004537031120198, "grad_norm": 8.523153345006731, "learning_rate": 2.4980480769064805e-07, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -604.0, "logps/rejected": -964.0, "loss": 0.1713, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.5625, "rewards/margins": 3.75, "rewards/rejected": -8.3125, "step": 8610 }, { "epoch": 1.1017317400472875, "grad_norm": 13.46349720976093, "learning_rate": 2.492471164397514e-07, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -616.0, "logps/rejected": -988.0, "loss": 0.1783, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.46875, "rewards/margins": 3.953125, "rewards/rejected": -8.4375, "step": 8620 }, { "epoch": 1.1030097769825549, "grad_norm": 12.692605192581905, "learning_rate": 2.486894289354521e-07, "logits/chosen": -3.046875, "logits/rejected": -2.875, "logps/chosen": -580.0, "logps/rejected": -1008.0, "loss": 0.1991, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.34375, "rewards/margins": 4.28125, "rewards/rejected": -8.625, "step": 8630 }, { "epoch": 1.1042878139178223, "grad_norm": 11.684460877635107, "learning_rate": 2.4813174795298753e-07, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -588.0, "logps/rejected": -980.0, "loss": 0.1707, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.28125, "rewards/margins": 4.1875, "rewards/rejected": -8.5, "step": 8640 }, { "epoch": 1.1055658508530897, "grad_norm": 11.138998731809462, "learning_rate": 2.475740762675623e-07, "logits/chosen": -2.75, "logits/rejected": -2.546875, "logps/chosen": -600.0, "logps/rejected": -968.0, "loss": 0.1698, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.25, "rewards/margins": 3.625, "rewards/rejected": -7.875, "step": 8650 }, { "epoch": 1.106843887788357, "grad_norm": 10.182232860207785, "learning_rate": 2.4701641665433514e-07, "logits/chosen": -2.78125, "logits/rejected": -2.578125, "logps/chosen": -644.0, "logps/rejected": -980.0, "loss": 0.1732, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.84375, "rewards/margins": 3.671875, "rewards/rejected": -8.5, "step": 8660 }, { "epoch": 1.1081219247236245, "grad_norm": 11.533017525853683, "learning_rate": 2.464587718884044e-07, "logits/chosen": -2.84375, "logits/rejected": -2.6875, "logps/chosen": -620.0, "logps/rejected": -996.0, "loss": 0.1505, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.65625, "rewards/margins": 3.984375, "rewards/rejected": -8.625, "step": 8670 }, { "epoch": 1.1093999616588919, "grad_norm": 10.19906423287545, "learning_rate": 2.459011447447947e-07, "logits/chosen": -2.953125, "logits/rejected": -2.859375, "logps/chosen": -600.0, "logps/rejected": -960.0, "loss": 0.1667, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.375, "rewards/margins": 3.8125, "rewards/rejected": -8.1875, "step": 8680 }, { "epoch": 1.1106779985941593, "grad_norm": 12.307725593269353, "learning_rate": 2.45343537998443e-07, "logits/chosen": -2.9375, "logits/rejected": -2.84375, "logps/chosen": -632.0, "logps/rejected": -1004.0, "loss": 0.1712, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.53125, "rewards/margins": 3.96875, "rewards/rejected": -8.5, "step": 8690 }, { "epoch": 1.111956035529427, "grad_norm": 13.901330347133428, "learning_rate": 2.4478595442418473e-07, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -640.0, "logps/rejected": -1016.0, "loss": 0.1714, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -4.78125, "rewards/margins": 4.03125, "rewards/rejected": -8.8125, "step": 8700 }, { "epoch": 1.1132340724646943, "grad_norm": 8.802904680915939, "learning_rate": 2.4422839679673996e-07, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -656.0, "logps/rejected": -984.0, "loss": 0.1607, "rewards/accuracies": 0.862500011920929, "rewards/chosen": -4.96875, "rewards/margins": 3.703125, "rewards/rejected": -8.6875, "step": 8710 }, { "epoch": 1.1145121093999617, "grad_norm": 10.39181707302118, "learning_rate": 2.4367086789069954e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -628.0, "logps/rejected": -1032.0, "loss": 0.1968, "rewards/accuracies": 0.875, "rewards/chosen": -4.84375, "rewards/margins": 4.0625, "rewards/rejected": -8.875, "step": 8720 }, { "epoch": 1.115790146335229, "grad_norm": 8.971322923592636, "learning_rate": 2.4311337048051163e-07, "logits/chosen": -3.0, "logits/rejected": -2.71875, "logps/chosen": -668.0, "logps/rejected": -1128.0, "loss": 0.1911, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.0, "rewards/margins": 4.5625, "rewards/rejected": -9.5625, "step": 8730 }, { "epoch": 1.1170681832704965, "grad_norm": 8.71816008693693, "learning_rate": 2.425559073404675e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -612.0, "logps/rejected": -1012.0, "loss": 0.1715, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.65625, "rewards/margins": 3.84375, "rewards/rejected": -8.5, "step": 8740 }, { "epoch": 1.118346220205764, "grad_norm": 11.110149770588588, "learning_rate": 2.41998481244688e-07, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -640.0, "logps/rejected": -1024.0, "loss": 0.1735, "rewards/accuracies": 0.9375, "rewards/chosen": -4.8125, "rewards/margins": 4.03125, "rewards/rejected": -8.8125, "step": 8750 }, { "epoch": 1.1196242571410313, "grad_norm": 8.80837913365687, "learning_rate": 2.4144109496710936e-07, "logits/chosen": -2.90625, "logits/rejected": -2.765625, "logps/chosen": -700.0, "logps/rejected": -1064.0, "loss": 0.1754, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.34375, "rewards/margins": 3.703125, "rewards/rejected": -9.0625, "step": 8760 }, { "epoch": 1.1209022940762987, "grad_norm": 8.1589054813181, "learning_rate": 2.4088375128146997e-07, "logits/chosen": -2.96875, "logits/rejected": -2.578125, "logps/chosen": -668.0, "logps/rejected": -1088.0, "loss": 0.161, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.15625, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 8770 }, { "epoch": 1.1221803310115663, "grad_norm": 11.945639642235644, "learning_rate": 2.403264529612962e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -644.0, "logps/rejected": -1020.0, "loss": 0.1504, "rewards/accuracies": 0.90625, "rewards/chosen": -4.78125, "rewards/margins": 3.8125, "rewards/rejected": -8.5625, "step": 8780 }, { "epoch": 1.1234583679468337, "grad_norm": 11.569953335116981, "learning_rate": 2.3976920277988854e-07, "logits/chosen": -2.828125, "logits/rejected": -2.703125, "logps/chosen": -628.0, "logps/rejected": -1048.0, "loss": 0.176, "rewards/accuracies": 0.9375, "rewards/chosen": -4.6875, "rewards/margins": 4.25, "rewards/rejected": -8.9375, "step": 8790 }, { "epoch": 1.1247364048821011, "grad_norm": 13.598732599331584, "learning_rate": 2.3921200351030807e-07, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -668.0, "logps/rejected": -1016.0, "loss": 0.1702, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 3.703125, "rewards/rejected": -8.875, "step": 8800 }, { "epoch": 1.1260144418173685, "grad_norm": 7.749172338773482, "learning_rate": 2.386548579253624e-07, "logits/chosen": -2.859375, "logits/rejected": -2.625, "logps/chosen": -680.0, "logps/rejected": -1040.0, "loss": 0.1687, "rewards/accuracies": 0.875, "rewards/chosen": -5.3125, "rewards/margins": 3.640625, "rewards/rejected": -9.0, "step": 8810 }, { "epoch": 1.127292478752636, "grad_norm": 9.869203051606881, "learning_rate": 2.3809776879759203e-07, "logits/chosen": -2.859375, "logits/rejected": -2.6875, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.179, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.5, "rewards/rejected": -9.9375, "step": 8820 }, { "epoch": 1.1285705156879033, "grad_norm": 9.212740949988463, "learning_rate": 2.3754073889925655e-07, "logits/chosen": -2.921875, "logits/rejected": -2.609375, "logps/chosen": -728.0, "logps/rejected": -1120.0, "loss": 0.1675, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.4375, "rewards/margins": 4.0625, "rewards/rejected": -9.5, "step": 8830 }, { "epoch": 1.1298485526231707, "grad_norm": 8.355055868788307, "learning_rate": 2.3698377100232073e-07, "logits/chosen": -2.9375, "logits/rejected": -2.625, "logps/chosen": -680.0, "logps/rejected": -1152.0, "loss": 0.1797, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.53125, "rewards/rejected": -9.875, "step": 8840 }, { "epoch": 1.1311265895584381, "grad_norm": 8.253615583958148, "learning_rate": 2.3642686787844095e-07, "logits/chosen": -2.9375, "logits/rejected": -2.625, "logps/chosen": -700.0, "logps/rejected": -1168.0, "loss": 0.1772, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.59375, "rewards/rejected": -9.9375, "step": 8850 }, { "epoch": 1.1324046264937055, "grad_norm": 11.200655287502787, "learning_rate": 2.3587003229895108e-07, "logits/chosen": -2.984375, "logits/rejected": -2.65625, "logps/chosen": -628.0, "logps/rejected": -1020.0, "loss": 0.1658, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.65625, "rewards/margins": 4.25, "rewards/rejected": -8.875, "step": 8860 }, { "epoch": 1.1336826634289732, "grad_norm": 9.215990244514574, "learning_rate": 2.3531326703484895e-07, "logits/chosen": -2.953125, "logits/rejected": -2.671875, "logps/chosen": -620.0, "logps/rejected": -1040.0, "loss": 0.1758, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.65625, "rewards/margins": 4.1875, "rewards/rejected": -8.8125, "step": 8870 }, { "epoch": 1.1349607003642406, "grad_norm": 14.804600873057607, "learning_rate": 2.3475657485678235e-07, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -616.0, "logps/rejected": -1008.0, "loss": 0.1901, "rewards/accuracies": 0.90625, "rewards/chosen": -4.625, "rewards/margins": 4.0, "rewards/rejected": -8.625, "step": 8880 }, { "epoch": 1.136238737299508, "grad_norm": 9.707687199553083, "learning_rate": 2.341999585350356e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -616.0, "logps/rejected": -976.0, "loss": 0.1887, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.6875, "rewards/margins": 3.8125, "rewards/rejected": -8.5, "step": 8890 }, { "epoch": 1.1375167742347754, "grad_norm": 12.930001235443862, "learning_rate": 2.336434208395154e-07, "logits/chosen": -3.0, "logits/rejected": -2.71875, "logps/chosen": -628.0, "logps/rejected": -1016.0, "loss": 0.18, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.84375, "rewards/margins": 3.96875, "rewards/rejected": -8.8125, "step": 8900 }, { "epoch": 1.1387948111700428, "grad_norm": 10.257419846335008, "learning_rate": 2.3308696453973726e-07, "logits/chosen": -2.953125, "logits/rejected": -2.6875, "logps/chosen": -664.0, "logps/rejected": -1064.0, "loss": 0.1751, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.09375, "rewards/margins": 4.28125, "rewards/rejected": -9.375, "step": 8910 }, { "epoch": 1.1400728481053102, "grad_norm": 8.426972164720423, "learning_rate": 2.325305924048115e-07, "logits/chosen": -2.875, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1080.0, "loss": 0.1769, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 3.859375, "rewards/rejected": -9.3125, "step": 8920 }, { "epoch": 1.1413508850405776, "grad_norm": 9.51570501986898, "learning_rate": 2.319743072034297e-07, "logits/chosen": -2.90625, "logits/rejected": -2.703125, "logps/chosen": -684.0, "logps/rejected": -1096.0, "loss": 0.1778, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.3125, "rewards/rejected": -9.5, "step": 8930 }, { "epoch": 1.1426289219758452, "grad_norm": 13.154249035166893, "learning_rate": 2.3141811170385078e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -632.0, "logps/rejected": -1080.0, "loss": 0.1927, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -4.8125, "rewards/margins": 4.65625, "rewards/rejected": -9.5, "step": 8940 }, { "epoch": 1.1439069589111126, "grad_norm": 8.837785292901133, "learning_rate": 2.308620086738874e-07, "logits/chosen": -2.9375, "logits/rejected": -2.59375, "logps/chosen": -668.0, "logps/rejected": -1012.0, "loss": 0.1807, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.09375, "rewards/margins": 3.578125, "rewards/rejected": -8.6875, "step": 8950 }, { "epoch": 1.14518499584638, "grad_norm": 10.641521132461545, "learning_rate": 2.303060008808919e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -656.0, "logps/rejected": -1080.0, "loss": 0.1879, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.03125, "rewards/margins": 4.375, "rewards/rejected": -9.4375, "step": 8960 }, { "epoch": 1.1464630327816474, "grad_norm": 7.139401203739757, "learning_rate": 2.2975009109174282e-07, "logits/chosen": -3.03125, "logits/rejected": -2.6875, "logps/chosen": -668.0, "logps/rejected": -1136.0, "loss": 0.1647, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.15625, "rewards/margins": 4.75, "rewards/rejected": -9.9375, "step": 8970 }, { "epoch": 1.1477410697169148, "grad_norm": 9.674884483133546, "learning_rate": 2.291942820728308e-07, "logits/chosen": -3.015625, "logits/rejected": -2.6875, "logps/chosen": -640.0, "logps/rejected": -1096.0, "loss": 0.183, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.875, "rewards/margins": 4.53125, "rewards/rejected": -9.375, "step": 8980 }, { "epoch": 1.1490191066521822, "grad_norm": 7.8995664380627435, "learning_rate": 2.286385765900453e-07, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -628.0, "logps/rejected": -1032.0, "loss": 0.1661, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.65625, "rewards/margins": 4.28125, "rewards/rejected": -8.9375, "step": 8990 }, { "epoch": 1.1502971435874496, "grad_norm": 10.201099604073594, "learning_rate": 2.2808297740876038e-07, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -668.0, "logps/rejected": -1024.0, "loss": 0.1818, "rewards/accuracies": 0.90625, "rewards/chosen": -5.03125, "rewards/margins": 3.71875, "rewards/rejected": -8.75, "step": 9000 }, { "epoch": 1.151575180522717, "grad_norm": 13.045559871799567, "learning_rate": 2.2752748729382108e-07, "logits/chosen": -3.046875, "logits/rejected": -2.65625, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.162, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.15625, "rewards/margins": 4.59375, "rewards/rejected": -9.75, "step": 9010 }, { "epoch": 1.1528532174579844, "grad_norm": 10.931551568686894, "learning_rate": 2.269721090095298e-07, "logits/chosen": -2.875, "logits/rejected": -2.78125, "logps/chosen": -752.0, "logps/rejected": -1168.0, "loss": 0.1973, "rewards/accuracies": 0.90625, "rewards/chosen": -5.8125, "rewards/margins": 4.21875, "rewards/rejected": -10.0, "step": 9020 }, { "epoch": 1.154131254393252, "grad_norm": 9.588251761779542, "learning_rate": 2.2641684531963233e-07, "logits/chosen": -2.96875, "logits/rejected": -2.65625, "logps/chosen": -696.0, "logps/rejected": -1056.0, "loss": 0.1832, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.375, "rewards/margins": 3.640625, "rewards/rejected": -9.0, "step": 9030 }, { "epoch": 1.1554092913285194, "grad_norm": 10.087702912942275, "learning_rate": 2.2586169898730425e-07, "logits/chosen": -2.921875, "logits/rejected": -2.65625, "logps/chosen": -660.0, "logps/rejected": -1024.0, "loss": 0.1692, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 3.875, "rewards/rejected": -9.0, "step": 9040 }, { "epoch": 1.1566873282637868, "grad_norm": 12.558033292582566, "learning_rate": 2.2530667277513713e-07, "logits/chosen": -2.890625, "logits/rejected": -2.59375, "logps/chosen": -688.0, "logps/rejected": -1064.0, "loss": 0.1698, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.375, "rewards/margins": 3.84375, "rewards/rejected": -9.1875, "step": 9050 }, { "epoch": 1.1579653651990542, "grad_norm": 9.862234243673335, "learning_rate": 2.247517694451248e-07, "logits/chosen": -2.84375, "logits/rejected": -2.6875, "logps/chosen": -736.0, "logps/rejected": -1144.0, "loss": 0.1714, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.3125, "rewards/rejected": -9.875, "step": 9060 }, { "epoch": 1.1592434021343216, "grad_norm": 8.483102506930782, "learning_rate": 2.2419699175864964e-07, "logits/chosen": -2.8125, "logits/rejected": -2.65625, "logps/chosen": -660.0, "logps/rejected": -1040.0, "loss": 0.1684, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -4.96875, "rewards/margins": 4.03125, "rewards/rejected": -9.0, "step": 9070 }, { "epoch": 1.160521439069589, "grad_norm": 10.365093327270756, "learning_rate": 2.2364234247646853e-07, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -660.0, "logps/rejected": -1032.0, "loss": 0.1785, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.03125, "rewards/margins": 3.859375, "rewards/rejected": -8.9375, "step": 9080 }, { "epoch": 1.1617994760048564, "grad_norm": 11.078748531859485, "learning_rate": 2.230878243586997e-07, "logits/chosen": -2.875, "logits/rejected": -2.65625, "logps/chosen": -696.0, "logps/rejected": -1048.0, "loss": 0.1895, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.4375, "rewards/margins": 3.828125, "rewards/rejected": -9.25, "step": 9090 }, { "epoch": 1.163077512940124, "grad_norm": 11.756344722412415, "learning_rate": 2.225334401648084e-07, "logits/chosen": -2.9375, "logits/rejected": -2.75, "logps/chosen": -756.0, "logps/rejected": -1144.0, "loss": 0.1722, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.8125, "rewards/margins": 4.21875, "rewards/rejected": -10.0625, "step": 9100 }, { "epoch": 1.1643555498753915, "grad_norm": 12.728237864635561, "learning_rate": 2.219791926535936e-07, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -728.0, "logps/rejected": -1096.0, "loss": 0.1621, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.59375, "rewards/margins": 3.90625, "rewards/rejected": -9.5, "step": 9110 }, { "epoch": 1.1656335868106589, "grad_norm": 8.251473794642155, "learning_rate": 2.2142508458317406e-07, "logits/chosen": -2.921875, "logits/rejected": -2.703125, "logps/chosen": -700.0, "logps/rejected": -1080.0, "loss": 0.172, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.03125, "rewards/rejected": -9.4375, "step": 9120 }, { "epoch": 1.1669116237459263, "grad_norm": 13.422178539672343, "learning_rate": 2.2087111871097455e-07, "logits/chosen": -3.03125, "logits/rejected": -2.65625, "logps/chosen": -744.0, "logps/rejected": -1176.0, "loss": 0.1675, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.96875, "rewards/margins": 4.28125, "rewards/rejected": -10.25, "step": 9130 }, { "epoch": 1.1681896606811937, "grad_norm": 15.75211685357834, "learning_rate": 2.2031729779371237e-07, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -748.0, "logps/rejected": -1176.0, "loss": 0.1719, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -6.0, "rewards/margins": 4.1875, "rewards/rejected": -10.1875, "step": 9140 }, { "epoch": 1.169467697616461, "grad_norm": 9.128192258566758, "learning_rate": 2.1976362458738333e-07, "logits/chosen": -2.890625, "logits/rejected": -2.703125, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1922, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.59375, "rewards/margins": 4.09375, "rewards/rejected": -9.6875, "step": 9150 }, { "epoch": 1.1707457345517285, "grad_norm": 9.468398215689838, "learning_rate": 2.1921010184724816e-07, "logits/chosen": -2.828125, "logits/rejected": -2.65625, "logps/chosen": -708.0, "logps/rejected": -1104.0, "loss": 0.163, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 9160 }, { "epoch": 1.1720237714869959, "grad_norm": 13.013947158173366, "learning_rate": 2.1865673232781901e-07, "logits/chosen": -2.859375, "logits/rejected": -2.6875, "logps/chosen": -716.0, "logps/rejected": -1112.0, "loss": 0.1881, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.03125, "rewards/rejected": -9.625, "step": 9170 }, { "epoch": 1.1733018084222633, "grad_norm": 10.06906757226758, "learning_rate": 2.1810351878284538e-07, "logits/chosen": -2.796875, "logits/rejected": -2.484375, "logps/chosen": -732.0, "logps/rejected": -1088.0, "loss": 0.1936, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.78125, "rewards/margins": 3.734375, "rewards/rejected": -9.5, "step": 9180 }, { "epoch": 1.174579845357531, "grad_norm": 13.95054752379249, "learning_rate": 2.1755046396530074e-07, "logits/chosen": -2.75, "logits/rejected": -2.65625, "logps/chosen": -712.0, "logps/rejected": -1064.0, "loss": 0.1618, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.59375, "rewards/margins": 3.734375, "rewards/rejected": -9.3125, "step": 9190 }, { "epoch": 1.1758578822927983, "grad_norm": 12.364774430012691, "learning_rate": 2.1699757062736845e-07, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.155, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.25, "rewards/rejected": -9.75, "step": 9200 }, { "epoch": 1.1771359192280657, "grad_norm": 8.761179985697268, "learning_rate": 2.164448415204285e-07, "logits/chosen": -2.875, "logits/rejected": -2.6875, "logps/chosen": -732.0, "logps/rejected": -1160.0, "loss": 0.1498, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.46875, "rewards/rejected": -10.0625, "step": 9210 }, { "epoch": 1.178413956163333, "grad_norm": 13.107311109960998, "learning_rate": 2.158922793950435e-07, "logits/chosen": -2.921875, "logits/rejected": -2.578125, "logps/chosen": -716.0, "logps/rejected": -1160.0, "loss": 0.1694, "rewards/accuracies": 0.9375, "rewards/chosen": -5.625, "rewards/margins": 4.6875, "rewards/rejected": -10.3125, "step": 9220 }, { "epoch": 1.1796919930986005, "grad_norm": 8.684428498052904, "learning_rate": 2.153398870009453e-07, "logits/chosen": -2.96875, "logits/rejected": -2.609375, "logps/chosen": -760.0, "logps/rejected": -1160.0, "loss": 0.1783, "rewards/accuracies": 0.90625, "rewards/chosen": -5.96875, "rewards/margins": 4.125, "rewards/rejected": -10.0625, "step": 9230 }, { "epoch": 1.180970030033868, "grad_norm": 8.989886993845838, "learning_rate": 2.1478766708702074e-07, "logits/chosen": -2.921875, "logits/rejected": -2.484375, "logps/chosen": -712.0, "logps/rejected": -1064.0, "loss": 0.157, "rewards/accuracies": 0.90625, "rewards/chosen": -5.5625, "rewards/margins": 3.796875, "rewards/rejected": -9.375, "step": 9240 }, { "epoch": 1.1822480669691353, "grad_norm": 9.158778773808871, "learning_rate": 2.1423562240129869e-07, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -732.0, "logps/rejected": -1152.0, "loss": 0.1642, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.625, "rewards/margins": 4.59375, "rewards/rejected": -10.25, "step": 9250 }, { "epoch": 1.183526103904403, "grad_norm": 12.167043112847624, "learning_rate": 2.1368375569093587e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1717, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.6875, "rewards/margins": 4.0, "rewards/rejected": -9.6875, "step": 9260 }, { "epoch": 1.1848041408396703, "grad_norm": 9.956515597502806, "learning_rate": 2.1313206970220338e-07, "logits/chosen": -2.921875, "logits/rejected": -2.671875, "logps/chosen": -716.0, "logps/rejected": -1096.0, "loss": 0.1736, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.53125, "rewards/margins": 3.96875, "rewards/rejected": -9.5, "step": 9270 }, { "epoch": 1.1860821777749377, "grad_norm": 8.82080127275406, "learning_rate": 2.1258056718047298e-07, "logits/chosen": -2.875, "logits/rejected": -2.625, "logps/chosen": -692.0, "logps/rejected": -1080.0, "loss": 0.1758, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.25, "rewards/margins": 4.15625, "rewards/rejected": -9.375, "step": 9280 }, { "epoch": 1.1873602147102051, "grad_norm": 10.48275029797209, "learning_rate": 2.1202925087020346e-07, "logits/chosen": -3.015625, "logits/rejected": -2.625, "logps/chosen": -716.0, "logps/rejected": -1136.0, "loss": 0.1759, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 9290 }, { "epoch": 1.1886382516454725, "grad_norm": 10.416324291817167, "learning_rate": 2.1147812351492687e-07, "logits/chosen": -2.90625, "logits/rejected": -2.734375, "logps/chosen": -700.0, "logps/rejected": -1096.0, "loss": 0.1829, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.0, "rewards/rejected": -9.3125, "step": 9300 }, { "epoch": 1.18991628858074, "grad_norm": 12.359961817177224, "learning_rate": 2.1092718785723497e-07, "logits/chosen": -2.90625, "logits/rejected": -2.78125, "logps/chosen": -724.0, "logps/rejected": -1120.0, "loss": 0.1622, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.25, "rewards/rejected": -9.8125, "step": 9310 }, { "epoch": 1.1911943255160073, "grad_norm": 9.133749012355143, "learning_rate": 2.1037644663876566e-07, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -740.0, "logps/rejected": -1120.0, "loss": 0.1815, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.8125, "rewards/margins": 4.0, "rewards/rejected": -9.8125, "step": 9320 }, { "epoch": 1.1924723624512747, "grad_norm": 8.607099471081787, "learning_rate": 2.0982590260018923e-07, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -720.0, "logps/rejected": -1144.0, "loss": 0.1873, "rewards/accuracies": 0.9375, "rewards/chosen": -5.59375, "rewards/margins": 4.4375, "rewards/rejected": -10.0625, "step": 9330 }, { "epoch": 1.1937503993865421, "grad_norm": 7.027959067289001, "learning_rate": 2.0927555848119472e-07, "logits/chosen": -2.859375, "logits/rejected": -2.65625, "logps/chosen": -720.0, "logps/rejected": -1096.0, "loss": 0.1758, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.625, "rewards/margins": 3.859375, "rewards/rejected": -9.5, "step": 9340 }, { "epoch": 1.1950284363218098, "grad_norm": 9.677106712552908, "learning_rate": 2.0872541702047635e-07, "logits/chosen": -2.953125, "logits/rejected": -2.796875, "logps/chosen": -728.0, "logps/rejected": -1088.0, "loss": 0.1637, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.65625, "rewards/margins": 3.84375, "rewards/rejected": -9.5, "step": 9350 }, { "epoch": 1.1963064732570772, "grad_norm": 9.172920122127982, "learning_rate": 2.0817548095571965e-07, "logits/chosen": -3.03125, "logits/rejected": -2.828125, "logps/chosen": -756.0, "logps/rejected": -1128.0, "loss": 0.1778, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.84375, "rewards/margins": 3.828125, "rewards/rejected": -9.6875, "step": 9360 }, { "epoch": 1.1975845101923446, "grad_norm": 14.141110110747428, "learning_rate": 2.0762575302358827e-07, "logits/chosen": -3.03125, "logits/rejected": -2.765625, "logps/chosen": -768.0, "logps/rejected": -1208.0, "loss": 0.1882, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -6.15625, "rewards/margins": 4.3125, "rewards/rejected": -10.4375, "step": 9370 }, { "epoch": 1.198862547127612, "grad_norm": 9.339406721803096, "learning_rate": 2.0707623595970995e-07, "logits/chosen": -2.96875, "logits/rejected": -2.75, "logps/chosen": -756.0, "logps/rejected": -1168.0, "loss": 0.1756, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.90625, "rewards/margins": 4.4375, "rewards/rejected": -10.3125, "step": 9380 }, { "epoch": 1.2001405840628794, "grad_norm": 10.981534315380854, "learning_rate": 2.0652693249866328e-07, "logits/chosen": -3.03125, "logits/rejected": -2.859375, "logps/chosen": -748.0, "logps/rejected": -1120.0, "loss": 0.1616, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.875, "rewards/margins": 4.03125, "rewards/rejected": -9.9375, "step": 9390 }, { "epoch": 1.2014186209981468, "grad_norm": 8.495947332387841, "learning_rate": 2.059778453739637e-07, "logits/chosen": -3.0, "logits/rejected": -2.671875, "logps/chosen": -708.0, "logps/rejected": -1072.0, "loss": 0.1483, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.5625, "rewards/margins": 3.71875, "rewards/rejected": -9.25, "step": 9400 }, { "epoch": 1.2026966579334142, "grad_norm": 10.158284185990194, "learning_rate": 2.0542897731805013e-07, "logits/chosen": -3.078125, "logits/rejected": -2.90625, "logps/chosen": -716.0, "logps/rejected": -1104.0, "loss": 0.1595, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.59375, "rewards/margins": 3.953125, "rewards/rejected": -9.5625, "step": 9410 }, { "epoch": 1.2039746948686818, "grad_norm": 9.832753006390973, "learning_rate": 2.0488033106227138e-07, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1120.0, "loss": 0.1718, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.53125, "rewards/margins": 4.15625, "rewards/rejected": -9.6875, "step": 9420 }, { "epoch": 1.2052527318039492, "grad_norm": 16.372998096961464, "learning_rate": 2.0433190933687246e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -716.0, "logps/rejected": -1080.0, "loss": 0.181, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.46875, "rewards/margins": 3.890625, "rewards/rejected": -9.375, "step": 9430 }, { "epoch": 1.2065307687392166, "grad_norm": 11.30856271193926, "learning_rate": 2.0378371487098115e-07, "logits/chosen": -2.8125, "logits/rejected": -2.6875, "logps/chosen": -700.0, "logps/rejected": -1072.0, "loss": 0.1747, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.0, "rewards/rejected": -9.3125, "step": 9440 }, { "epoch": 1.207808805674484, "grad_norm": 9.448324520200229, "learning_rate": 2.0323575039259419e-07, "logits/chosen": -2.953125, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1064.0, "loss": 0.1657, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.375, "rewards/margins": 3.8125, "rewards/rejected": -9.1875, "step": 9450 }, { "epoch": 1.2090868426097514, "grad_norm": 8.391376110479527, "learning_rate": 2.0268801862856383e-07, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.163, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.5625, "rewards/rejected": -9.75, "step": 9460 }, { "epoch": 1.2103648795450188, "grad_norm": 11.48008422005276, "learning_rate": 2.0214052230458434e-07, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1579, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.28125, "rewards/rejected": -9.6875, "step": 9470 }, { "epoch": 1.2116429164802862, "grad_norm": 9.857979206562652, "learning_rate": 2.015932641451783e-07, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -688.0, "logps/rejected": -1072.0, "loss": 0.1567, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.09375, "rewards/margins": 4.09375, "rewards/rejected": -9.1875, "step": 9480 }, { "epoch": 1.2129209534155536, "grad_norm": 13.596627334327412, "learning_rate": 2.010462468736831e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -640.0, "logps/rejected": -1004.0, "loss": 0.1696, "rewards/accuracies": 0.9375, "rewards/chosen": -4.8125, "rewards/margins": 3.953125, "rewards/rejected": -8.75, "step": 9490 }, { "epoch": 1.214198990350821, "grad_norm": 8.00173857521025, "learning_rate": 2.004994732122376e-07, "logits/chosen": -3.109375, "logits/rejected": -2.84375, "logps/chosen": -648.0, "logps/rejected": -1048.0, "loss": 0.158, "rewards/accuracies": 0.9375, "rewards/chosen": -4.8125, "rewards/margins": 4.3125, "rewards/rejected": -9.125, "step": 9500 }, { "epoch": 1.2154770272860886, "grad_norm": 13.184302458769404, "learning_rate": 1.9995294588176787e-07, "logits/chosen": -3.0625, "logits/rejected": -2.703125, "logps/chosen": -664.0, "logps/rejected": -1064.0, "loss": 0.1756, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0625, "rewards/margins": 4.15625, "rewards/rejected": -9.25, "step": 9510 }, { "epoch": 1.216755064221356, "grad_norm": 7.550889466931553, "learning_rate": 1.9940666760197469e-07, "logits/chosen": -3.046875, "logits/rejected": -2.71875, "logps/chosen": -684.0, "logps/rejected": -1120.0, "loss": 0.153, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.1875, "rewards/margins": 4.4375, "rewards/rejected": -9.625, "step": 9520 }, { "epoch": 1.2180331011566234, "grad_norm": 9.583242395878697, "learning_rate": 1.9886064109131917e-07, "logits/chosen": -3.0625, "logits/rejected": -2.84375, "logps/chosen": -656.0, "logps/rejected": -1024.0, "loss": 0.1665, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.84375, "rewards/margins": 3.921875, "rewards/rejected": -8.75, "step": 9530 }, { "epoch": 1.2193111380918908, "grad_norm": 8.254257411573414, "learning_rate": 1.9831486906700966e-07, "logits/chosen": -2.921875, "logits/rejected": -2.640625, "logps/chosen": -636.0, "logps/rejected": -1024.0, "loss": 0.1581, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.75, "rewards/margins": 4.09375, "rewards/rejected": -8.875, "step": 9540 }, { "epoch": 1.2205891750271582, "grad_norm": 10.525671293076426, "learning_rate": 1.9776935424498812e-07, "logits/chosen": -3.09375, "logits/rejected": -2.796875, "logps/chosen": -652.0, "logps/rejected": -1096.0, "loss": 0.1593, "rewards/accuracies": 0.96875, "rewards/chosen": -4.875, "rewards/margins": 4.78125, "rewards/rejected": -9.6875, "step": 9550 }, { "epoch": 1.2218672119624256, "grad_norm": 12.569317958096246, "learning_rate": 1.9722409933991647e-07, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -644.0, "logps/rejected": -1080.0, "loss": 0.16, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.9375, "rewards/margins": 4.4375, "rewards/rejected": -9.375, "step": 9560 }, { "epoch": 1.223145248897693, "grad_norm": 8.083359903906649, "learning_rate": 1.9667910706516336e-07, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -628.0, "logps/rejected": -1096.0, "loss": 0.1593, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.84375, "rewards/margins": 4.59375, "rewards/rejected": -9.4375, "step": 9570 }, { "epoch": 1.2244232858329607, "grad_norm": 9.861583608780453, "learning_rate": 1.9613438013279028e-07, "logits/chosen": -3.0, "logits/rejected": -2.59375, "logps/chosen": -664.0, "logps/rejected": -1096.0, "loss": 0.1597, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.03125, "rewards/margins": 4.59375, "rewards/rejected": -9.625, "step": 9580 }, { "epoch": 1.225701322768228, "grad_norm": 8.484208509238131, "learning_rate": 1.955899212535385e-07, "logits/chosen": -2.984375, "logits/rejected": -2.65625, "logps/chosen": -708.0, "logps/rejected": -1152.0, "loss": 0.1682, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.4375, "rewards/margins": 4.46875, "rewards/rejected": -9.9375, "step": 9590 }, { "epoch": 1.2269793597034955, "grad_norm": 8.632396255358831, "learning_rate": 1.950457331368153e-07, "logits/chosen": -2.84375, "logits/rejected": -2.703125, "logps/chosen": -680.0, "logps/rejected": -1064.0, "loss": 0.1537, "rewards/accuracies": 0.9375, "rewards/chosen": -5.25, "rewards/margins": 4.03125, "rewards/rejected": -9.25, "step": 9600 }, { "epoch": 1.2282573966387629, "grad_norm": 12.19240726948117, "learning_rate": 1.945018184906805e-07, "logits/chosen": -2.90625, "logits/rejected": -2.59375, "logps/chosen": -672.0, "logps/rejected": -1032.0, "loss": 0.1755, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 3.890625, "rewards/rejected": -9.0, "step": 9610 }, { "epoch": 1.2295354335740303, "grad_norm": 9.047734921539412, "learning_rate": 1.9395818002183316e-07, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -672.0, "logps/rejected": -1040.0, "loss": 0.1563, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.28125, "rewards/margins": 3.828125, "rewards/rejected": -9.125, "step": 9620 }, { "epoch": 1.2308134705092977, "grad_norm": 11.425663980736191, "learning_rate": 1.934148204355978e-07, "logits/chosen": -2.953125, "logits/rejected": -2.515625, "logps/chosen": -700.0, "logps/rejected": -1088.0, "loss": 0.1648, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.375, "rewards/margins": 4.0625, "rewards/rejected": -9.4375, "step": 9630 }, { "epoch": 1.232091507444565, "grad_norm": 8.965728172000757, "learning_rate": 1.9287174243591134e-07, "logits/chosen": -2.8125, "logits/rejected": -2.546875, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1582, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.125, "rewards/rejected": -9.5, "step": 9640 }, { "epoch": 1.2333695443798325, "grad_norm": 8.57673320822646, "learning_rate": 1.9232894872530928e-07, "logits/chosen": -2.84375, "logits/rejected": -2.59375, "logps/chosen": -708.0, "logps/rejected": -1112.0, "loss": 0.188, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.375, "rewards/margins": 4.15625, "rewards/rejected": -9.5, "step": 9650 }, { "epoch": 1.2346475813150999, "grad_norm": 11.365561069929564, "learning_rate": 1.9178644200491248e-07, "logits/chosen": -2.796875, "logits/rejected": -2.75, "logps/chosen": -728.0, "logps/rejected": -1104.0, "loss": 0.1525, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.21875, "rewards/rejected": -9.75, "step": 9660 }, { "epoch": 1.2359256182503675, "grad_norm": 12.077874161518164, "learning_rate": 1.9124422497441366e-07, "logits/chosen": -2.984375, "logits/rejected": -2.84375, "logps/chosen": -688.0, "logps/rejected": -1080.0, "loss": 0.1489, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.25, "rewards/margins": 4.25, "rewards/rejected": -9.5, "step": 9670 }, { "epoch": 1.237203655185635, "grad_norm": 11.7683730356947, "learning_rate": 1.9070230033206385e-07, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -732.0, "logps/rejected": -1096.0, "loss": 0.1536, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.8125, "rewards/margins": 3.953125, "rewards/rejected": -9.75, "step": 9680 }, { "epoch": 1.2384816921209023, "grad_norm": 9.769742084109476, "learning_rate": 1.9016067077465912e-07, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -716.0, "logps/rejected": -1176.0, "loss": 0.1748, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.75, "rewards/margins": 4.625, "rewards/rejected": -10.375, "step": 9690 }, { "epoch": 1.2397597290561697, "grad_norm": 10.711727609047516, "learning_rate": 1.896193389975271e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1578, "rewards/accuracies": 0.9375, "rewards/chosen": -5.65625, "rewards/margins": 4.21875, "rewards/rejected": -9.875, "step": 9700 }, { "epoch": 1.2410377659914371, "grad_norm": 8.709962769552265, "learning_rate": 1.890783076945136e-07, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -752.0, "logps/rejected": -1104.0, "loss": 0.1737, "rewards/accuracies": 0.90625, "rewards/chosen": -6.0, "rewards/margins": 3.671875, "rewards/rejected": -9.6875, "step": 9710 }, { "epoch": 1.2423158029267045, "grad_norm": 15.964742767418263, "learning_rate": 1.8853757955796924e-07, "logits/chosen": -3.015625, "logits/rejected": -2.875, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1666, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.59375, "rewards/margins": 4.4375, "rewards/rejected": -10.0, "step": 9720 }, { "epoch": 1.243593839861972, "grad_norm": 11.204694589329842, "learning_rate": 1.8799715727873565e-07, "logits/chosen": -3.03125, "logits/rejected": -2.671875, "logps/chosen": -732.0, "logps/rejected": -1128.0, "loss": 0.1752, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.75, "rewards/margins": 4.1875, "rewards/rejected": -9.9375, "step": 9730 }, { "epoch": 1.2448718767972395, "grad_norm": 13.582403295228488, "learning_rate": 1.8745704354613278e-07, "logits/chosen": -2.9375, "logits/rejected": -2.734375, "logps/chosen": -736.0, "logps/rejected": -1136.0, "loss": 0.1652, "rewards/accuracies": 0.9375, "rewards/chosen": -5.75, "rewards/margins": 4.125, "rewards/rejected": -9.875, "step": 9740 }, { "epoch": 1.246149913732507, "grad_norm": 9.514114838209917, "learning_rate": 1.8691724104794493e-07, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -744.0, "logps/rejected": -1168.0, "loss": 0.1769, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.84375, "rewards/margins": 4.3125, "rewards/rejected": -10.1875, "step": 9750 }, { "epoch": 1.2474279506677743, "grad_norm": 10.536194154771247, "learning_rate": 1.8637775247040777e-07, "logits/chosen": -2.921875, "logits/rejected": -2.828125, "logps/chosen": -776.0, "logps/rejected": -1128.0, "loss": 0.1832, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -6.03125, "rewards/margins": 3.96875, "rewards/rejected": -10.0, "step": 9760 }, { "epoch": 1.2487059876030417, "grad_norm": 13.45511967286933, "learning_rate": 1.8583858049819464e-07, "logits/chosen": -2.875, "logits/rejected": -2.796875, "logps/chosen": -760.0, "logps/rejected": -1096.0, "loss": 0.1734, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.9375, "rewards/margins": 3.671875, "rewards/rejected": -9.625, "step": 9770 }, { "epoch": 1.2499840245383091, "grad_norm": 13.317071101199458, "learning_rate": 1.8529972781440345e-07, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -744.0, "logps/rejected": -1160.0, "loss": 0.17, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.875, "rewards/margins": 4.1875, "rewards/rejected": -10.0625, "step": 9780 }, { "epoch": 1.2512620614735765, "grad_norm": 10.23423676486201, "learning_rate": 1.8476119710054323e-07, "logits/chosen": -2.84375, "logits/rejected": -2.640625, "logps/chosen": -728.0, "logps/rejected": -1064.0, "loss": 0.166, "rewards/accuracies": 0.875, "rewards/chosen": -5.78125, "rewards/margins": 3.46875, "rewards/rejected": -9.25, "step": 9790 }, { "epoch": 1.252540098408844, "grad_norm": 10.124955326944248, "learning_rate": 1.8422299103652063e-07, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -724.0, "logps/rejected": -1160.0, "loss": 0.1892, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.875, "rewards/margins": 4.3125, "rewards/rejected": -10.1875, "step": 9800 }, { "epoch": 1.2538181353441114, "grad_norm": 8.042244637589551, "learning_rate": 1.8368511230062695e-07, "logits/chosen": -2.90625, "logits/rejected": -2.734375, "logps/chosen": -760.0, "logps/rejected": -1144.0, "loss": 0.1708, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -6.09375, "rewards/margins": 3.875, "rewards/rejected": -9.9375, "step": 9810 }, { "epoch": 1.2550961722793788, "grad_norm": 9.73056988456358, "learning_rate": 1.8314756356952438e-07, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -752.0, "logps/rejected": -1152.0, "loss": 0.1722, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.96875, "rewards/margins": 4.0625, "rewards/rejected": -10.0625, "step": 9820 }, { "epoch": 1.2563742092146464, "grad_norm": 9.299217725085608, "learning_rate": 1.8261034751823316e-07, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -768.0, "logps/rejected": -1160.0, "loss": 0.161, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -6.09375, "rewards/margins": 4.0625, "rewards/rejected": -10.125, "step": 9830 }, { "epoch": 1.2576522461499138, "grad_norm": 12.508446663068236, "learning_rate": 1.820734668201177e-07, "logits/chosen": -3.0625, "logits/rejected": -2.78125, "logps/chosen": -772.0, "logps/rejected": -1216.0, "loss": 0.1526, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.9375, "rewards/margins": 4.65625, "rewards/rejected": -10.625, "step": 9840 }, { "epoch": 1.2589302830851812, "grad_norm": 12.023850516348142, "learning_rate": 1.815369241468738e-07, "logits/chosen": -2.96875, "logits/rejected": -2.78125, "logps/chosen": -736.0, "logps/rejected": -1136.0, "loss": 0.1774, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.71875, "rewards/margins": 4.21875, "rewards/rejected": -9.9375, "step": 9850 }, { "epoch": 1.2602083200204486, "grad_norm": 14.490328800702788, "learning_rate": 1.8100072216851508e-07, "logits/chosen": -3.1875, "logits/rejected": -2.828125, "logps/chosen": -772.0, "logps/rejected": -1136.0, "loss": 0.175, "rewards/accuracies": 0.875, "rewards/chosen": -6.0625, "rewards/margins": 3.734375, "rewards/rejected": -9.8125, "step": 9860 }, { "epoch": 1.261486356955716, "grad_norm": 10.975165721589365, "learning_rate": 1.8046486355335972e-07, "logits/chosen": -3.109375, "logits/rejected": -2.8125, "logps/chosen": -732.0, "logps/rejected": -1168.0, "loss": 0.1568, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.5, "rewards/rejected": -10.125, "step": 9870 }, { "epoch": 1.2627643938909834, "grad_norm": 9.49015858824327, "learning_rate": 1.799293509680173e-07, "logits/chosen": -3.046875, "logits/rejected": -2.765625, "logps/chosen": -752.0, "logps/rejected": -1136.0, "loss": 0.1659, "rewards/accuracies": 0.875, "rewards/chosen": -5.96875, "rewards/margins": 3.9375, "rewards/rejected": -9.875, "step": 9880 }, { "epoch": 1.264042430826251, "grad_norm": 9.126750240492946, "learning_rate": 1.793941870773753e-07, "logits/chosen": -2.96875, "logits/rejected": -2.875, "logps/chosen": -732.0, "logps/rejected": -1104.0, "loss": 0.1575, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.625, "rewards/margins": 3.953125, "rewards/rejected": -9.5625, "step": 9890 }, { "epoch": 1.2653204677615184, "grad_norm": 8.488616696424357, "learning_rate": 1.7885937454458605e-07, "logits/chosen": -2.953125, "logits/rejected": -2.875, "logps/chosen": -708.0, "logps/rejected": -1072.0, "loss": 0.1381, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.0, "rewards/rejected": -9.4375, "step": 9900 }, { "epoch": 1.2665985046967858, "grad_norm": 9.060288256850551, "learning_rate": 1.7832491603105346e-07, "logits/chosen": -2.96875, "logits/rejected": -2.859375, "logps/chosen": -740.0, "logps/rejected": -1160.0, "loss": 0.1625, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.78125, "rewards/margins": 4.25, "rewards/rejected": -10.0, "step": 9910 }, { "epoch": 1.2678765416320532, "grad_norm": 13.196118121665412, "learning_rate": 1.7779081419641963e-07, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -728.0, "logps/rejected": -1176.0, "loss": 0.1741, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.84375, "rewards/margins": 4.40625, "rewards/rejected": -10.25, "step": 9920 }, { "epoch": 1.2691545785673206, "grad_norm": 16.944706503313817, "learning_rate": 1.772570716985519e-07, "logits/chosen": -2.921875, "logits/rejected": -2.765625, "logps/chosen": -744.0, "logps/rejected": -1112.0, "loss": 0.176, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.84375, "rewards/margins": 3.875, "rewards/rejected": -9.75, "step": 9930 }, { "epoch": 1.270432615502588, "grad_norm": 14.012718602248546, "learning_rate": 1.7672369119352903e-07, "logits/chosen": -2.859375, "logits/rejected": -2.625, "logps/chosen": -692.0, "logps/rejected": -1096.0, "loss": 0.1807, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.125, "rewards/rejected": -9.4375, "step": 9940 }, { "epoch": 1.2717106524378554, "grad_norm": 13.799438014368707, "learning_rate": 1.7619067533562887e-07, "logits/chosen": -2.84375, "logits/rejected": -2.65625, "logps/chosen": -704.0, "logps/rejected": -1112.0, "loss": 0.1762, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.0625, "rewards/rejected": -9.5, "step": 9950 }, { "epoch": 1.2729886893731228, "grad_norm": 9.789102158125887, "learning_rate": 1.7565802677731423e-07, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -736.0, "logps/rejected": -1104.0, "loss": 0.1588, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.625, "rewards/margins": 3.9375, "rewards/rejected": -9.5625, "step": 9960 }, { "epoch": 1.2742667263083902, "grad_norm": 11.670905648113973, "learning_rate": 1.7512574816922044e-07, "logits/chosen": -2.921875, "logits/rejected": -2.6875, "logps/chosen": -728.0, "logps/rejected": -1112.0, "loss": 0.1619, "rewards/accuracies": 0.90625, "rewards/chosen": -5.40625, "rewards/margins": 4.15625, "rewards/rejected": -9.5625, "step": 9970 }, { "epoch": 1.2755447632436576, "grad_norm": 10.907827426264625, "learning_rate": 1.745938421601417e-07, "logits/chosen": -2.859375, "logits/rejected": -2.71875, "logps/chosen": -680.0, "logps/rejected": -1064.0, "loss": 0.1774, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.125, "rewards/margins": 4.125, "rewards/rejected": -9.25, "step": 9980 }, { "epoch": 1.2768228001789252, "grad_norm": 9.678391040150258, "learning_rate": 1.7406231139701806e-07, "logits/chosen": -2.796875, "logits/rejected": -2.640625, "logps/chosen": -688.0, "logps/rejected": -1144.0, "loss": 0.173, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.59375, "rewards/rejected": -9.875, "step": 9990 }, { "epoch": 1.2781008371141926, "grad_norm": 13.97763085395111, "learning_rate": 1.735311585249222e-07, "logits/chosen": -2.875, "logits/rejected": -2.796875, "logps/chosen": -736.0, "logps/rejected": -1088.0, "loss": 0.1801, "rewards/accuracies": 0.90625, "rewards/chosen": -5.625, "rewards/margins": 3.9375, "rewards/rejected": -9.5625, "step": 10000 }, { "epoch": 1.2781008371141926, "eval_logits/chosen": -2.984375, "eval_logits/rejected": -2.765625, "eval_logps/chosen": -720.0, "eval_logps/rejected": -1080.0, "eval_loss": 0.23926551640033722, "eval_rewards/accuracies": 0.8974654674530029, "eval_rewards/chosen": -5.625, "eval_rewards/margins": 3.78125, "eval_rewards/rejected": -9.375, "eval_runtime": 1884.4699, "eval_samples_per_second": 29.454, "eval_steps_per_second": 0.461, "step": 10000 }, { "epoch": 1.27937887404946, "grad_norm": 11.61678325539143, "learning_rate": 1.7300038618704626e-07, "logits/chosen": -2.921875, "logits/rejected": -2.890625, "logps/chosen": -692.0, "logps/rejected": -1072.0, "loss": 0.1605, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.0625, "rewards/rejected": -9.375, "step": 10010 }, { "epoch": 1.2806569109847274, "grad_norm": 7.438166564405569, "learning_rate": 1.7246999702468884e-07, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -668.0, "logps/rejected": -1072.0, "loss": 0.1613, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.1875, "rewards/rejected": -9.375, "step": 10020 }, { "epoch": 1.2819349479199948, "grad_norm": 7.976608253366539, "learning_rate": 1.7193999367724156e-07, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1434, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.375, "rewards/rejected": -10.0, "step": 10030 }, { "epoch": 1.2832129848552623, "grad_norm": 12.29189907278104, "learning_rate": 1.7141037878217624e-07, "logits/chosen": -3.09375, "logits/rejected": -2.78125, "logps/chosen": -688.0, "logps/rejected": -1120.0, "loss": 0.1792, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 10040 }, { "epoch": 1.2844910217905299, "grad_norm": 8.263784179594328, "learning_rate": 1.7088115497503163e-07, "logits/chosen": -2.96875, "logits/rejected": -2.796875, "logps/chosen": -684.0, "logps/rejected": -1096.0, "loss": 0.1529, "rewards/accuracies": 0.9375, "rewards/chosen": -5.25, "rewards/margins": 4.15625, "rewards/rejected": -9.4375, "step": 10050 }, { "epoch": 1.2857690587257973, "grad_norm": 9.788017604997307, "learning_rate": 1.7035232488940013e-07, "logits/chosen": -3.078125, "logits/rejected": -2.9375, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.1743, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 10060 }, { "epoch": 1.2870470956610647, "grad_norm": 12.141864032176112, "learning_rate": 1.6982389115691502e-07, "logits/chosen": -3.109375, "logits/rejected": -2.8125, "logps/chosen": -688.0, "logps/rejected": -1112.0, "loss": 0.1547, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.4375, "rewards/rejected": -9.75, "step": 10070 }, { "epoch": 1.288325132596332, "grad_norm": 8.15304150250457, "learning_rate": 1.6929585640723715e-07, "logits/chosen": -3.125, "logits/rejected": -2.921875, "logps/chosen": -684.0, "logps/rejected": -1120.0, "loss": 0.141, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.21875, "rewards/margins": 4.5625, "rewards/rejected": -9.75, "step": 10080 }, { "epoch": 1.2896031695315995, "grad_norm": 12.365972275709149, "learning_rate": 1.687682232680419e-07, "logits/chosen": -3.109375, "logits/rejected": -2.90625, "logps/chosen": -692.0, "logps/rejected": -1080.0, "loss": 0.168, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.40625, "rewards/margins": 3.96875, "rewards/rejected": -9.375, "step": 10090 }, { "epoch": 1.2908812064668669, "grad_norm": 11.148985181311243, "learning_rate": 1.682409943650061e-07, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -668.0, "logps/rejected": -1096.0, "loss": 0.147, "rewards/accuracies": 0.9375, "rewards/chosen": -5.28125, "rewards/margins": 4.28125, "rewards/rejected": -9.5625, "step": 10100 }, { "epoch": 1.2921592434021343, "grad_norm": 9.910706559244055, "learning_rate": 1.6771417232179491e-07, "logits/chosen": -3.078125, "logits/rejected": -2.828125, "logps/chosen": -644.0, "logps/rejected": -1048.0, "loss": 0.1705, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.84375, "rewards/margins": 4.25, "rewards/rejected": -9.125, "step": 10110 }, { "epoch": 1.2934372803374017, "grad_norm": 14.49757489823456, "learning_rate": 1.6718775976004896e-07, "logits/chosen": -2.984375, "logits/rejected": -2.625, "logps/chosen": -656.0, "logps/rejected": -1048.0, "loss": 0.1625, "rewards/accuracies": 0.9375, "rewards/chosen": -4.84375, "rewards/margins": 4.28125, "rewards/rejected": -9.125, "step": 10120 }, { "epoch": 1.294715317272669, "grad_norm": 8.542572240275735, "learning_rate": 1.66661759299371e-07, "logits/chosen": -2.859375, "logits/rejected": -2.671875, "logps/chosen": -688.0, "logps/rejected": -1104.0, "loss": 0.1454, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.21875, "rewards/rejected": -9.5, "step": 10130 }, { "epoch": 1.2959933542079365, "grad_norm": 8.145781278602318, "learning_rate": 1.6613617355731323e-07, "logits/chosen": -2.859375, "logits/rejected": -2.859375, "logps/chosen": -692.0, "logps/rejected": -1064.0, "loss": 0.1521, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.25, "rewards/margins": 3.9375, "rewards/rejected": -9.1875, "step": 10140 }, { "epoch": 1.2972713911432041, "grad_norm": 13.633869435963755, "learning_rate": 1.6561100514936384e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1618, "rewards/accuracies": 0.9375, "rewards/chosen": -5.34375, "rewards/margins": 4.03125, "rewards/rejected": -9.375, "step": 10150 }, { "epoch": 1.2985494280784715, "grad_norm": 12.090786431572226, "learning_rate": 1.650862566889343e-07, "logits/chosen": -2.953125, "logits/rejected": -2.734375, "logps/chosen": -680.0, "logps/rejected": -1168.0, "loss": 0.1743, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.15625, "rewards/margins": 5.03125, "rewards/rejected": -10.1875, "step": 10160 }, { "epoch": 1.299827465013739, "grad_norm": 10.97838475586465, "learning_rate": 1.6456193078734648e-07, "logits/chosen": -2.828125, "logits/rejected": -2.671875, "logps/chosen": -668.0, "logps/rejected": -1104.0, "loss": 0.1583, "rewards/accuracies": 0.9375, "rewards/chosen": -5.125, "rewards/margins": 4.53125, "rewards/rejected": -9.625, "step": 10170 }, { "epoch": 1.3011055019490063, "grad_norm": 9.673857765098305, "learning_rate": 1.6403803005381913e-07, "logits/chosen": -2.796875, "logits/rejected": -2.6875, "logps/chosen": -700.0, "logps/rejected": -1080.0, "loss": 0.1574, "rewards/accuracies": 0.9375, "rewards/chosen": -5.15625, "rewards/margins": 4.15625, "rewards/rejected": -9.3125, "step": 10180 }, { "epoch": 1.3023835388842737, "grad_norm": 12.851507523488168, "learning_rate": 1.635145570954555e-07, "logits/chosen": -2.984375, "logits/rejected": -2.71875, "logps/chosen": -708.0, "logps/rejected": -1104.0, "loss": 0.1687, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.625, "rewards/margins": 4.03125, "rewards/rejected": -9.6875, "step": 10190 }, { "epoch": 1.3036615758195411, "grad_norm": 11.353888250383354, "learning_rate": 1.6299151451722991e-07, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -768.0, "logps/rejected": -1152.0, "loss": 0.1637, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -6.0625, "rewards/margins": 4.1875, "rewards/rejected": -10.25, "step": 10200 }, { "epoch": 1.3049396127548087, "grad_norm": 9.29453192994112, "learning_rate": 1.6246890492197516e-07, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1136.0, "loss": 0.1554, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.375, "rewards/rejected": -9.9375, "step": 10210 }, { "epoch": 1.3062176496900761, "grad_norm": 11.056729555742567, "learning_rate": 1.619467309103692e-07, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.148, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.71875, "rewards/margins": 4.0625, "rewards/rejected": -9.75, "step": 10220 }, { "epoch": 1.3074956866253435, "grad_norm": 9.619998922145582, "learning_rate": 1.614249950809224e-07, "logits/chosen": -3.0, "logits/rejected": -2.65625, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1532, "rewards/accuracies": 0.9375, "rewards/chosen": -5.15625, "rewards/margins": 4.59375, "rewards/rejected": -9.75, "step": 10230 }, { "epoch": 1.308773723560611, "grad_norm": 13.306882835676573, "learning_rate": 1.6090370002996458e-07, "logits/chosen": -3.03125, "logits/rejected": -2.875, "logps/chosen": -704.0, "logps/rejected": -1120.0, "loss": 0.1372, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.3125, "rewards/rejected": -9.6875, "step": 10240 }, { "epoch": 1.3100517604958783, "grad_norm": 13.535705683821893, "learning_rate": 1.6038284835163227e-07, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1096.0, "loss": 0.152, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.1875, "rewards/rejected": -9.375, "step": 10250 }, { "epoch": 1.3113297974311457, "grad_norm": 9.303734802663968, "learning_rate": 1.5986244263785543e-07, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -676.0, "logps/rejected": -1136.0, "loss": 0.154, "rewards/accuracies": 0.9375, "rewards/chosen": -5.28125, "rewards/margins": 4.71875, "rewards/rejected": -10.0, "step": 10260 }, { "epoch": 1.3126078343664132, "grad_norm": 12.80378526578538, "learning_rate": 1.5934248547834493e-07, "logits/chosen": -2.96875, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1168.0, "loss": 0.1562, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.46875, "rewards/rejected": -9.9375, "step": 10270 }, { "epoch": 1.3138858713016806, "grad_norm": 9.8353883410435, "learning_rate": 1.5882297946057922e-07, "logits/chosen": -3.0, "logits/rejected": -2.796875, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1494, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.25, "rewards/margins": 4.5, "rewards/rejected": -9.75, "step": 10280 }, { "epoch": 1.315163908236948, "grad_norm": 10.334995200129333, "learning_rate": 1.58303927169792e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -676.0, "logps/rejected": -1104.0, "loss": 0.1507, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.1875, "rewards/margins": 4.40625, "rewards/rejected": -9.5625, "step": 10290 }, { "epoch": 1.3164419451722154, "grad_norm": 15.548721968057658, "learning_rate": 1.5778533118895893e-07, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -660.0, "logps/rejected": -1048.0, "loss": 0.1984, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.0625, "rewards/margins": 3.984375, "rewards/rejected": -9.0, "step": 10300 }, { "epoch": 1.317719982107483, "grad_norm": 9.965301691514663, "learning_rate": 1.5726719409878497e-07, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -628.0, "logps/rejected": -1032.0, "loss": 0.1522, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -4.75, "rewards/margins": 4.1875, "rewards/rejected": -8.9375, "step": 10310 }, { "epoch": 1.3189980190427504, "grad_norm": 11.883410302000634, "learning_rate": 1.5674951847769143e-07, "logits/chosen": -3.015625, "logits/rejected": -2.90625, "logps/chosen": -652.0, "logps/rejected": -1032.0, "loss": 0.1708, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.9375, "rewards/margins": 3.96875, "rewards/rejected": -8.875, "step": 10320 }, { "epoch": 1.3202760559780178, "grad_norm": 10.437698687232169, "learning_rate": 1.5623230690180328e-07, "logits/chosen": -2.9375, "logits/rejected": -2.578125, "logps/chosen": -656.0, "logps/rejected": -1072.0, "loss": 0.1715, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.96875, "rewards/margins": 4.28125, "rewards/rejected": -9.25, "step": 10330 }, { "epoch": 1.3215540929132852, "grad_norm": 11.893821802995003, "learning_rate": 1.5571556194493615e-07, "logits/chosen": -2.921875, "logits/rejected": -2.640625, "logps/chosen": -672.0, "logps/rejected": -1088.0, "loss": 0.1591, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.9375, "rewards/margins": 4.3125, "rewards/rejected": -9.25, "step": 10340 }, { "epoch": 1.3228321298485526, "grad_norm": 10.899103930979841, "learning_rate": 1.5519928617858373e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.162, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.34375, "rewards/rejected": -9.6875, "step": 10350 }, { "epoch": 1.32411016678382, "grad_norm": 11.345899735465693, "learning_rate": 1.5468348217190492e-07, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -692.0, "logps/rejected": -1096.0, "loss": 0.163, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.375, "rewards/rejected": -9.625, "step": 10360 }, { "epoch": 1.3253882037190876, "grad_norm": 12.30558614485336, "learning_rate": 1.541681524917106e-07, "logits/chosen": -2.8125, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1072.0, "loss": 0.1729, "rewards/accuracies": 0.90625, "rewards/chosen": -5.46875, "rewards/margins": 3.84375, "rewards/rejected": -9.3125, "step": 10370 }, { "epoch": 1.326666240654355, "grad_norm": 10.070533445591067, "learning_rate": 1.5365329970245173e-07, "logits/chosen": -2.859375, "logits/rejected": -2.59375, "logps/chosen": -696.0, "logps/rejected": -1104.0, "loss": 0.1567, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.28125, "rewards/rejected": -9.625, "step": 10380 }, { "epoch": 1.3279442775896224, "grad_norm": 9.555068680623833, "learning_rate": 1.5313892636620572e-07, "logits/chosen": -2.921875, "logits/rejected": -2.6875, "logps/chosen": -716.0, "logps/rejected": -1072.0, "loss": 0.1376, "rewards/accuracies": 0.96875, "rewards/chosen": -5.4375, "rewards/margins": 3.90625, "rewards/rejected": -9.3125, "step": 10390 }, { "epoch": 1.3292223145248898, "grad_norm": 12.8211655716973, "learning_rate": 1.5262503504266441e-07, "logits/chosen": -2.9375, "logits/rejected": -2.6875, "logps/chosen": -664.0, "logps/rejected": -1088.0, "loss": 0.1643, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.15625, "rewards/margins": 4.21875, "rewards/rejected": -9.375, "step": 10400 }, { "epoch": 1.3305003514601572, "grad_norm": 7.33947423183935, "learning_rate": 1.5211162828912073e-07, "logits/chosen": -2.921875, "logits/rejected": -2.703125, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.146, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 10410 }, { "epoch": 1.3317783883954246, "grad_norm": 16.553187562669738, "learning_rate": 1.5159870866045632e-07, "logits/chosen": -3.03125, "logits/rejected": -2.90625, "logps/chosen": -708.0, "logps/rejected": -1152.0, "loss": 0.1598, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5, "rewards/margins": 4.625, "rewards/rejected": -10.125, "step": 10420 }, { "epoch": 1.333056425330692, "grad_norm": 8.249018554748698, "learning_rate": 1.5108627870912877e-07, "logits/chosen": -3.046875, "logits/rejected": -2.9375, "logps/chosen": -696.0, "logps/rejected": -1144.0, "loss": 0.1599, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 4.75, "rewards/rejected": -9.9375, "step": 10430 }, { "epoch": 1.3343344622659594, "grad_norm": 11.885385205335053, "learning_rate": 1.5057434098515871e-07, "logits/chosen": -2.96875, "logits/rejected": -2.890625, "logps/chosen": -672.0, "logps/rejected": -1080.0, "loss": 0.1588, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.09375, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 10440 }, { "epoch": 1.3356124992012268, "grad_norm": 12.606879170463433, "learning_rate": 1.500628980361175e-07, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -724.0, "logps/rejected": -1112.0, "loss": 0.1643, "rewards/accuracies": 0.90625, "rewards/chosen": -5.6875, "rewards/margins": 3.96875, "rewards/rejected": -9.625, "step": 10450 }, { "epoch": 1.3368905361364942, "grad_norm": 8.510890461580383, "learning_rate": 1.4955195240711416e-07, "logits/chosen": -3.078125, "logits/rejected": -2.890625, "logps/chosen": -712.0, "logps/rejected": -1136.0, "loss": 0.1511, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.34375, "rewards/rejected": -9.8125, "step": 10460 }, { "epoch": 1.3381685730717618, "grad_norm": 13.307022774845727, "learning_rate": 1.49041506640783e-07, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -672.0, "logps/rejected": -1112.0, "loss": 0.1608, "rewards/accuracies": 0.9375, "rewards/chosen": -5.1875, "rewards/margins": 4.3125, "rewards/rejected": -9.5, "step": 10470 }, { "epoch": 1.3394466100070292, "grad_norm": 10.194898408199766, "learning_rate": 1.4853156327727083e-07, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -640.0, "logps/rejected": -1056.0, "loss": 0.1436, "rewards/accuracies": 0.96875, "rewards/chosen": -4.9375, "rewards/margins": 4.1875, "rewards/rejected": -9.125, "step": 10480 }, { "epoch": 1.3407246469422966, "grad_norm": 9.840893728306792, "learning_rate": 1.4802212485422423e-07, "logits/chosen": -3.078125, "logits/rejected": -2.6875, "logps/chosen": -684.0, "logps/rejected": -1088.0, "loss": 0.1665, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.3125, "rewards/margins": 4.09375, "rewards/rejected": -9.375, "step": 10490 }, { "epoch": 1.342002683877564, "grad_norm": 16.030708438907578, "learning_rate": 1.475131939067771e-07, "logits/chosen": -3.125, "logits/rejected": -2.90625, "logps/chosen": -668.0, "logps/rejected": -1096.0, "loss": 0.1865, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.15625, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 10500 }, { "epoch": 1.3432807208128315, "grad_norm": 14.6898022036467, "learning_rate": 1.4700477296753805e-07, "logits/chosen": -2.921875, "logits/rejected": -2.75, "logps/chosen": -640.0, "logps/rejected": -1040.0, "loss": 0.1643, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.8125, "rewards/margins": 4.0625, "rewards/rejected": -8.875, "step": 10510 }, { "epoch": 1.3445587577480989, "grad_norm": 7.640691836956104, "learning_rate": 1.4649686456657764e-07, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -652.0, "logps/rejected": -1056.0, "loss": 0.1655, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -4.9375, "rewards/margins": 4.125, "rewards/rejected": -9.0625, "step": 10520 }, { "epoch": 1.3458367946833665, "grad_norm": 10.471801000652848, "learning_rate": 1.4598947123141593e-07, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -648.0, "logps/rejected": -1056.0, "loss": 0.1324, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -4.90625, "rewards/margins": 4.3125, "rewards/rejected": -9.1875, "step": 10530 }, { "epoch": 1.3471148316186339, "grad_norm": 14.327591713140697, "learning_rate": 1.4548259548700958e-07, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1104.0, "loss": 0.1583, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.25, "rewards/margins": 4.34375, "rewards/rejected": -9.5625, "step": 10540 }, { "epoch": 1.3483928685539013, "grad_norm": 10.818716941448443, "learning_rate": 1.4497623985574003e-07, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -720.0, "logps/rejected": -1144.0, "loss": 0.1545, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.65625, "rewards/margins": 4.34375, "rewards/rejected": -10.0, "step": 10550 }, { "epoch": 1.3496709054891687, "grad_norm": 10.701973869348382, "learning_rate": 1.4447040685740011e-07, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -680.0, "logps/rejected": -1088.0, "loss": 0.1774, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.03125, "rewards/margins": 4.5, "rewards/rejected": -9.5, "step": 10560 }, { "epoch": 1.350948942424436, "grad_norm": 10.557227140466773, "learning_rate": 1.439650990091819e-07, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -680.0, "logps/rejected": -1088.0, "loss": 0.1641, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.28125, "rewards/rejected": -9.5625, "step": 10570 }, { "epoch": 1.3522269793597035, "grad_norm": 11.232062901600258, "learning_rate": 1.4346031882566422e-07, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -672.0, "logps/rejected": -1064.0, "loss": 0.1688, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.1875, "rewards/margins": 4.09375, "rewards/rejected": -9.25, "step": 10580 }, { "epoch": 1.3535050162949709, "grad_norm": 8.121003160184413, "learning_rate": 1.4295606881879994e-07, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -708.0, "logps/rejected": -1096.0, "loss": 0.1371, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.25, "rewards/margins": 4.0625, "rewards/rejected": -9.3125, "step": 10590 }, { "epoch": 1.3547830532302383, "grad_norm": 11.168731072310887, "learning_rate": 1.4245235149790393e-07, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -664.0, "logps/rejected": -1112.0, "loss": 0.1594, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -4.96875, "rewards/margins": 4.625, "rewards/rejected": -9.625, "step": 10600 }, { "epoch": 1.3560610901655057, "grad_norm": 10.801917621179639, "learning_rate": 1.4194916936963975e-07, "logits/chosen": -2.859375, "logits/rejected": -2.796875, "logps/chosen": -668.0, "logps/rejected": -1080.0, "loss": 0.166, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.03125, "rewards/margins": 4.25, "rewards/rejected": -9.3125, "step": 10610 }, { "epoch": 1.357339127100773, "grad_norm": 12.255284522095959, "learning_rate": 1.4144652493800813e-07, "logits/chosen": -3.015625, "logits/rejected": -2.6875, "logps/chosen": -640.0, "logps/rejected": -1040.0, "loss": 0.1662, "rewards/accuracies": 0.9375, "rewards/chosen": -4.875, "rewards/margins": 4.0, "rewards/rejected": -8.875, "step": 10620 }, { "epoch": 1.3586171640360407, "grad_norm": 9.63680109894397, "learning_rate": 1.409444207043337e-07, "logits/chosen": -3.046875, "logits/rejected": -2.90625, "logps/chosen": -672.0, "logps/rejected": -1056.0, "loss": 0.1636, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.15625, "rewards/margins": 4.09375, "rewards/rejected": -9.25, "step": 10630 }, { "epoch": 1.3598952009713081, "grad_norm": 11.406361911753756, "learning_rate": 1.40442859167253e-07, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1589, "rewards/accuracies": 0.9375, "rewards/chosen": -5.40625, "rewards/margins": 4.28125, "rewards/rejected": -9.6875, "step": 10640 }, { "epoch": 1.3611732379065755, "grad_norm": 8.724674918938067, "learning_rate": 1.3994184282270212e-07, "logits/chosen": -3.0625, "logits/rejected": -2.90625, "logps/chosen": -732.0, "logps/rejected": -1136.0, "loss": 0.166, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.8125, "rewards/margins": 4.125, "rewards/rejected": -9.9375, "step": 10650 }, { "epoch": 1.362451274841843, "grad_norm": 8.713189202090533, "learning_rate": 1.3944137416390367e-07, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -740.0, "logps/rejected": -1176.0, "loss": 0.1326, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.65625, "rewards/margins": 4.53125, "rewards/rejected": -10.1875, "step": 10660 }, { "epoch": 1.3637293117771103, "grad_norm": 11.092642328430658, "learning_rate": 1.3894145568135532e-07, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -732.0, "logps/rejected": -1080.0, "loss": 0.1563, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.875, "rewards/margins": 3.640625, "rewards/rejected": -9.5, "step": 10670 }, { "epoch": 1.3650073487123777, "grad_norm": 15.316488295385675, "learning_rate": 1.3844208986281647e-07, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1637, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.3125, "rewards/rejected": -9.6875, "step": 10680 }, { "epoch": 1.3662853856476453, "grad_norm": 11.215655709405532, "learning_rate": 1.379432791932963e-07, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -740.0, "logps/rejected": -1184.0, "loss": 0.1711, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.6875, "rewards/margins": 4.46875, "rewards/rejected": -10.1875, "step": 10690 }, { "epoch": 1.3675634225829127, "grad_norm": 8.734870719545054, "learning_rate": 1.3744502615504177e-07, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -736.0, "logps/rejected": -1152.0, "loss": 0.1546, "rewards/accuracies": 0.90625, "rewards/chosen": -5.75, "rewards/margins": 4.375, "rewards/rejected": -10.125, "step": 10700 }, { "epoch": 1.3688414595181801, "grad_norm": 10.872876644870878, "learning_rate": 1.3694733322752433e-07, "logits/chosen": -2.984375, "logits/rejected": -2.828125, "logps/chosen": -696.0, "logps/rejected": -1152.0, "loss": 0.1501, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 10710 }, { "epoch": 1.3701194964534475, "grad_norm": 11.29460016079079, "learning_rate": 1.3645020288742872e-07, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -708.0, "logps/rejected": -1104.0, "loss": 0.1671, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 10720 }, { "epoch": 1.371397533388715, "grad_norm": 13.2399522838317, "learning_rate": 1.3595363760863958e-07, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -668.0, "logps/rejected": -1056.0, "loss": 0.1558, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.15625, "rewards/margins": 4.0625, "rewards/rejected": -9.1875, "step": 10730 }, { "epoch": 1.3726755703239824, "grad_norm": 9.725638941764556, "learning_rate": 1.3545763986222973e-07, "logits/chosen": -2.859375, "logits/rejected": -2.8125, "logps/chosen": -652.0, "logps/rejected": -1080.0, "loss": 0.1628, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.09375, "rewards/margins": 4.21875, "rewards/rejected": -9.3125, "step": 10740 }, { "epoch": 1.3739536072592498, "grad_norm": 16.159984692053747, "learning_rate": 1.349622121164481e-07, "logits/chosen": -3.09375, "logits/rejected": -2.65625, "logps/chosen": -668.0, "logps/rejected": -1096.0, "loss": 0.1739, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.125, "rewards/margins": 4.40625, "rewards/rejected": -9.5, "step": 10750 }, { "epoch": 1.3752316441945172, "grad_norm": 10.45347181336293, "learning_rate": 1.344673568367065e-07, "logits/chosen": -2.9375, "logits/rejected": -2.71875, "logps/chosen": -700.0, "logps/rejected": -1128.0, "loss": 0.151, "rewards/accuracies": 0.90625, "rewards/chosen": -5.4375, "rewards/margins": 4.21875, "rewards/rejected": -9.625, "step": 10760 }, { "epoch": 1.3765096811297846, "grad_norm": 12.043752161415977, "learning_rate": 1.3397307648556861e-07, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -724.0, "logps/rejected": -1096.0, "loss": 0.1639, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.71875, "rewards/margins": 3.828125, "rewards/rejected": -9.5625, "step": 10770 }, { "epoch": 1.377787718065052, "grad_norm": 8.3067657319515, "learning_rate": 1.334793735227366e-07, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -720.0, "logps/rejected": -1176.0, "loss": 0.1579, "rewards/accuracies": 0.9375, "rewards/chosen": -5.65625, "rewards/margins": 4.65625, "rewards/rejected": -10.3125, "step": 10780 }, { "epoch": 1.3790657550003196, "grad_norm": 10.000745025884786, "learning_rate": 1.329862504050395e-07, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.145, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.34375, "rewards/rejected": -9.75, "step": 10790 }, { "epoch": 1.380343791935587, "grad_norm": 9.27281728403169, "learning_rate": 1.32493709586421e-07, "logits/chosen": -2.9375, "logits/rejected": -2.671875, "logps/chosen": -668.0, "logps/rejected": -1072.0, "loss": 0.1402, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.28125, "rewards/margins": 3.96875, "rewards/rejected": -9.25, "step": 10800 }, { "epoch": 1.3816218288708544, "grad_norm": 11.177095928758252, "learning_rate": 1.3200175351792676e-07, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -724.0, "logps/rejected": -1144.0, "loss": 0.159, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.375, "rewards/rejected": -9.6875, "step": 10810 }, { "epoch": 1.3828998658061218, "grad_norm": 13.81811898136344, "learning_rate": 1.3151038464769288e-07, "logits/chosen": -3.125, "logits/rejected": -2.84375, "logps/chosen": -720.0, "logps/rejected": -1120.0, "loss": 0.167, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.6875, "rewards/margins": 4.1875, "rewards/rejected": -9.875, "step": 10820 }, { "epoch": 1.3841779027413892, "grad_norm": 12.668241973578079, "learning_rate": 1.3101960542093306e-07, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -692.0, "logps/rejected": -1128.0, "loss": 0.1615, "rewards/accuracies": 0.9375, "rewards/chosen": -5.1875, "rewards/margins": 4.375, "rewards/rejected": -9.625, "step": 10830 }, { "epoch": 1.3854559396766566, "grad_norm": 14.666040788124336, "learning_rate": 1.3052941827992708e-07, "logits/chosen": -2.953125, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1056.0, "loss": 0.1744, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.65625, "rewards/margins": 3.578125, "rewards/rejected": -9.25, "step": 10840 }, { "epoch": 1.3867339766119242, "grad_norm": 8.587958353569526, "learning_rate": 1.3003982566400802e-07, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -700.0, "logps/rejected": -1136.0, "loss": 0.1293, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.375, "rewards/rejected": -9.75, "step": 10850 }, { "epoch": 1.3880120135471916, "grad_norm": 10.462369727921462, "learning_rate": 1.295508300095503e-07, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -724.0, "logps/rejected": -1176.0, "loss": 0.1488, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.65625, "rewards/rejected": -10.1875, "step": 10860 }, { "epoch": 1.389290050482459, "grad_norm": 7.445306803662397, "learning_rate": 1.2906243374995818e-07, "logits/chosen": -2.96875, "logits/rejected": -2.78125, "logps/chosen": -748.0, "logps/rejected": -1192.0, "loss": 0.164, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.8125, "rewards/margins": 4.53125, "rewards/rejected": -10.3125, "step": 10870 }, { "epoch": 1.3905680874177264, "grad_norm": 10.953212359779702, "learning_rate": 1.2857463931565256e-07, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -720.0, "logps/rejected": -1176.0, "loss": 0.1529, "rewards/accuracies": 0.96875, "rewards/chosen": -5.625, "rewards/margins": 4.65625, "rewards/rejected": -10.3125, "step": 10880 }, { "epoch": 1.3918461243529938, "grad_norm": 6.7514582868915225, "learning_rate": 1.280874491340599e-07, "logits/chosen": -3.109375, "logits/rejected": -2.75, "logps/chosen": -720.0, "logps/rejected": -1168.0, "loss": 0.1522, "rewards/accuracies": 0.9375, "rewards/chosen": -5.53125, "rewards/margins": 4.65625, "rewards/rejected": -10.1875, "step": 10890 }, { "epoch": 1.3931241612882612, "grad_norm": 14.749439493951504, "learning_rate": 1.2760086562959937e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -752.0, "logps/rejected": -1152.0, "loss": 0.1664, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.8125, "rewards/margins": 4.3125, "rewards/rejected": -10.125, "step": 10900 }, { "epoch": 1.3944021982235286, "grad_norm": 10.625720834052649, "learning_rate": 1.2711489122367124e-07, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -668.0, "logps/rejected": -1088.0, "loss": 0.1532, "rewards/accuracies": 0.96875, "rewards/chosen": -5.15625, "rewards/margins": 4.4375, "rewards/rejected": -9.5625, "step": 10910 }, { "epoch": 1.395680235158796, "grad_norm": 12.248756314421437, "learning_rate": 1.2662952833464485e-07, "logits/chosen": -2.96875, "logits/rejected": -2.6875, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1634, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.46875, "rewards/rejected": -9.6875, "step": 10920 }, { "epoch": 1.3969582720940634, "grad_norm": 9.829887044914619, "learning_rate": 1.2614477937784617e-07, "logits/chosen": -2.953125, "logits/rejected": -2.84375, "logps/chosen": -656.0, "logps/rejected": -1072.0, "loss": 0.1473, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.09375, "rewards/margins": 4.25, "rewards/rejected": -9.3125, "step": 10930 }, { "epoch": 1.3982363090293308, "grad_norm": 11.549751875940906, "learning_rate": 1.2566064676554628e-07, "logits/chosen": -3.09375, "logits/rejected": -2.890625, "logps/chosen": -660.0, "logps/rejected": -1080.0, "loss": 0.166, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 4.25, "rewards/rejected": -9.375, "step": 10940 }, { "epoch": 1.3995143459645984, "grad_norm": 10.503560647884871, "learning_rate": 1.2517713290694892e-07, "logits/chosen": -3.171875, "logits/rejected": -3.0, "logps/chosen": -692.0, "logps/rejected": -1104.0, "loss": 0.1522, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.28125, "rewards/margins": 4.4375, "rewards/rejected": -9.6875, "step": 10950 }, { "epoch": 1.4007923828998659, "grad_norm": 14.679390504503392, "learning_rate": 1.246942402081788e-07, "logits/chosen": -3.171875, "logits/rejected": -3.015625, "logps/chosen": -704.0, "logps/rejected": -1120.0, "loss": 0.1611, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.3125, "rewards/rejected": -9.75, "step": 10960 }, { "epoch": 1.4020704198351333, "grad_norm": 8.841059671351234, "learning_rate": 1.242119710722696e-07, "logits/chosen": -3.078125, "logits/rejected": -2.9375, "logps/chosen": -668.0, "logps/rejected": -1072.0, "loss": 0.1721, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.21875, "rewards/margins": 4.1875, "rewards/rejected": -9.4375, "step": 10970 }, { "epoch": 1.4033484567704007, "grad_norm": 11.857415726709224, "learning_rate": 1.2373032789915176e-07, "logits/chosen": -3.09375, "logits/rejected": -2.90625, "logps/chosen": -608.0, "logps/rejected": -1016.0, "loss": 0.1661, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -4.71875, "rewards/margins": 4.125, "rewards/rejected": -8.8125, "step": 10980 }, { "epoch": 1.404626493705668, "grad_norm": 9.710749089708392, "learning_rate": 1.2324931308564103e-07, "logits/chosen": -3.0625, "logits/rejected": -2.78125, "logps/chosen": -652.0, "logps/rejected": -1040.0, "loss": 0.1659, "rewards/accuracies": 0.90625, "rewards/chosen": -5.03125, "rewards/margins": 3.96875, "rewards/rejected": -9.0, "step": 10990 }, { "epoch": 1.4059045306409355, "grad_norm": 11.438152431601049, "learning_rate": 1.2276892902542596e-07, "logits/chosen": -2.84375, "logits/rejected": -2.796875, "logps/chosen": -744.0, "logps/rejected": -1120.0, "loss": 0.1669, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.3125, "rewards/rejected": -9.75, "step": 11000 }, { "epoch": 1.407182567576203, "grad_norm": 9.709697271547052, "learning_rate": 1.2228917810905631e-07, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -652.0, "logps/rejected": -1072.0, "loss": 0.1474, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -4.9375, "rewards/margins": 4.375, "rewards/rejected": -9.3125, "step": 11010 }, { "epoch": 1.4084606045114705, "grad_norm": 10.24605596244792, "learning_rate": 1.2181006272393115e-07, "logits/chosen": -2.984375, "logits/rejected": -2.65625, "logps/chosen": -648.0, "logps/rejected": -1072.0, "loss": 0.1763, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.9375, "rewards/margins": 4.28125, "rewards/rejected": -9.25, "step": 11020 }, { "epoch": 1.4097386414467379, "grad_norm": 12.25722659236921, "learning_rate": 1.2133158525428686e-07, "logits/chosen": -2.90625, "logits/rejected": -2.8125, "logps/chosen": -688.0, "logps/rejected": -1032.0, "loss": 0.1666, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.0625, "rewards/margins": 3.828125, "rewards/rejected": -8.875, "step": 11030 }, { "epoch": 1.4110166783820053, "grad_norm": 12.072026666485739, "learning_rate": 1.2085374808118558e-07, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -640.0, "logps/rejected": -1072.0, "loss": 0.1611, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -4.9375, "rewards/margins": 4.53125, "rewards/rejected": -9.5, "step": 11040 }, { "epoch": 1.4122947153172727, "grad_norm": 7.932169471982789, "learning_rate": 1.2037655358250282e-07, "logits/chosen": -3.09375, "logits/rejected": -2.78125, "logps/chosen": -696.0, "logps/rejected": -1104.0, "loss": 0.1501, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.25, "rewards/margins": 4.3125, "rewards/rejected": -9.5625, "step": 11050 }, { "epoch": 1.41357275225254, "grad_norm": 7.953752156821214, "learning_rate": 1.199000041329162e-07, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -664.0, "logps/rejected": -1080.0, "loss": 0.1624, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.15625, "rewards/margins": 4.21875, "rewards/rejected": -9.375, "step": 11060 }, { "epoch": 1.4148507891878075, "grad_norm": 13.426621520621833, "learning_rate": 1.1942410210389317e-07, "logits/chosen": -3.0625, "logits/rejected": -2.765625, "logps/chosen": -700.0, "logps/rejected": -1136.0, "loss": 0.1732, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 11070 }, { "epoch": 1.416128826123075, "grad_norm": 9.374559807282663, "learning_rate": 1.1894884986367934e-07, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -720.0, "logps/rejected": -1112.0, "loss": 0.1555, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.21875, "rewards/rejected": -9.8125, "step": 11080 }, { "epoch": 1.4174068630583423, "grad_norm": 9.973310613947785, "learning_rate": 1.184742497772871e-07, "logits/chosen": -2.984375, "logits/rejected": -2.875, "logps/chosen": -728.0, "logps/rejected": -1184.0, "loss": 0.1436, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.84375, "rewards/rejected": -10.3125, "step": 11090 }, { "epoch": 1.4186848999936097, "grad_norm": 13.250219412695905, "learning_rate": 1.1800030420648313e-07, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1619, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.53125, "rewards/margins": 4.1875, "rewards/rejected": -9.75, "step": 11100 }, { "epoch": 1.4199629369288773, "grad_norm": 13.773396859073182, "learning_rate": 1.1752701550977715e-07, "logits/chosen": -3.078125, "logits/rejected": -2.828125, "logps/chosen": -732.0, "logps/rejected": -1240.0, "loss": 0.1541, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.59375, "rewards/margins": 5.0625, "rewards/rejected": -10.6875, "step": 11110 }, { "epoch": 1.4212409738641447, "grad_norm": 8.993057070559015, "learning_rate": 1.1705438604241025e-07, "logits/chosen": -2.96875, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1160.0, "loss": 0.1575, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.59375, "rewards/rejected": -10.1875, "step": 11120 }, { "epoch": 1.4225190107994121, "grad_norm": 12.61693065803415, "learning_rate": 1.1658241815634259e-07, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -716.0, "logps/rejected": -1072.0, "loss": 0.1807, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.59375, "rewards/margins": 3.703125, "rewards/rejected": -9.25, "step": 11130 }, { "epoch": 1.4237970477346795, "grad_norm": 11.057436063689746, "learning_rate": 1.1611111420024255e-07, "logits/chosen": -3.0625, "logits/rejected": -2.765625, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.145, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.125, "rewards/rejected": -9.75, "step": 11140 }, { "epoch": 1.425075084669947, "grad_norm": 12.071787660113134, "learning_rate": 1.1564047651947423e-07, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -668.0, "logps/rejected": -1136.0, "loss": 0.1734, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.5625, "rewards/rejected": -9.75, "step": 11150 }, { "epoch": 1.4263531216052143, "grad_norm": 8.663142571001057, "learning_rate": 1.1517050745608611e-07, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1694, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5625, "rewards/margins": 3.8125, "rewards/rejected": -9.375, "step": 11160 }, { "epoch": 1.427631158540482, "grad_norm": 10.17682640212931, "learning_rate": 1.147012093487997e-07, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -664.0, "logps/rejected": -1080.0, "loss": 0.1556, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.09375, "rewards/margins": 4.28125, "rewards/rejected": -9.375, "step": 11170 }, { "epoch": 1.4289091954757493, "grad_norm": 9.483691786180756, "learning_rate": 1.1423258453299728e-07, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -688.0, "logps/rejected": -1088.0, "loss": 0.145, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.1875, "rewards/rejected": -9.5, "step": 11180 }, { "epoch": 1.4301872324110168, "grad_norm": 9.211669175543367, "learning_rate": 1.1376463534071093e-07, "logits/chosen": -3.109375, "logits/rejected": -2.953125, "logps/chosen": -676.0, "logps/rejected": -1072.0, "loss": 0.1559, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.1875, "rewards/margins": 4.0, "rewards/rejected": -9.1875, "step": 11190 }, { "epoch": 1.4314652693462842, "grad_norm": 12.093964271729028, "learning_rate": 1.1329736410061033e-07, "logits/chosen": -3.0625, "logits/rejected": -2.953125, "logps/chosen": -704.0, "logps/rejected": -1168.0, "loss": 0.1554, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.78125, "rewards/rejected": -10.1875, "step": 11200 }, { "epoch": 1.4327433062815516, "grad_norm": 9.29114146915148, "learning_rate": 1.128307731379915e-07, "logits/chosen": -3.15625, "logits/rejected": -2.859375, "logps/chosen": -700.0, "logps/rejected": -1104.0, "loss": 0.1694, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.1875, "rewards/rejected": -9.5, "step": 11210 }, { "epoch": 1.434021343216819, "grad_norm": 12.657395691854575, "learning_rate": 1.1236486477476551e-07, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -696.0, "logps/rejected": -1104.0, "loss": 0.1662, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.25, "rewards/rejected": -9.5625, "step": 11220 }, { "epoch": 1.4352993801520864, "grad_norm": 11.348547431972344, "learning_rate": 1.1189964132944588e-07, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -684.0, "logps/rejected": -1056.0, "loss": 0.1622, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 4.15625, "rewards/rejected": -9.3125, "step": 11230 }, { "epoch": 1.4365774170873538, "grad_norm": 8.516011381449768, "learning_rate": 1.1143510511713849e-07, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -648.0, "logps/rejected": -1064.0, "loss": 0.1633, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.125, "rewards/margins": 4.1875, "rewards/rejected": -9.3125, "step": 11240 }, { "epoch": 1.4378554540226212, "grad_norm": 12.710258105717188, "learning_rate": 1.1097125844952885e-07, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -672.0, "logps/rejected": -1072.0, "loss": 0.1456, "rewards/accuracies": 0.9375, "rewards/chosen": -5.1875, "rewards/margins": 4.28125, "rewards/rejected": -9.4375, "step": 11250 }, { "epoch": 1.4391334909578886, "grad_norm": 9.096742033397492, "learning_rate": 1.1050810363487139e-07, "logits/chosen": -3.078125, "logits/rejected": -2.828125, "logps/chosen": -680.0, "logps/rejected": -1136.0, "loss": 0.1644, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.28125, "rewards/margins": 4.65625, "rewards/rejected": -9.9375, "step": 11260 }, { "epoch": 1.4404115278931562, "grad_norm": 9.27229207508295, "learning_rate": 1.1004564297797741e-07, "logits/chosen": -3.21875, "logits/rejected": -2.953125, "logps/chosen": -680.0, "logps/rejected": -1120.0, "loss": 0.1396, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 4.65625, "rewards/rejected": -9.875, "step": 11270 }, { "epoch": 1.4416895648284236, "grad_norm": 14.009678507550513, "learning_rate": 1.0958387878020381e-07, "logits/chosen": -3.15625, "logits/rejected": -2.9375, "logps/chosen": -700.0, "logps/rejected": -1128.0, "loss": 0.1682, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.46875, "rewards/rejected": -9.8125, "step": 11280 }, { "epoch": 1.442967601763691, "grad_norm": 12.862568319486723, "learning_rate": 1.0912281333944204e-07, "logits/chosen": -3.1875, "logits/rejected": -2.9375, "logps/chosen": -724.0, "logps/rejected": -1128.0, "loss": 0.133, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.6875, "rewards/margins": 4.15625, "rewards/rejected": -9.875, "step": 11290 }, { "epoch": 1.4442456386989584, "grad_norm": 10.33161006641078, "learning_rate": 1.0866244895010584e-07, "logits/chosen": -3.09375, "logits/rejected": -2.8125, "logps/chosen": -724.0, "logps/rejected": -1168.0, "loss": 0.1567, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.59375, "rewards/rejected": -10.25, "step": 11300 }, { "epoch": 1.4455236756342258, "grad_norm": 14.720700918357485, "learning_rate": 1.082027879031207e-07, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -684.0, "logps/rejected": -1120.0, "loss": 0.1529, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.34375, "rewards/rejected": -9.625, "step": 11310 }, { "epoch": 1.4468017125694932, "grad_norm": 9.199043302635625, "learning_rate": 1.077438324859118e-07, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -656.0, "logps/rejected": -1112.0, "loss": 0.181, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.0, "rewards/margins": 4.5, "rewards/rejected": -9.5, "step": 11320 }, { "epoch": 1.4480797495047608, "grad_norm": 9.957779437846934, "learning_rate": 1.072855849823928e-07, "logits/chosen": -3.046875, "logits/rejected": -2.875, "logps/chosen": -676.0, "logps/rejected": -1056.0, "loss": 0.1457, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.15625, "rewards/margins": 4.0625, "rewards/rejected": -9.1875, "step": 11330 }, { "epoch": 1.4493577864400282, "grad_norm": 10.62616037496165, "learning_rate": 1.0682804767295486e-07, "logits/chosen": -3.078125, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1080.0, "loss": 0.1583, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.125, "rewards/margins": 4.3125, "rewards/rejected": -9.4375, "step": 11340 }, { "epoch": 1.4506358233752956, "grad_norm": 13.753631734488085, "learning_rate": 1.0637122283445466e-07, "logits/chosen": -2.96875, "logits/rejected": -2.84375, "logps/chosen": -720.0, "logps/rejected": -1128.0, "loss": 0.1532, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.4375, "rewards/rejected": -10.0, "step": 11350 }, { "epoch": 1.451913860310563, "grad_norm": 10.84099368492692, "learning_rate": 1.0591511274020364e-07, "logits/chosen": -3.0, "logits/rejected": -2.703125, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1693, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.46875, "rewards/rejected": -9.8125, "step": 11360 }, { "epoch": 1.4531918972458304, "grad_norm": 10.784148259297346, "learning_rate": 1.0545971965995634e-07, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1176.0, "loss": 0.141, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.875, "rewards/rejected": -10.1875, "step": 11370 }, { "epoch": 1.4544699341810978, "grad_norm": 8.135502688800099, "learning_rate": 1.05005045859899e-07, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -668.0, "logps/rejected": -1088.0, "loss": 0.1616, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.3125, "rewards/rejected": -9.625, "step": 11380 }, { "epoch": 1.4557479711163652, "grad_norm": 8.00246592998164, "learning_rate": 1.0455109360263889e-07, "logits/chosen": -2.875, "logits/rejected": -2.796875, "logps/chosen": -692.0, "logps/rejected": -1104.0, "loss": 0.1657, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.25, "rewards/margins": 4.25, "rewards/rejected": -9.5, "step": 11390 }, { "epoch": 1.4570260080516326, "grad_norm": 9.797097532732542, "learning_rate": 1.0409786514719223e-07, "logits/chosen": -2.984375, "logits/rejected": -2.875, "logps/chosen": -728.0, "logps/rejected": -1096.0, "loss": 0.1604, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 11400 }, { "epoch": 1.4583040449869, "grad_norm": 9.694349670220472, "learning_rate": 1.0364536274897376e-07, "logits/chosen": -3.046875, "logits/rejected": -2.703125, "logps/chosen": -668.0, "logps/rejected": -1144.0, "loss": 0.1455, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.1875, "rewards/margins": 4.75, "rewards/rejected": -9.9375, "step": 11410 }, { "epoch": 1.4595820819221674, "grad_norm": 12.38283245894702, "learning_rate": 1.031935886597848e-07, "logits/chosen": -3.015625, "logits/rejected": -2.640625, "logps/chosen": -680.0, "logps/rejected": -1184.0, "loss": 0.1535, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.28125, "rewards/margins": 5.1875, "rewards/rejected": -10.5, "step": 11420 }, { "epoch": 1.460860118857435, "grad_norm": 8.936517008363177, "learning_rate": 1.027425451278025e-07, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -692.0, "logps/rejected": -1208.0, "loss": 0.153, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.28125, "rewards/margins": 5.15625, "rewards/rejected": -10.4375, "step": 11430 }, { "epoch": 1.4621381557927025, "grad_norm": 10.831220985154136, "learning_rate": 1.0229223439756851e-07, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.1656, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.46875, "rewards/rejected": -9.9375, "step": 11440 }, { "epoch": 1.4634161927279699, "grad_norm": 8.556358700103702, "learning_rate": 1.0184265870997768e-07, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -720.0, "logps/rejected": -1128.0, "loss": 0.1367, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.375, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 11450 }, { "epoch": 1.4646942296632373, "grad_norm": 12.077581357401632, "learning_rate": 1.0139382030226737e-07, "logits/chosen": -2.96875, "logits/rejected": -2.9375, "logps/chosen": -720.0, "logps/rejected": -1200.0, "loss": 0.1411, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.5625, "rewards/margins": 4.96875, "rewards/rejected": -10.5625, "step": 11460 }, { "epoch": 1.4659722665985047, "grad_norm": 10.787672273909923, "learning_rate": 1.0094572140800558e-07, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1739, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.375, "rewards/rejected": -9.875, "step": 11470 }, { "epoch": 1.467250303533772, "grad_norm": 9.948886157796718, "learning_rate": 1.004983642570807e-07, "logits/chosen": -2.890625, "logits/rejected": -2.71875, "logps/chosen": -716.0, "logps/rejected": -1112.0, "loss": 0.168, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.21875, "rewards/rejected": -9.625, "step": 11480 }, { "epoch": 1.4685283404690397, "grad_norm": 15.864211760797932, "learning_rate": 1.0005175107568953e-07, "logits/chosen": -2.890625, "logits/rejected": -2.6875, "logps/chosen": -696.0, "logps/rejected": -1128.0, "loss": 0.1541, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 11490 }, { "epoch": 1.469806377404307, "grad_norm": 12.462596811906824, "learning_rate": 9.960588408632683e-08, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1088.0, "loss": 0.1577, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.0, "rewards/rejected": -9.4375, "step": 11500 }, { "epoch": 1.4710844143395745, "grad_norm": 9.807942803597852, "learning_rate": 9.916076550777417e-08, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.1628, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.3125, "rewards/rejected": -9.6875, "step": 11510 }, { "epoch": 1.4723624512748419, "grad_norm": 12.380375604436118, "learning_rate": 9.871639755508852e-08, "logits/chosen": -3.109375, "logits/rejected": -2.84375, "logps/chosen": -704.0, "logps/rejected": -1096.0, "loss": 0.178, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.375, "rewards/margins": 4.21875, "rewards/rejected": -9.625, "step": 11520 }, { "epoch": 1.4736404882101093, "grad_norm": 12.77801489441428, "learning_rate": 9.827278243959176e-08, "logits/chosen": -2.921875, "logits/rejected": -2.8125, "logps/chosen": -700.0, "logps/rejected": -1096.0, "loss": 0.1753, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.15625, "rewards/rejected": -9.625, "step": 11530 }, { "epoch": 1.4749185251453767, "grad_norm": 10.35489180421786, "learning_rate": 9.78299223688592e-08, "logits/chosen": -3.078125, "logits/rejected": -2.75, "logps/chosen": -684.0, "logps/rejected": -1112.0, "loss": 0.1516, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.28125, "rewards/rejected": -9.625, "step": 11540 }, { "epoch": 1.476196562080644, "grad_norm": 14.764091673164994, "learning_rate": 9.73878195467088e-08, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -704.0, "logps/rejected": -1128.0, "loss": 0.1589, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.4375, "rewards/margins": 4.40625, "rewards/rejected": -9.875, "step": 11550 }, { "epoch": 1.4774745990159115, "grad_norm": 8.91578021211425, "learning_rate": 9.694647617319044e-08, "logits/chosen": -3.140625, "logits/rejected": -2.90625, "logps/chosen": -692.0, "logps/rejected": -1136.0, "loss": 0.1648, "rewards/accuracies": 0.9375, "rewards/chosen": -5.25, "rewards/margins": 4.59375, "rewards/rejected": -9.8125, "step": 11560 }, { "epoch": 1.478752635951179, "grad_norm": 10.9841675460549, "learning_rate": 9.650589444457435e-08, "logits/chosen": -3.0625, "logits/rejected": -2.890625, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1397, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.625, "rewards/margins": 4.03125, "rewards/rejected": -9.625, "step": 11570 }, { "epoch": 1.4800306728864463, "grad_norm": 9.89509879622352, "learning_rate": 9.606607655334103e-08, "logits/chosen": -3.125, "logits/rejected": -2.765625, "logps/chosen": -744.0, "logps/rejected": -1176.0, "loss": 0.1384, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.8125, "rewards/margins": 4.53125, "rewards/rejected": -10.375, "step": 11580 }, { "epoch": 1.481308709821714, "grad_norm": 8.810197963037124, "learning_rate": 9.562702468816946e-08, "logits/chosen": -3.09375, "logits/rejected": -2.6875, "logps/chosen": -680.0, "logps/rejected": -1136.0, "loss": 0.1529, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.5, "rewards/rejected": -9.9375, "step": 11590 }, { "epoch": 1.4825867467569813, "grad_norm": 11.187235524321606, "learning_rate": 9.518874103392673e-08, "logits/chosen": -3.015625, "logits/rejected": -2.890625, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1545, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.28125, "rewards/margins": 4.34375, "rewards/rejected": -9.625, "step": 11600 }, { "epoch": 1.4838647836922487, "grad_norm": 8.308168361331765, "learning_rate": 9.475122777165729e-08, "logits/chosen": -3.21875, "logits/rejected": -2.96875, "logps/chosen": -720.0, "logps/rejected": -1128.0, "loss": 0.1392, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.21875, "rewards/rejected": -9.75, "step": 11610 }, { "epoch": 1.4851428206275161, "grad_norm": 12.759456416581942, "learning_rate": 9.431448707857148e-08, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -724.0, "logps/rejected": -1104.0, "loss": 0.1407, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.6875, "rewards/margins": 3.984375, "rewards/rejected": -9.625, "step": 11620 }, { "epoch": 1.4864208575627835, "grad_norm": 9.265918162123217, "learning_rate": 9.387852112803546e-08, "logits/chosen": -2.984375, "logits/rejected": -2.828125, "logps/chosen": -740.0, "logps/rejected": -1152.0, "loss": 0.146, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.78125, "rewards/margins": 4.15625, "rewards/rejected": -9.9375, "step": 11630 }, { "epoch": 1.487698894498051, "grad_norm": 8.332858163699392, "learning_rate": 9.34433320895598e-08, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -724.0, "logps/rejected": -1160.0, "loss": 0.1364, "rewards/accuracies": 0.9375, "rewards/chosen": -5.71875, "rewards/margins": 4.53125, "rewards/rejected": -10.25, "step": 11640 }, { "epoch": 1.4889769314333186, "grad_norm": 12.235024940851668, "learning_rate": 9.300892212878886e-08, "logits/chosen": -2.984375, "logits/rejected": -2.859375, "logps/chosen": -716.0, "logps/rejected": -1088.0, "loss": 0.1507, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 4.1875, "rewards/rejected": -9.6875, "step": 11650 }, { "epoch": 1.490254968368586, "grad_norm": 13.052203756824563, "learning_rate": 9.257529340749021e-08, "logits/chosen": -3.078125, "logits/rejected": -2.78125, "logps/chosen": -720.0, "logps/rejected": -1144.0, "loss": 0.1563, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.625, "rewards/margins": 4.46875, "rewards/rejected": -10.125, "step": 11660 }, { "epoch": 1.4915330053038534, "grad_norm": 10.747014663632283, "learning_rate": 9.214244808354351e-08, "logits/chosen": -3.0625, "logits/rejected": -2.84375, "logps/chosen": -676.0, "logps/rejected": -1112.0, "loss": 0.178, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.25, "rewards/margins": 4.40625, "rewards/rejected": -9.625, "step": 11670 }, { "epoch": 1.4928110422391208, "grad_norm": 10.663336523174248, "learning_rate": 9.171038831093028e-08, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1654, "rewards/accuracies": 0.9375, "rewards/chosen": -5.5, "rewards/margins": 4.25, "rewards/rejected": -9.75, "step": 11680 }, { "epoch": 1.4940890791743882, "grad_norm": 9.739080278989183, "learning_rate": 9.127911623972253e-08, "logits/chosen": -3.03125, "logits/rejected": -2.625, "logps/chosen": -700.0, "logps/rejected": -1120.0, "loss": 0.1298, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.1875, "rewards/rejected": -9.75, "step": 11690 }, { "epoch": 1.4953671161096556, "grad_norm": 10.515221950290618, "learning_rate": 9.084863401607273e-08, "logits/chosen": -3.140625, "logits/rejected": -2.8125, "logps/chosen": -648.0, "logps/rejected": -1144.0, "loss": 0.1541, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.09375, "rewards/margins": 4.96875, "rewards/rejected": -10.0625, "step": 11700 }, { "epoch": 1.496645153044923, "grad_norm": 10.246625757921912, "learning_rate": 9.041894378220258e-08, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -752.0, "logps/rejected": -1136.0, "loss": 0.1573, "rewards/accuracies": 0.9375, "rewards/chosen": -5.8125, "rewards/margins": 4.09375, "rewards/rejected": -9.875, "step": 11710 }, { "epoch": 1.4979231899801904, "grad_norm": 12.99095492752731, "learning_rate": 8.999004767639256e-08, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1136.0, "loss": 0.1511, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.4375, "rewards/rejected": -10.0, "step": 11720 }, { "epoch": 1.4992012269154578, "grad_norm": 10.437109592299482, "learning_rate": 8.95619478329716e-08, "logits/chosen": -2.953125, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1152.0, "loss": 0.1588, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 4.46875, "rewards/rejected": -10.0, "step": 11730 }, { "epoch": 1.5004792638507252, "grad_norm": 10.487452986000255, "learning_rate": 8.913464638230575e-08, "logits/chosen": -3.046875, "logits/rejected": -2.859375, "logps/chosen": -736.0, "logps/rejected": -1096.0, "loss": 0.1591, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.71875, "rewards/margins": 3.921875, "rewards/rejected": -9.625, "step": 11740 }, { "epoch": 1.5017573007859926, "grad_norm": 9.032769583796481, "learning_rate": 8.870814545078839e-08, "logits/chosen": -3.125, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1459, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5, "rewards/margins": 4.34375, "rewards/rejected": -9.8125, "step": 11750 }, { "epoch": 1.5030353377212602, "grad_norm": 12.474367808376874, "learning_rate": 8.828244716082897e-08, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1176.0, "loss": 0.1642, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.625, "rewards/margins": 4.53125, "rewards/rejected": -10.1875, "step": 11760 }, { "epoch": 1.5043133746565276, "grad_norm": 13.831164839214665, "learning_rate": 8.785755363084277e-08, "logits/chosen": -3.0625, "logits/rejected": -2.90625, "logps/chosen": -696.0, "logps/rejected": -1152.0, "loss": 0.1741, "rewards/accuracies": 0.9375, "rewards/chosen": -5.5, "rewards/margins": 4.6875, "rewards/rejected": -10.1875, "step": 11770 }, { "epoch": 1.505591411591795, "grad_norm": 11.80319684012857, "learning_rate": 8.743346697524054e-08, "logits/chosen": -2.90625, "logits/rejected": -2.921875, "logps/chosen": -728.0, "logps/rejected": -1128.0, "loss": 0.1588, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.3125, "rewards/rejected": -9.875, "step": 11780 }, { "epoch": 1.5068694485270624, "grad_norm": 11.340568045210375, "learning_rate": 8.701018930441739e-08, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -756.0, "logps/rejected": -1184.0, "loss": 0.1596, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.875, "rewards/margins": 4.4375, "rewards/rejected": -10.3125, "step": 11790 }, { "epoch": 1.50814748546233, "grad_norm": 10.031009539482778, "learning_rate": 8.658772272474307e-08, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1714, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.25, "rewards/rejected": -9.75, "step": 11800 }, { "epoch": 1.5094255223975974, "grad_norm": 10.222246813978815, "learning_rate": 8.616606933855077e-08, "logits/chosen": -3.109375, "logits/rejected": -2.78125, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1406, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.65625, "rewards/margins": 4.25, "rewards/rejected": -9.9375, "step": 11810 }, { "epoch": 1.5107035593328648, "grad_norm": 10.522623597133993, "learning_rate": 8.574523124412697e-08, "logits/chosen": -3.046875, "logits/rejected": -2.765625, "logps/chosen": -728.0, "logps/rejected": -1168.0, "loss": 0.1654, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.75, "rewards/margins": 4.34375, "rewards/rejected": -10.0625, "step": 11820 }, { "epoch": 1.5119815962681322, "grad_norm": 11.181079924622459, "learning_rate": 8.532521053570121e-08, "logits/chosen": -3.03125, "logits/rejected": -2.828125, "logps/chosen": -732.0, "logps/rejected": -1112.0, "loss": 0.1502, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.71875, "rewards/margins": 4.09375, "rewards/rejected": -9.8125, "step": 11830 }, { "epoch": 1.5132596332033996, "grad_norm": 7.667202990627593, "learning_rate": 8.490600930343514e-08, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -724.0, "logps/rejected": -1152.0, "loss": 0.1623, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.71875, "rewards/margins": 4.1875, "rewards/rejected": -9.875, "step": 11840 }, { "epoch": 1.514537670138667, "grad_norm": 9.646765032562787, "learning_rate": 8.44876296334128e-08, "logits/chosen": -3.125, "logits/rejected": -2.765625, "logps/chosen": -744.0, "logps/rejected": -1192.0, "loss": 0.1487, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.75, "rewards/margins": 4.65625, "rewards/rejected": -10.375, "step": 11850 }, { "epoch": 1.5158157070739344, "grad_norm": 10.012416218533366, "learning_rate": 8.407007360762955e-08, "logits/chosen": -3.0, "logits/rejected": -2.875, "logps/chosen": -700.0, "logps/rejected": -1112.0, "loss": 0.159, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.375, "rewards/margins": 4.21875, "rewards/rejected": -9.625, "step": 11860 }, { "epoch": 1.5170937440092018, "grad_norm": 8.150935233146557, "learning_rate": 8.36533433039821e-08, "logits/chosen": -2.953125, "logits/rejected": -2.640625, "logps/chosen": -736.0, "logps/rejected": -1136.0, "loss": 0.1648, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.8125, "rewards/margins": 4.15625, "rewards/rejected": -10.0, "step": 11870 }, { "epoch": 1.5183717809444692, "grad_norm": 10.803015402481362, "learning_rate": 8.323744079625819e-08, "logits/chosen": -3.03125, "logits/rejected": -2.765625, "logps/chosen": -716.0, "logps/rejected": -1112.0, "loss": 0.1701, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.75, "rewards/margins": 3.96875, "rewards/rejected": -9.6875, "step": 11880 }, { "epoch": 1.5196498178797366, "grad_norm": 11.958688449421555, "learning_rate": 8.282236815412599e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1648, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.625, "rewards/margins": 4.28125, "rewards/rejected": -9.875, "step": 11890 }, { "epoch": 1.520927854815004, "grad_norm": 10.106095459532787, "learning_rate": 8.240812744312428e-08, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -684.0, "logps/rejected": -1128.0, "loss": 0.1532, "rewards/accuracies": 0.90625, "rewards/chosen": -5.34375, "rewards/margins": 4.5, "rewards/rejected": -9.8125, "step": 11900 }, { "epoch": 1.5222058917502714, "grad_norm": 12.37592965783719, "learning_rate": 8.199472072465153e-08, "logits/chosen": -2.90625, "logits/rejected": -2.875, "logps/chosen": -704.0, "logps/rejected": -1096.0, "loss": 0.1578, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.21875, "rewards/rejected": -9.6875, "step": 11910 }, { "epoch": 1.523483928685539, "grad_norm": 10.377992192363955, "learning_rate": 8.158215005595628e-08, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -712.0, "logps/rejected": -1136.0, "loss": 0.1483, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.40625, "rewards/rejected": -9.875, "step": 11920 }, { "epoch": 1.5247619656208065, "grad_norm": 10.792908757384343, "learning_rate": 8.117041749012649e-08, "logits/chosen": -3.09375, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1531, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.375, "rewards/margins": 4.15625, "rewards/rejected": -9.5, "step": 11930 }, { "epoch": 1.5260400025560739, "grad_norm": 11.001792418578807, "learning_rate": 8.075952507607931e-08, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -744.0, "logps/rejected": -1160.0, "loss": 0.1603, "rewards/accuracies": 0.90625, "rewards/chosen": -5.78125, "rewards/margins": 4.34375, "rewards/rejected": -10.125, "step": 11940 }, { "epoch": 1.5273180394913413, "grad_norm": 10.242356822888798, "learning_rate": 8.034947485855131e-08, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -728.0, "logps/rejected": -1160.0, "loss": 0.166, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.5625, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 11950 }, { "epoch": 1.5285960764266089, "grad_norm": 9.251414278750838, "learning_rate": 7.994026887808772e-08, "logits/chosen": -3.046875, "logits/rejected": -2.859375, "logps/chosen": -724.0, "logps/rejected": -1192.0, "loss": 0.1411, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.75, "rewards/margins": 4.75, "rewards/rejected": -10.5, "step": 11960 }, { "epoch": 1.5298741133618763, "grad_norm": 11.853751107996388, "learning_rate": 7.95319091710328e-08, "logits/chosen": -3.09375, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1152.0, "loss": 0.1534, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.6875, "rewards/margins": 4.5, "rewards/rejected": -10.1875, "step": 11970 }, { "epoch": 1.5311521502971437, "grad_norm": 14.040158388167082, "learning_rate": 7.912439776951935e-08, "logits/chosen": -2.859375, "logits/rejected": -2.78125, "logps/chosen": -700.0, "logps/rejected": -1136.0, "loss": 0.1735, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.34375, "rewards/rejected": -9.75, "step": 11980 }, { "epoch": 1.532430187232411, "grad_norm": 9.573096797550356, "learning_rate": 7.871773670145862e-08, "logits/chosen": -2.96875, "logits/rejected": -2.90625, "logps/chosen": -708.0, "logps/rejected": -1080.0, "loss": 0.1562, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.4375, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 11990 }, { "epoch": 1.5337082241676785, "grad_norm": 9.396314193149403, "learning_rate": 7.831192799053066e-08, "logits/chosen": -3.09375, "logits/rejected": -2.78125, "logps/chosen": -700.0, "logps/rejected": -1160.0, "loss": 0.1399, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.5, "rewards/margins": 4.59375, "rewards/rejected": -10.125, "step": 12000 }, { "epoch": 1.534986261102946, "grad_norm": 16.69418623861175, "learning_rate": 7.790697365617347e-08, "logits/chosen": -3.109375, "logits/rejected": -2.875, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1415, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.375, "rewards/rejected": -9.8125, "step": 12010 }, { "epoch": 1.5362642980382133, "grad_norm": 10.567016182091837, "learning_rate": 7.750287571357377e-08, "logits/chosen": -3.109375, "logits/rejected": -2.796875, "logps/chosen": -700.0, "logps/rejected": -1128.0, "loss": 0.1586, "rewards/accuracies": 0.9375, "rewards/chosen": -5.375, "rewards/margins": 4.5, "rewards/rejected": -9.875, "step": 12020 }, { "epoch": 1.5375423349734807, "grad_norm": 11.50615338821914, "learning_rate": 7.709963617365642e-08, "logits/chosen": -3.15625, "logits/rejected": -2.8125, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1753, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5, "rewards/margins": 4.34375, "rewards/rejected": -9.8125, "step": 12030 }, { "epoch": 1.538820371908748, "grad_norm": 10.65143104102225, "learning_rate": 7.669725704307445e-08, "logits/chosen": -3.109375, "logits/rejected": -2.6875, "logps/chosen": -712.0, "logps/rejected": -1120.0, "loss": 0.1668, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.25, "rewards/rejected": -9.75, "step": 12040 }, { "epoch": 1.5400984088440155, "grad_norm": 10.025028948194606, "learning_rate": 7.629574032419958e-08, "logits/chosen": -3.09375, "logits/rejected": -2.90625, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1433, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.25, "rewards/rejected": -9.875, "step": 12050 }, { "epoch": 1.541376445779283, "grad_norm": 9.686910173982989, "learning_rate": 7.589508801511143e-08, "logits/chosen": -3.140625, "logits/rejected": -2.96875, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.142, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 12060 }, { "epoch": 1.5426544827145503, "grad_norm": 7.269425221448671, "learning_rate": 7.549530210958854e-08, "logits/chosen": -3.046875, "logits/rejected": -2.703125, "logps/chosen": -684.0, "logps/rejected": -1096.0, "loss": 0.1409, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.25, "rewards/rejected": -9.6875, "step": 12070 }, { "epoch": 1.543932519649818, "grad_norm": 12.223446432641076, "learning_rate": 7.509638459709739e-08, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1120.0, "loss": 0.1642, "rewards/accuracies": 0.90625, "rewards/chosen": -5.65625, "rewards/margins": 4.03125, "rewards/rejected": -9.6875, "step": 12080 }, { "epoch": 1.5452105565850853, "grad_norm": 8.641546278848681, "learning_rate": 7.469833746278346e-08, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -696.0, "logps/rejected": -1128.0, "loss": 0.1477, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.34375, "rewards/rejected": -9.8125, "step": 12090 }, { "epoch": 1.5464885935203527, "grad_norm": 9.917101746029982, "learning_rate": 7.430116268746078e-08, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -692.0, "logps/rejected": -1152.0, "loss": 0.1714, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.625, "rewards/rejected": -10.0625, "step": 12100 }, { "epoch": 1.5477666304556201, "grad_norm": 13.761179512544745, "learning_rate": 7.39048622476022e-08, "logits/chosen": -3.0625, "logits/rejected": -2.859375, "logps/chosen": -724.0, "logps/rejected": -1152.0, "loss": 0.1411, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 12110 }, { "epoch": 1.5490446673908878, "grad_norm": 11.119246022232648, "learning_rate": 7.350943811532976e-08, "logits/chosen": -2.859375, "logits/rejected": -2.6875, "logps/chosen": -728.0, "logps/rejected": -1120.0, "loss": 0.1634, "rewards/accuracies": 0.90625, "rewards/chosen": -5.8125, "rewards/margins": 4.03125, "rewards/rejected": -9.875, "step": 12120 }, { "epoch": 1.5503227043261552, "grad_norm": 9.249195090157448, "learning_rate": 7.311489225840444e-08, "logits/chosen": -2.921875, "logits/rejected": -2.859375, "logps/chosen": -732.0, "logps/rejected": -1128.0, "loss": 0.1543, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.71875, "rewards/margins": 4.25, "rewards/rejected": -9.9375, "step": 12130 }, { "epoch": 1.5516007412614226, "grad_norm": 10.447651696391043, "learning_rate": 7.272122664021695e-08, "logits/chosen": -2.9375, "logits/rejected": -2.65625, "logps/chosen": -744.0, "logps/rejected": -1152.0, "loss": 0.1677, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.75, "rewards/margins": 4.4375, "rewards/rejected": -10.1875, "step": 12140 }, { "epoch": 1.55287877819669, "grad_norm": 10.864502757673646, "learning_rate": 7.232844321977738e-08, "logits/chosen": -2.828125, "logits/rejected": -2.703125, "logps/chosen": -748.0, "logps/rejected": -1160.0, "loss": 0.1459, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -6.0, "rewards/margins": 4.28125, "rewards/rejected": -10.25, "step": 12150 }, { "epoch": 1.5541568151319574, "grad_norm": 8.85828421121414, "learning_rate": 7.193654395170573e-08, "logits/chosen": -2.953125, "logits/rejected": -2.609375, "logps/chosen": -736.0, "logps/rejected": -1168.0, "loss": 0.1511, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.84375, "rewards/margins": 4.375, "rewards/rejected": -10.25, "step": 12160 }, { "epoch": 1.5554348520672248, "grad_norm": 16.50517282532781, "learning_rate": 7.154553078622239e-08, "logits/chosen": -2.9375, "logits/rejected": -2.609375, "logps/chosen": -732.0, "logps/rejected": -1112.0, "loss": 0.1755, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.625, "rewards/margins": 4.15625, "rewards/rejected": -9.8125, "step": 12170 }, { "epoch": 1.5567128890024922, "grad_norm": 8.574202855048568, "learning_rate": 7.115540566913797e-08, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -712.0, "logps/rejected": -1136.0, "loss": 0.1436, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.375, "rewards/margins": 4.40625, "rewards/rejected": -9.8125, "step": 12180 }, { "epoch": 1.5579909259377596, "grad_norm": 10.810888515464885, "learning_rate": 7.076617054184389e-08, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.1574, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 12190 }, { "epoch": 1.559268962873027, "grad_norm": 7.457960346140366, "learning_rate": 7.03778273413029e-08, "logits/chosen": -2.890625, "logits/rejected": -2.734375, "logps/chosen": -732.0, "logps/rejected": -1104.0, "loss": 0.15, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.8125, "rewards/margins": 3.890625, "rewards/rejected": -9.6875, "step": 12200 }, { "epoch": 1.5605469998082944, "grad_norm": 7.279709759728956, "learning_rate": 6.99903780000389e-08, "logits/chosen": -2.921875, "logits/rejected": -2.625, "logps/chosen": -748.0, "logps/rejected": -1200.0, "loss": 0.1338, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.875, "rewards/margins": 4.78125, "rewards/rejected": -10.625, "step": 12210 }, { "epoch": 1.5618250367435618, "grad_norm": 9.062437716884702, "learning_rate": 6.960382444612801e-08, "logits/chosen": -2.953125, "logits/rejected": -2.5625, "logps/chosen": -728.0, "logps/rejected": -1224.0, "loss": 0.1722, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.6875, "rewards/margins": 5.125, "rewards/rejected": -10.8125, "step": 12220 }, { "epoch": 1.5631030736788292, "grad_norm": 13.459270325223686, "learning_rate": 6.92181686031883e-08, "logits/chosen": -3.015625, "logits/rejected": -2.75, "logps/chosen": -764.0, "logps/rejected": -1216.0, "loss": 0.1632, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -6.0625, "rewards/margins": 4.625, "rewards/rejected": -10.6875, "step": 12230 }, { "epoch": 1.5643811106140968, "grad_norm": 11.420081863784505, "learning_rate": 6.883341239037064e-08, "logits/chosen": -2.90625, "logits/rejected": -2.796875, "logps/chosen": -720.0, "logps/rejected": -1144.0, "loss": 0.1602, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.625, "rewards/margins": 4.4375, "rewards/rejected": -10.0625, "step": 12240 }, { "epoch": 1.5656591475493642, "grad_norm": 13.417134293230164, "learning_rate": 6.844955772234917e-08, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -672.0, "logps/rejected": -1096.0, "loss": 0.1536, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.3125, "rewards/margins": 4.3125, "rewards/rejected": -9.625, "step": 12250 }, { "epoch": 1.5669371844846316, "grad_norm": 9.667616154823163, "learning_rate": 6.806660650931142e-08, "logits/chosen": -2.984375, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1072.0, "loss": 0.1428, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.21875, "rewards/margins": 4.21875, "rewards/rejected": -9.4375, "step": 12260 }, { "epoch": 1.568215221419899, "grad_norm": 8.997991857537537, "learning_rate": 6.76845606569493e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -1104.0, "loss": 0.1452, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.375, "rewards/margins": 4.4375, "rewards/rejected": -9.8125, "step": 12270 }, { "epoch": 1.5694932583551666, "grad_norm": 9.04008816490061, "learning_rate": 6.73034220664491e-08, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -700.0, "logps/rejected": -1104.0, "loss": 0.1521, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5, "rewards/margins": 4.125, "rewards/rejected": -9.625, "step": 12280 }, { "epoch": 1.570771295290434, "grad_norm": 9.432414905692434, "learning_rate": 6.69231926344824e-08, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -732.0, "logps/rejected": -1152.0, "loss": 0.1475, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.84375, "rewards/margins": 4.34375, "rewards/rejected": -10.1875, "step": 12290 }, { "epoch": 1.5720493322257014, "grad_norm": 12.035295485184237, "learning_rate": 6.654387425319646e-08, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -724.0, "logps/rejected": -1168.0, "loss": 0.1629, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.65625, "rewards/margins": 4.5, "rewards/rejected": -10.1875, "step": 12300 }, { "epoch": 1.5733273691609688, "grad_norm": 14.534600801356092, "learning_rate": 6.616546881020479e-08, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -740.0, "logps/rejected": -1160.0, "loss": 0.1555, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.8125, "rewards/margins": 4.15625, "rewards/rejected": -9.9375, "step": 12310 }, { "epoch": 1.5746054060962362, "grad_norm": 10.748006072313448, "learning_rate": 6.578797818857806e-08, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -728.0, "logps/rejected": -1160.0, "loss": 0.1497, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.625, "rewards/rejected": -10.1875, "step": 12320 }, { "epoch": 1.5758834430315036, "grad_norm": 9.366457097549711, "learning_rate": 6.54114042668342e-08, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1628, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.71875, "rewards/margins": 4.0625, "rewards/rejected": -9.75, "step": 12330 }, { "epoch": 1.577161479966771, "grad_norm": 14.29220140565374, "learning_rate": 6.503574891892954e-08, "logits/chosen": -3.109375, "logits/rejected": -2.78125, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1496, "rewards/accuracies": 0.9375, "rewards/chosen": -5.53125, "rewards/margins": 4.6875, "rewards/rejected": -10.25, "step": 12340 }, { "epoch": 1.5784395169020384, "grad_norm": 14.439666635744588, "learning_rate": 6.46610140142492e-08, "logits/chosen": -3.15625, "logits/rejected": -2.796875, "logps/chosen": -768.0, "logps/rejected": -1192.0, "loss": 0.1514, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -6.09375, "rewards/margins": 4.25, "rewards/rejected": -10.3125, "step": 12350 }, { "epoch": 1.5797175538373058, "grad_norm": 9.351590074519404, "learning_rate": 6.428720141759778e-08, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -752.0, "logps/rejected": -1144.0, "loss": 0.163, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.8125, "rewards/margins": 4.25, "rewards/rejected": -10.0625, "step": 12360 }, { "epoch": 1.5809955907725732, "grad_norm": 13.470209400982355, "learning_rate": 6.39143129891904e-08, "logits/chosen": -3.109375, "logits/rejected": -2.828125, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1572, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.4375, "rewards/rejected": -10.0, "step": 12370 }, { "epoch": 1.5822736277078406, "grad_norm": 11.158096087311577, "learning_rate": 6.354235058464295e-08, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1698, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.1875, "rewards/rejected": -9.6875, "step": 12380 }, { "epoch": 1.583551664643108, "grad_norm": 9.951528485560331, "learning_rate": 6.317131605496334e-08, "logits/chosen": -3.046875, "logits/rejected": -2.609375, "logps/chosen": -684.0, "logps/rejected": -1136.0, "loss": 0.1495, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.25, "rewards/margins": 4.5625, "rewards/rejected": -9.8125, "step": 12390 }, { "epoch": 1.5848297015783757, "grad_norm": 10.644975027048147, "learning_rate": 6.280121124654201e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -712.0, "logps/rejected": -1120.0, "loss": 0.136, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.21875, "rewards/rejected": -9.8125, "step": 12400 }, { "epoch": 1.586107738513643, "grad_norm": 11.807603959654157, "learning_rate": 6.243203800114257e-08, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -716.0, "logps/rejected": -1176.0, "loss": 0.1471, "rewards/accuracies": 0.9375, "rewards/chosen": -5.59375, "rewards/margins": 4.65625, "rewards/rejected": -10.25, "step": 12410 }, { "epoch": 1.5873857754489105, "grad_norm": 7.108280544801186, "learning_rate": 6.206379815589333e-08, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1216.0, "loss": 0.1553, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.4375, "rewards/margins": 5.21875, "rewards/rejected": -10.6875, "step": 12420 }, { "epoch": 1.5886638123841779, "grad_norm": 12.019450109413693, "learning_rate": 6.169649354327719e-08, "logits/chosen": -3.171875, "logits/rejected": -2.875, "logps/chosen": -724.0, "logps/rejected": -1152.0, "loss": 0.1654, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.53125, "rewards/rejected": -10.125, "step": 12430 }, { "epoch": 1.5899418493194455, "grad_norm": 13.551368457160118, "learning_rate": 6.133012599112341e-08, "logits/chosen": -3.09375, "logits/rejected": -2.875, "logps/chosen": -724.0, "logps/rejected": -1136.0, "loss": 0.1552, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.71875, "rewards/margins": 4.3125, "rewards/rejected": -10.0, "step": 12440 }, { "epoch": 1.591219886254713, "grad_norm": 17.599148853785525, "learning_rate": 6.096469732259795e-08, "logits/chosen": -3.046875, "logits/rejected": -2.890625, "logps/chosen": -696.0, "logps/rejected": -1120.0, "loss": 0.1427, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.34375, "rewards/margins": 4.46875, "rewards/rejected": -9.8125, "step": 12450 }, { "epoch": 1.5924979231899803, "grad_norm": 9.710638561537065, "learning_rate": 6.06002093561945e-08, "logits/chosen": -3.15625, "logits/rejected": -2.875, "logps/chosen": -728.0, "logps/rejected": -1200.0, "loss": 0.1316, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.6875, "rewards/margins": 4.65625, "rewards/rejected": -10.375, "step": 12460 }, { "epoch": 1.5937759601252477, "grad_norm": 11.967214456719551, "learning_rate": 6.023666390572576e-08, "logits/chosen": -3.15625, "logits/rejected": -2.796875, "logps/chosen": -736.0, "logps/rejected": -1272.0, "loss": 0.1695, "rewards/accuracies": 0.9375, "rewards/chosen": -5.8125, "rewards/margins": 5.21875, "rewards/rejected": -11.0, "step": 12470 }, { "epoch": 1.595053997060515, "grad_norm": 9.109161920815124, "learning_rate": 5.987406278031384e-08, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -720.0, "logps/rejected": -1160.0, "loss": 0.1404, "rewards/accuracies": 0.9375, "rewards/chosen": -5.625, "rewards/margins": 4.6875, "rewards/rejected": -10.3125, "step": 12480 }, { "epoch": 1.5963320339957825, "grad_norm": 9.579879903189044, "learning_rate": 5.951240778438188e-08, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -756.0, "logps/rejected": -1136.0, "loss": 0.1656, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -6.0, "rewards/margins": 4.03125, "rewards/rejected": -10.0, "step": 12490 }, { "epoch": 1.59761007093105, "grad_norm": 9.228594375064981, "learning_rate": 5.915170071764461e-08, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -696.0, "logps/rejected": -1088.0, "loss": 0.1375, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.21875, "rewards/rejected": -9.6875, "step": 12500 }, { "epoch": 1.5988881078663173, "grad_norm": 12.547914385348141, "learning_rate": 5.879194337509949e-08, "logits/chosen": -3.03125, "logits/rejected": -2.859375, "logps/chosen": -744.0, "logps/rejected": -1136.0, "loss": 0.1435, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.84375, "rewards/margins": 4.21875, "rewards/rejected": -10.0625, "step": 12510 }, { "epoch": 1.6001661448015847, "grad_norm": 13.744472128530129, "learning_rate": 5.843313754701798e-08, "logits/chosen": -3.046875, "logits/rejected": -2.859375, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1396, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.625, "rewards/margins": 4.34375, "rewards/rejected": -10.0, "step": 12520 }, { "epoch": 1.601444181736852, "grad_norm": 9.517400972891124, "learning_rate": 5.8075285018936325e-08, "logits/chosen": -3.1875, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1136.0, "loss": 0.1186, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 12530 }, { "epoch": 1.6027222186721195, "grad_norm": 10.271480402842696, "learning_rate": 5.77183875716471e-08, "logits/chosen": -3.03125, "logits/rejected": -2.890625, "logps/chosen": -712.0, "logps/rejected": -1112.0, "loss": 0.1534, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.75, "rewards/margins": 4.0625, "rewards/rejected": -9.8125, "step": 12540 }, { "epoch": 1.604000255607387, "grad_norm": 12.276563592099997, "learning_rate": 5.7362446981189775e-08, "logits/chosen": -3.015625, "logits/rejected": -2.890625, "logps/chosen": -720.0, "logps/rejected": -1144.0, "loss": 0.1523, "rewards/accuracies": 0.90625, "rewards/chosen": -5.5, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 12550 }, { "epoch": 1.6052782925426545, "grad_norm": 10.450582546903629, "learning_rate": 5.7007465018842435e-08, "logits/chosen": -3.125, "logits/rejected": -2.734375, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1688, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 4.65625, "rewards/rejected": -10.125, "step": 12560 }, { "epoch": 1.606556329477922, "grad_norm": 12.859229216456761, "learning_rate": 5.6653443451112585e-08, "logits/chosen": -3.046875, "logits/rejected": -2.84375, "logps/chosen": -736.0, "logps/rejected": -1152.0, "loss": 0.1518, "rewards/accuracies": 0.9375, "rewards/chosen": -5.75, "rewards/margins": 4.3125, "rewards/rejected": -10.0625, "step": 12570 }, { "epoch": 1.6078343664131893, "grad_norm": 10.183791967823463, "learning_rate": 5.6300384039728424e-08, "logits/chosen": -3.09375, "logits/rejected": -2.71875, "logps/chosen": -724.0, "logps/rejected": -1152.0, "loss": 0.1614, "rewards/accuracies": 0.9375, "rewards/chosen": -5.625, "rewards/margins": 4.3125, "rewards/rejected": -9.9375, "step": 12580 }, { "epoch": 1.6091124033484567, "grad_norm": 12.161865277394519, "learning_rate": 5.594828854163036e-08, "logits/chosen": -3.078125, "logits/rejected": -2.703125, "logps/chosen": -716.0, "logps/rejected": -1152.0, "loss": 0.1582, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.75, "rewards/margins": 4.28125, "rewards/rejected": -10.0, "step": 12590 }, { "epoch": 1.6103904402837244, "grad_norm": 8.506154794757979, "learning_rate": 5.559715870896178e-08, "logits/chosen": -3.125, "logits/rejected": -2.890625, "logps/chosen": -736.0, "logps/rejected": -1120.0, "loss": 0.1548, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.75, "rewards/margins": 4.03125, "rewards/rejected": -9.75, "step": 12600 }, { "epoch": 1.6116684772189918, "grad_norm": 10.724959243801147, "learning_rate": 5.52469962890609e-08, "logits/chosen": -2.96875, "logits/rejected": -2.828125, "logps/chosen": -716.0, "logps/rejected": -1096.0, "loss": 0.1491, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.09375, "rewards/rejected": -9.5, "step": 12610 }, { "epoch": 1.6129465141542592, "grad_norm": 11.613970832785595, "learning_rate": 5.4897803024451545e-08, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1596, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.59375, "rewards/margins": 4.15625, "rewards/rejected": -9.75, "step": 12620 }, { "epoch": 1.6142245510895266, "grad_norm": 15.729825565373556, "learning_rate": 5.454958065283466e-08, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -748.0, "logps/rejected": -1168.0, "loss": 0.1657, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.75, "rewards/margins": 4.46875, "rewards/rejected": -10.1875, "step": 12630 }, { "epoch": 1.615502588024794, "grad_norm": 12.184563504498001, "learning_rate": 5.420233090708001e-08, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -732.0, "logps/rejected": -1112.0, "loss": 0.1444, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.75, "rewards/margins": 4.09375, "rewards/rejected": -9.875, "step": 12640 }, { "epoch": 1.6167806249600614, "grad_norm": 8.8771990213809, "learning_rate": 5.3856055515216906e-08, "logits/chosen": -3.09375, "logits/rejected": -2.71875, "logps/chosen": -692.0, "logps/rejected": -1232.0, "loss": 0.1403, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.46875, "rewards/margins": 5.0625, "rewards/rejected": -10.5625, "step": 12650 }, { "epoch": 1.6180586618953288, "grad_norm": 12.427781605965897, "learning_rate": 5.351075620042622e-08, "logits/chosen": -3.09375, "logits/rejected": -2.765625, "logps/chosen": -708.0, "logps/rejected": -1152.0, "loss": 0.1534, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.53125, "rewards/rejected": -10.0, "step": 12660 }, { "epoch": 1.6193366988305962, "grad_norm": 8.807307663163346, "learning_rate": 5.316643468103138e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -696.0, "logps/rejected": -1120.0, "loss": 0.1491, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.375, "rewards/rejected": -9.8125, "step": 12670 }, { "epoch": 1.6206147357658636, "grad_norm": 9.757673892289993, "learning_rate": 5.282309267048995e-08, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -712.0, "logps/rejected": -1120.0, "loss": 0.1662, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.25, "rewards/rejected": -9.75, "step": 12680 }, { "epoch": 1.621892772701131, "grad_norm": 11.35318485146273, "learning_rate": 5.2480731877385356e-08, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -708.0, "logps/rejected": -1160.0, "loss": 0.1416, "rewards/accuracies": 0.90625, "rewards/chosen": -5.4375, "rewards/margins": 4.625, "rewards/rejected": -10.0625, "step": 12690 }, { "epoch": 1.6231708096363984, "grad_norm": 9.466002182166104, "learning_rate": 5.2139354005417856e-08, "logits/chosen": -3.109375, "logits/rejected": -2.84375, "logps/chosen": -712.0, "logps/rejected": -1120.0, "loss": 0.1443, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.25, "rewards/rejected": -9.8125, "step": 12700 }, { "epoch": 1.6244488465716658, "grad_norm": 9.89133474990578, "learning_rate": 5.179896075339671e-08, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -684.0, "logps/rejected": -1128.0, "loss": 0.1601, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.5625, "rewards/rejected": -10.0, "step": 12710 }, { "epoch": 1.6257268835069334, "grad_norm": 9.510344550642348, "learning_rate": 5.1459553815231094e-08, "logits/chosen": -2.984375, "logits/rejected": -2.78125, "logps/chosen": -740.0, "logps/rejected": -1088.0, "loss": 0.1591, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.71875, "rewards/margins": 3.75, "rewards/rejected": -9.5, "step": 12720 }, { "epoch": 1.6270049204422008, "grad_norm": 14.356022509608808, "learning_rate": 5.112113487992209e-08, "logits/chosen": -2.984375, "logits/rejected": -2.703125, "logps/chosen": -732.0, "logps/rejected": -1184.0, "loss": 0.1766, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.71875, "rewards/margins": 4.65625, "rewards/rejected": -10.375, "step": 12730 }, { "epoch": 1.6282829573774682, "grad_norm": 10.447214245640378, "learning_rate": 5.078370563155413e-08, "logits/chosen": -3.0, "logits/rejected": -2.6875, "logps/chosen": -708.0, "logps/rejected": -1168.0, "loss": 0.1386, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.71875, "rewards/rejected": -10.3125, "step": 12740 }, { "epoch": 1.6295609943127356, "grad_norm": 11.237649051877147, "learning_rate": 5.044726774928657e-08, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -744.0, "logps/rejected": -1128.0, "loss": 0.163, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.9375, "rewards/margins": 4.09375, "rewards/rejected": -10.0625, "step": 12750 }, { "epoch": 1.6308390312480032, "grad_norm": 10.054893207508396, "learning_rate": 5.011182290734561e-08, "logits/chosen": -3.1875, "logits/rejected": -2.90625, "logps/chosen": -736.0, "logps/rejected": -1136.0, "loss": 0.1602, "rewards/accuracies": 0.875, "rewards/chosen": -5.8125, "rewards/margins": 4.125, "rewards/rejected": -9.9375, "step": 12760 }, { "epoch": 1.6321170681832706, "grad_norm": 10.084195714615372, "learning_rate": 4.9777372775015456e-08, "logits/chosen": -3.140625, "logits/rejected": -2.9375, "logps/chosen": -748.0, "logps/rejected": -1176.0, "loss": 0.1447, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.84375, "rewards/margins": 4.5625, "rewards/rejected": -10.375, "step": 12770 }, { "epoch": 1.633395105118538, "grad_norm": 13.263484573524702, "learning_rate": 4.9443919016630645e-08, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -748.0, "logps/rejected": -1176.0, "loss": 0.1703, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.875, "rewards/margins": 4.34375, "rewards/rejected": -10.25, "step": 12780 }, { "epoch": 1.6346731420538054, "grad_norm": 8.383512346411408, "learning_rate": 4.9111463291567125e-08, "logits/chosen": -3.09375, "logits/rejected": -2.890625, "logps/chosen": -708.0, "logps/rejected": -1256.0, "loss": 0.1559, "rewards/accuracies": 0.9375, "rewards/chosen": -5.59375, "rewards/margins": 5.5625, "rewards/rejected": -11.125, "step": 12790 }, { "epoch": 1.6359511789890728, "grad_norm": 10.580332639012939, "learning_rate": 4.8780007254234394e-08, "logits/chosen": -3.15625, "logits/rejected": -2.90625, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1439, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.46875, "rewards/rejected": -9.9375, "step": 12800 }, { "epoch": 1.6372292159243402, "grad_norm": 19.51848775776392, "learning_rate": 4.844955255406732e-08, "logits/chosen": -3.015625, "logits/rejected": -2.71875, "logps/chosen": -700.0, "logps/rejected": -1120.0, "loss": 0.1828, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.1875, "rewards/rejected": -9.8125, "step": 12810 }, { "epoch": 1.6385072528596076, "grad_norm": 11.931564428067116, "learning_rate": 4.812010083551757e-08, "logits/chosen": -3.09375, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1136.0, "loss": 0.16, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.125, "rewards/rejected": -9.625, "step": 12820 }, { "epoch": 1.639785289794875, "grad_norm": 11.081839896001414, "learning_rate": 4.779165373804586e-08, "logits/chosen": -3.125, "logits/rejected": -2.875, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1284, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.53125, "rewards/rejected": -10.0, "step": 12830 }, { "epoch": 1.6410633267301424, "grad_norm": 8.912648287414607, "learning_rate": 4.746421289611338e-08, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -752.0, "logps/rejected": -1160.0, "loss": 0.1423, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.78125, "rewards/margins": 4.25, "rewards/rejected": -10.0, "step": 12840 }, { "epoch": 1.6423413636654098, "grad_norm": 11.476184210415996, "learning_rate": 4.7137779939173893e-08, "logits/chosen": -3.125, "logits/rejected": -2.8125, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1474, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5, "rewards/margins": 4.375, "rewards/rejected": -9.875, "step": 12850 }, { "epoch": 1.6436194006006772, "grad_norm": 12.492110875247722, "learning_rate": 4.6812356491665764e-08, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1192.0, "loss": 0.1505, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.53125, "rewards/margins": 4.96875, "rewards/rejected": -10.5, "step": 12860 }, { "epoch": 1.6448974375359446, "grad_norm": 10.879943565961865, "learning_rate": 4.6487944173003476e-08, "logits/chosen": -3.046875, "logits/rejected": -2.765625, "logps/chosen": -736.0, "logps/rejected": -1128.0, "loss": 0.1442, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.71875, "rewards/margins": 4.3125, "rewards/rejected": -10.0, "step": 12870 }, { "epoch": 1.6461754744712123, "grad_norm": 15.928020637284542, "learning_rate": 4.616454459757002e-08, "logits/chosen": -3.171875, "logits/rejected": -2.90625, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1507, "rewards/accuracies": 0.96875, "rewards/chosen": -5.5, "rewards/margins": 4.59375, "rewards/rejected": -10.0625, "step": 12880 }, { "epoch": 1.6474535114064797, "grad_norm": 7.269730657997192, "learning_rate": 4.584215937470848e-08, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -724.0, "logps/rejected": -1160.0, "loss": 0.1529, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.65625, "rewards/rejected": -10.25, "step": 12890 }, { "epoch": 1.648731548341747, "grad_norm": 10.28920428876983, "learning_rate": 4.552079010871415e-08, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1585, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.5625, "rewards/rejected": -10.0, "step": 12900 }, { "epoch": 1.6500095852770145, "grad_norm": 11.316966507611001, "learning_rate": 4.520043839882684e-08, "logits/chosen": -3.0625, "logits/rejected": -2.71875, "logps/chosen": -724.0, "logps/rejected": -1120.0, "loss": 0.1616, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.625, "rewards/margins": 4.1875, "rewards/rejected": -9.8125, "step": 12910 }, { "epoch": 1.651287622212282, "grad_norm": 9.069174759597042, "learning_rate": 4.488110583922236e-08, "logits/chosen": -3.125, "logits/rejected": -2.78125, "logps/chosen": -700.0, "logps/rejected": -1160.0, "loss": 0.1694, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.46875, "rewards/rejected": -9.9375, "step": 12920 }, { "epoch": 1.6525656591475495, "grad_norm": 12.417263699859777, "learning_rate": 4.456279401900523e-08, "logits/chosen": -3.171875, "logits/rejected": -2.75, "logps/chosen": -744.0, "logps/rejected": -1176.0, "loss": 0.153, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.75, "rewards/margins": 4.4375, "rewards/rejected": -10.1875, "step": 12930 }, { "epoch": 1.653843696082817, "grad_norm": 17.65844858555798, "learning_rate": 4.424550452219994e-08, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -748.0, "logps/rejected": -1160.0, "loss": 0.1847, "rewards/accuracies": 0.9375, "rewards/chosen": -5.6875, "rewards/margins": 4.1875, "rewards/rejected": -9.875, "step": 12940 }, { "epoch": 1.6551217330180843, "grad_norm": 9.332315629733964, "learning_rate": 4.392923892774411e-08, "logits/chosen": -3.109375, "logits/rejected": -2.84375, "logps/chosen": -744.0, "logps/rejected": -1112.0, "loss": 0.1723, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.65625, "rewards/margins": 4.03125, "rewards/rejected": -9.6875, "step": 12950 }, { "epoch": 1.6563997699533517, "grad_norm": 11.230195658475209, "learning_rate": 4.361399880947972e-08, "logits/chosen": -3.09375, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1616, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 4.53125, "rewards/rejected": -9.6875, "step": 12960 }, { "epoch": 1.657677806888619, "grad_norm": 10.757072801678362, "learning_rate": 4.329978573614568e-08, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.167, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.3125, "rewards/rejected": -9.6875, "step": 12970 }, { "epoch": 1.6589558438238865, "grad_norm": 12.568097778396032, "learning_rate": 4.298660127137024e-08, "logits/chosen": -3.078125, "logits/rejected": -2.828125, "logps/chosen": -680.0, "logps/rejected": -1128.0, "loss": 0.1395, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.09375, "rewards/margins": 4.71875, "rewards/rejected": -9.8125, "step": 12980 }, { "epoch": 1.660233880759154, "grad_norm": 8.681036794407342, "learning_rate": 4.26744469736626e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1503, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.65625, "rewards/rejected": -10.125, "step": 12990 }, { "epoch": 1.6615119176944213, "grad_norm": 10.748730385170303, "learning_rate": 4.2363324396405805e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -676.0, "logps/rejected": -1072.0, "loss": 0.1422, "rewards/accuracies": 0.90625, "rewards/chosen": -5.40625, "rewards/margins": 4.03125, "rewards/rejected": -9.4375, "step": 13000 }, { "epoch": 1.6627899546296887, "grad_norm": 9.54558033538067, "learning_rate": 4.205323508784855e-08, "logits/chosen": -2.96875, "logits/rejected": -2.921875, "logps/chosen": -708.0, "logps/rejected": -1088.0, "loss": 0.1411, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.0, "rewards/rejected": -9.5, "step": 13010 }, { "epoch": 1.664067991564956, "grad_norm": 9.69882496731971, "learning_rate": 4.1744180591097605e-08, "logits/chosen": -3.09375, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1144.0, "loss": 0.1611, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.28125, "rewards/margins": 4.59375, "rewards/rejected": -9.875, "step": 13020 }, { "epoch": 1.6653460285002235, "grad_norm": 9.404584846040253, "learning_rate": 4.1436162444110386e-08, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -692.0, "logps/rejected": -1152.0, "loss": 0.1633, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.65625, "rewards/rejected": -10.0625, "step": 13030 }, { "epoch": 1.6666240654354911, "grad_norm": 9.751104808340164, "learning_rate": 4.112918217968683e-08, "logits/chosen": -3.109375, "logits/rejected": -2.75, "logps/chosen": -696.0, "logps/rejected": -1168.0, "loss": 0.1687, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.6875, "rewards/rejected": -10.125, "step": 13040 }, { "epoch": 1.6679021023707585, "grad_norm": 9.070501787724893, "learning_rate": 4.082324132546231e-08, "logits/chosen": -3.078125, "logits/rejected": -2.890625, "logps/chosen": -692.0, "logps/rejected": -1112.0, "loss": 0.1596, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.34375, "rewards/margins": 4.34375, "rewards/rejected": -9.6875, "step": 13050 }, { "epoch": 1.669180139306026, "grad_norm": 9.824126557878227, "learning_rate": 4.051834140389951e-08, "logits/chosen": -3.09375, "logits/rejected": -2.875, "logps/chosen": -704.0, "logps/rejected": -1104.0, "loss": 0.1308, "rewards/accuracies": 0.9375, "rewards/chosen": -5.5625, "rewards/margins": 4.1875, "rewards/rejected": -9.75, "step": 13060 }, { "epoch": 1.6704581762412933, "grad_norm": 13.590904771054559, "learning_rate": 4.0214483932281106e-08, "logits/chosen": -3.15625, "logits/rejected": -2.6875, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1474, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.34375, "rewards/rejected": -9.9375, "step": 13070 }, { "epoch": 1.671736213176561, "grad_norm": 10.099487968529644, "learning_rate": 3.991167042270241e-08, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -716.0, "logps/rejected": -1136.0, "loss": 0.1365, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.65625, "rewards/margins": 4.46875, "rewards/rejected": -10.125, "step": 13080 }, { "epoch": 1.6730142501118284, "grad_norm": 13.315177673304635, "learning_rate": 3.9609902382063353e-08, "logits/chosen": -3.171875, "logits/rejected": -2.96875, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1655, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.34375, "rewards/rejected": -9.875, "step": 13090 }, { "epoch": 1.6742922870470958, "grad_norm": 10.878570202665678, "learning_rate": 3.930918131206157e-08, "logits/chosen": -3.171875, "logits/rejected": -2.90625, "logps/chosen": -704.0, "logps/rejected": -1152.0, "loss": 0.1493, "rewards/accuracies": 0.9375, "rewards/chosen": -5.65625, "rewards/margins": 4.65625, "rewards/rejected": -10.3125, "step": 13100 }, { "epoch": 1.6755703239823632, "grad_norm": 10.960054640663829, "learning_rate": 3.900950870918435e-08, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -704.0, "logps/rejected": -1176.0, "loss": 0.1381, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.875, "rewards/rejected": -10.3125, "step": 13110 }, { "epoch": 1.6768483609176306, "grad_norm": 11.730572519108025, "learning_rate": 3.871088606470152e-08, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1144.0, "loss": 0.1578, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 13120 }, { "epoch": 1.678126397852898, "grad_norm": 16.351532699106595, "learning_rate": 3.841331486465818e-08, "logits/chosen": -3.109375, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1627, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 13130 }, { "epoch": 1.6794044347881654, "grad_norm": 10.074693097060957, "learning_rate": 3.81167965898668e-08, "logits/chosen": -3.09375, "logits/rejected": -2.84375, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1565, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 13140 }, { "epoch": 1.6806824717234328, "grad_norm": 10.356973881084844, "learning_rate": 3.782133271590046e-08, "logits/chosen": -3.0, "logits/rejected": -2.734375, "logps/chosen": -672.0, "logps/rejected": -1136.0, "loss": 0.124, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.25, "rewards/margins": 4.71875, "rewards/rejected": -10.0, "step": 13150 }, { "epoch": 1.6819605086587002, "grad_norm": 7.5109538996121135, "learning_rate": 3.7526924713084776e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -692.0, "logps/rejected": -1136.0, "loss": 0.146, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.3125, "rewards/rejected": -9.9375, "step": 13160 }, { "epoch": 1.6832385455939676, "grad_norm": 7.43584171219965, "learning_rate": 3.7233574046491386e-08, "logits/chosen": -2.984375, "logits/rejected": -2.6875, "logps/chosen": -744.0, "logps/rejected": -1136.0, "loss": 0.1689, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.78125, "rewards/margins": 4.1875, "rewards/rejected": -9.9375, "step": 13170 }, { "epoch": 1.684516582529235, "grad_norm": 8.091461209680139, "learning_rate": 3.694128217593012e-08, "logits/chosen": -2.984375, "logits/rejected": -2.859375, "logps/chosen": -760.0, "logps/rejected": -1168.0, "loss": 0.1472, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -6.0, "rewards/margins": 4.28125, "rewards/rejected": -10.3125, "step": 13180 }, { "epoch": 1.6857946194645024, "grad_norm": 11.278883290132143, "learning_rate": 3.665005055594183e-08, "logits/chosen": -3.078125, "logits/rejected": -2.890625, "logps/chosen": -720.0, "logps/rejected": -1152.0, "loss": 0.1577, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.6875, "rewards/margins": 4.4375, "rewards/rejected": -10.1875, "step": 13190 }, { "epoch": 1.68707265639977, "grad_norm": 7.690946330940168, "learning_rate": 3.635988063579137e-08, "logits/chosen": -3.03125, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1192.0, "loss": 0.1398, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.34375, "rewards/margins": 5.0625, "rewards/rejected": -10.4375, "step": 13200 }, { "epoch": 1.6883506933350374, "grad_norm": 14.1181961523169, "learning_rate": 3.607077385946003e-08, "logits/chosen": -3.015625, "logits/rejected": -2.921875, "logps/chosen": -732.0, "logps/rejected": -1200.0, "loss": 0.156, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.53125, "rewards/margins": 4.84375, "rewards/rejected": -10.375, "step": 13210 }, { "epoch": 1.6896287302703048, "grad_norm": 12.185505585818044, "learning_rate": 3.578273166563878e-08, "logits/chosen": -3.109375, "logits/rejected": -2.8125, "logps/chosen": -664.0, "logps/rejected": -1128.0, "loss": 0.157, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.40625, "rewards/rejected": -9.6875, "step": 13220 }, { "epoch": 1.6909067672055722, "grad_norm": 10.486951579951219, "learning_rate": 3.549575548772066e-08, "logits/chosen": -3.125, "logits/rejected": -2.984375, "logps/chosen": -732.0, "logps/rejected": -1144.0, "loss": 0.1339, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 13230 }, { "epoch": 1.6921848041408398, "grad_norm": 10.874078113401561, "learning_rate": 3.52098467537939e-08, "logits/chosen": -3.046875, "logits/rejected": -2.859375, "logps/chosen": -708.0, "logps/rejected": -1168.0, "loss": 0.1447, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.5625, "rewards/rejected": -10.125, "step": 13240 }, { "epoch": 1.6934628410761072, "grad_norm": 8.559857941272588, "learning_rate": 3.492500688663494e-08, "logits/chosen": -3.03125, "logits/rejected": -2.90625, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.1439, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.34375, "rewards/rejected": -9.5625, "step": 13250 }, { "epoch": 1.6947408780113746, "grad_norm": 11.945439184291894, "learning_rate": 3.4641237303701e-08, "logits/chosen": -3.125, "logits/rejected": -2.8125, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1382, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.5, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 13260 }, { "epoch": 1.696018914946642, "grad_norm": 10.703184305003559, "learning_rate": 3.435853941712322e-08, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1200.0, "loss": 0.1502, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.96875, "rewards/rejected": -10.4375, "step": 13270 }, { "epoch": 1.6972969518819094, "grad_norm": 9.473424218245944, "learning_rate": 3.4076914633699825e-08, "logits/chosen": -3.0625, "logits/rejected": -2.9375, "logps/chosen": -728.0, "logps/rejected": -1128.0, "loss": 0.1615, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.875, "rewards/margins": 4.25, "rewards/rejected": -10.125, "step": 13280 }, { "epoch": 1.6985749888171768, "grad_norm": 12.309402631962879, "learning_rate": 3.379636435488864e-08, "logits/chosen": -3.109375, "logits/rejected": -2.765625, "logps/chosen": -688.0, "logps/rejected": -1200.0, "loss": 0.1562, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.4375, "rewards/margins": 5.09375, "rewards/rejected": -10.5, "step": 13290 }, { "epoch": 1.6998530257524442, "grad_norm": 12.714540982337835, "learning_rate": 3.351688997680069e-08, "logits/chosen": -3.09375, "logits/rejected": -2.921875, "logps/chosen": -728.0, "logps/rejected": -1168.0, "loss": 0.1582, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.6875, "rewards/margins": 4.40625, "rewards/rejected": -10.125, "step": 13300 }, { "epoch": 1.7011310626877116, "grad_norm": 12.042285791699674, "learning_rate": 3.323849289019273e-08, "logits/chosen": -3.078125, "logits/rejected": -2.921875, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1688, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.625, "rewards/margins": 4.3125, "rewards/rejected": -9.9375, "step": 13310 }, { "epoch": 1.702409099622979, "grad_norm": 8.951963033610214, "learning_rate": 3.296117448046062e-08, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -748.0, "logps/rejected": -1136.0, "loss": 0.1688, "rewards/accuracies": 0.90625, "rewards/chosen": -5.875, "rewards/margins": 4.125, "rewards/rejected": -10.0, "step": 13320 }, { "epoch": 1.7036871365582464, "grad_norm": 8.891445504358778, "learning_rate": 3.268493612763246e-08, "logits/chosen": -3.109375, "logits/rejected": -2.796875, "logps/chosen": -712.0, "logps/rejected": -1144.0, "loss": 0.1341, "rewards/accuracies": 0.90625, "rewards/chosen": -5.34375, "rewards/margins": 4.5, "rewards/rejected": -9.875, "step": 13330 }, { "epoch": 1.7049651734935138, "grad_norm": 9.866638871693574, "learning_rate": 3.240977920636148e-08, "logits/chosen": -3.1875, "logits/rejected": -2.828125, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1541, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.1875, "rewards/rejected": -9.875, "step": 13340 }, { "epoch": 1.7062432104287812, "grad_norm": 15.709750146528357, "learning_rate": 3.2135705085919534e-08, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -728.0, "logps/rejected": -1160.0, "loss": 0.158, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.71875, "rewards/margins": 4.40625, "rewards/rejected": -10.125, "step": 13350 }, { "epoch": 1.7075212473640489, "grad_norm": 14.42470319889697, "learning_rate": 3.186271513018993e-08, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -728.0, "logps/rejected": -1160.0, "loss": 0.1594, "rewards/accuracies": 0.9375, "rewards/chosen": -5.6875, "rewards/margins": 4.5625, "rewards/rejected": -10.25, "step": 13360 }, { "epoch": 1.7087992842993163, "grad_norm": 8.930022776253411, "learning_rate": 3.15908106976609e-08, "logits/chosen": -2.96875, "logits/rejected": -2.703125, "logps/chosen": -672.0, "logps/rejected": -1152.0, "loss": 0.1438, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.125, "rewards/margins": 4.875, "rewards/rejected": -10.0, "step": 13370 }, { "epoch": 1.7100773212345837, "grad_norm": 10.266994731297885, "learning_rate": 3.131999314141873e-08, "logits/chosen": -2.984375, "logits/rejected": -2.6875, "logps/chosen": -712.0, "logps/rejected": -1184.0, "loss": 0.1345, "rewards/accuracies": 0.9375, "rewards/chosen": -5.59375, "rewards/margins": 4.5625, "rewards/rejected": -10.1875, "step": 13380 }, { "epoch": 1.711355358169851, "grad_norm": 7.991014015948914, "learning_rate": 3.1050263809141e-08, "logits/chosen": -3.15625, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1176.0, "loss": 0.1413, "rewards/accuracies": 0.9375, "rewards/chosen": -5.5625, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 13390 }, { "epoch": 1.7126333951051187, "grad_norm": 9.533028636727256, "learning_rate": 3.078162404309009e-08, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1437, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.625, "rewards/margins": 4.3125, "rewards/rejected": -9.9375, "step": 13400 }, { "epoch": 1.713911432040386, "grad_norm": 10.166254139871006, "learning_rate": 3.051407518010612e-08, "logits/chosen": -3.0, "logits/rejected": -2.796875, "logps/chosen": -668.0, "logps/rejected": -1128.0, "loss": 0.1374, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.71875, "rewards/rejected": -9.9375, "step": 13410 }, { "epoch": 1.7151894689756535, "grad_norm": 20.501536359023945, "learning_rate": 3.024761855160077e-08, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -736.0, "logps/rejected": -1176.0, "loss": 0.1432, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.625, "rewards/margins": 4.59375, "rewards/rejected": -10.1875, "step": 13420 }, { "epoch": 1.716467505910921, "grad_norm": 12.16897294342861, "learning_rate": 2.9982255483550154e-08, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1642, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.5, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 13430 }, { "epoch": 1.7177455428461883, "grad_norm": 7.095714092157576, "learning_rate": 2.9717987296488555e-08, "logits/chosen": -3.109375, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1144.0, "loss": 0.164, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.75, "rewards/rejected": -10.0625, "step": 13440 }, { "epoch": 1.7190235797814557, "grad_norm": 10.085330881143296, "learning_rate": 2.9454815305501867e-08, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -700.0, "logps/rejected": -1184.0, "loss": 0.15, "rewards/accuracies": 0.9375, "rewards/chosen": -5.375, "rewards/margins": 4.90625, "rewards/rejected": -10.3125, "step": 13450 }, { "epoch": 1.720301616716723, "grad_norm": 10.080544930632923, "learning_rate": 2.9192740820220773e-08, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -696.0, "logps/rejected": -1104.0, "loss": 0.1782, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.25, "rewards/rejected": -9.6875, "step": 13460 }, { "epoch": 1.7215796536519905, "grad_norm": 11.512873439831605, "learning_rate": 2.893176514481463e-08, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -732.0, "logps/rejected": -1120.0, "loss": 0.1597, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.71875, "rewards/margins": 3.984375, "rewards/rejected": -9.6875, "step": 13470 }, { "epoch": 1.722857690587258, "grad_norm": 16.09918469753171, "learning_rate": 2.8671889577984527e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1817, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.65625, "rewards/rejected": -10.0625, "step": 13480 }, { "epoch": 1.7241357275225253, "grad_norm": 8.307363455025255, "learning_rate": 2.8413115412957127e-08, "logits/chosen": -2.953125, "logits/rejected": -2.859375, "logps/chosen": -696.0, "logps/rejected": -1128.0, "loss": 0.1534, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.5, "rewards/rejected": -9.8125, "step": 13490 }, { "epoch": 1.7254137644577927, "grad_norm": 14.318859134640787, "learning_rate": 2.8155443937478264e-08, "logits/chosen": -3.09375, "logits/rejected": -2.640625, "logps/chosen": -684.0, "logps/rejected": -1160.0, "loss": 0.165, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.8125, "rewards/rejected": -10.125, "step": 13500 }, { "epoch": 1.72669180139306, "grad_norm": 8.121559167464675, "learning_rate": 2.78988764338062e-08, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1144.0, "loss": 0.1377, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.34375, "rewards/margins": 4.5625, "rewards/rejected": -9.9375, "step": 13510 }, { "epoch": 1.7279698383283277, "grad_norm": 11.152849450848047, "learning_rate": 2.7643414178705748e-08, "logits/chosen": -3.125, "logits/rejected": -2.890625, "logps/chosen": -660.0, "logps/rejected": -1072.0, "loss": 0.1507, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.21875, "rewards/margins": 4.125, "rewards/rejected": -9.375, "step": 13520 }, { "epoch": 1.7292478752635951, "grad_norm": 9.953380185633824, "learning_rate": 2.738905844344144e-08, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -660.0, "logps/rejected": -1080.0, "loss": 0.1679, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.125, "rewards/rejected": -9.375, "step": 13530 }, { "epoch": 1.7305259121988625, "grad_norm": 9.098672192531483, "learning_rate": 2.7135810493771366e-08, "logits/chosen": -2.890625, "logits/rejected": -2.78125, "logps/chosen": -688.0, "logps/rejected": -1056.0, "loss": 0.1465, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.21875, "rewards/margins": 3.953125, "rewards/rejected": -9.1875, "step": 13540 }, { "epoch": 1.73180394913413, "grad_norm": 13.080306160681328, "learning_rate": 2.688367158994115e-08, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -724.0, "logps/rejected": -1112.0, "loss": 0.1576, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5, "rewards/margins": 4.1875, "rewards/rejected": -9.6875, "step": 13550 }, { "epoch": 1.7330819860693976, "grad_norm": 11.270529218973435, "learning_rate": 2.663264298667711e-08, "logits/chosen": -2.9375, "logits/rejected": -2.796875, "logps/chosen": -732.0, "logps/rejected": -1128.0, "loss": 0.1499, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.625, "rewards/margins": 4.1875, "rewards/rejected": -9.8125, "step": 13560 }, { "epoch": 1.734360023004665, "grad_norm": 10.513696283143453, "learning_rate": 2.638272593318072e-08, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -672.0, "logps/rejected": -1120.0, "loss": 0.1435, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.3125, "rewards/margins": 4.4375, "rewards/rejected": -9.75, "step": 13570 }, { "epoch": 1.7356380599399324, "grad_norm": 8.904662666522817, "learning_rate": 2.613392167312159e-08, "logits/chosen": -2.90625, "logits/rejected": -2.6875, "logps/chosen": -712.0, "logps/rejected": -1136.0, "loss": 0.1451, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.40625, "rewards/rejected": -10.0, "step": 13580 }, { "epoch": 1.7369160968751998, "grad_norm": 6.71524883761725, "learning_rate": 2.588623144463206e-08, "logits/chosen": -3.046875, "logits/rejected": -2.59375, "logps/chosen": -712.0, "logps/rejected": -1224.0, "loss": 0.1465, "rewards/accuracies": 0.9375, "rewards/chosen": -5.65625, "rewards/margins": 4.90625, "rewards/rejected": -10.5625, "step": 13590 }, { "epoch": 1.7381941338104672, "grad_norm": 7.921988234050129, "learning_rate": 2.563965648030053e-08, "logits/chosen": -2.90625, "logits/rejected": -2.734375, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1424, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.375, "rewards/rejected": -9.8125, "step": 13600 }, { "epoch": 1.7394721707457346, "grad_norm": 10.377797310654964, "learning_rate": 2.539419800716541e-08, "logits/chosen": -3.03125, "logits/rejected": -2.828125, "logps/chosen": -688.0, "logps/rejected": -1160.0, "loss": 0.1698, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.6875, "rewards/rejected": -10.125, "step": 13610 }, { "epoch": 1.740750207681002, "grad_norm": 10.564238013114593, "learning_rate": 2.5149857246709262e-08, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -692.0, "logps/rejected": -1144.0, "loss": 0.1525, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.34375, "rewards/margins": 4.59375, "rewards/rejected": -9.9375, "step": 13620 }, { "epoch": 1.7420282446162694, "grad_norm": 11.163471055282269, "learning_rate": 2.4906635414852328e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -704.0, "logps/rejected": -1176.0, "loss": 0.1483, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.8125, "rewards/rejected": -10.25, "step": 13630 }, { "epoch": 1.7433062815515368, "grad_norm": 7.886940638043434, "learning_rate": 2.4664533721946906e-08, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -692.0, "logps/rejected": -1128.0, "loss": 0.1346, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 13640 }, { "epoch": 1.7445843184868042, "grad_norm": 8.139565850536563, "learning_rate": 2.4423553372770995e-08, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -680.0, "logps/rejected": -1096.0, "loss": 0.1431, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.3125, "rewards/rejected": -9.625, "step": 13650 }, { "epoch": 1.7458623554220716, "grad_norm": 10.93235959546461, "learning_rate": 2.4183695566522332e-08, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -728.0, "logps/rejected": -1112.0, "loss": 0.146, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.6875, "rewards/margins": 3.953125, "rewards/rejected": -9.625, "step": 13660 }, { "epoch": 1.747140392357339, "grad_norm": 12.129375239791715, "learning_rate": 2.3944961496812744e-08, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -672.0, "logps/rejected": -1120.0, "loss": 0.159, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.6875, "rewards/rejected": -9.875, "step": 13670 }, { "epoch": 1.7484184292926066, "grad_norm": 9.591038353306839, "learning_rate": 2.370735235166174e-08, "logits/chosen": -3.09375, "logits/rejected": -2.734375, "logps/chosen": -700.0, "logps/rejected": -1160.0, "loss": 0.1538, "rewards/accuracies": 0.96875, "rewards/chosen": -5.5, "rewards/margins": 4.8125, "rewards/rejected": -10.3125, "step": 13680 }, { "epoch": 1.749696466227874, "grad_norm": 8.64814249707982, "learning_rate": 2.3470869313490987e-08, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -672.0, "logps/rejected": -1096.0, "loss": 0.1526, "rewards/accuracies": 0.9375, "rewards/chosen": -5.21875, "rewards/margins": 4.40625, "rewards/rejected": -9.625, "step": 13690 }, { "epoch": 1.7509745031631414, "grad_norm": 12.638604359599785, "learning_rate": 2.3235513559118207e-08, "logits/chosen": -3.15625, "logits/rejected": -2.75, "logps/chosen": -660.0, "logps/rejected": -1128.0, "loss": 0.1536, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.15625, "rewards/margins": 4.84375, "rewards/rejected": -10.0, "step": 13700 }, { "epoch": 1.7522525400984088, "grad_norm": 9.80199416228049, "learning_rate": 2.3001286259751313e-08, "logits/chosen": -3.046875, "logits/rejected": -2.890625, "logps/chosen": -680.0, "logps/rejected": -1088.0, "loss": 0.143, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.3125, "rewards/margins": 4.25, "rewards/rejected": -9.5625, "step": 13710 }, { "epoch": 1.7535305770336764, "grad_norm": 11.838193078041444, "learning_rate": 2.2768188580982866e-08, "logits/chosen": -3.125, "logits/rejected": -2.90625, "logps/chosen": -720.0, "logps/rejected": -1104.0, "loss": 0.1622, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.65625, "rewards/margins": 4.0, "rewards/rejected": -9.6875, "step": 13720 }, { "epoch": 1.7548086139689438, "grad_norm": 11.25848518919155, "learning_rate": 2.2536221682783794e-08, "logits/chosen": -3.03125, "logits/rejected": -2.78125, "logps/chosen": -712.0, "logps/rejected": -1144.0, "loss": 0.1411, "rewards/accuracies": 0.9375, "rewards/chosen": -5.59375, "rewards/margins": 4.5, "rewards/rejected": -10.125, "step": 13730 }, { "epoch": 1.7560866509042112, "grad_norm": 10.808883838407986, "learning_rate": 2.2305386719498165e-08, "logits/chosen": -3.171875, "logits/rejected": -2.953125, "logps/chosen": -688.0, "logps/rejected": -1160.0, "loss": 0.155, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.28125, "rewards/margins": 4.875, "rewards/rejected": -10.125, "step": 13740 }, { "epoch": 1.7573646878394786, "grad_norm": 12.822469636687895, "learning_rate": 2.2075684839836933e-08, "logits/chosen": -3.03125, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1152.0, "loss": 0.1472, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.25, "rewards/margins": 4.75, "rewards/rejected": -10.0, "step": 13750 }, { "epoch": 1.758642724774746, "grad_norm": 10.580264723553709, "learning_rate": 2.1847117186872516e-08, "logits/chosen": -3.09375, "logits/rejected": -2.90625, "logps/chosen": -696.0, "logps/rejected": -1096.0, "loss": 0.1519, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.15625, "rewards/rejected": -9.5625, "step": 13760 }, { "epoch": 1.7599207617100134, "grad_norm": 13.29768195211083, "learning_rate": 2.1619684898033174e-08, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -720.0, "logps/rejected": -1184.0, "loss": 0.1479, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.8125, "rewards/rejected": -10.3125, "step": 13770 }, { "epoch": 1.7611987986452808, "grad_norm": 9.244016852518765, "learning_rate": 2.1393389105097026e-08, "logits/chosen": -2.96875, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1104.0, "loss": 0.1529, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5, "rewards/margins": 4.21875, "rewards/rejected": -9.6875, "step": 13780 }, { "epoch": 1.7624768355805482, "grad_norm": 8.150376865571264, "learning_rate": 2.1168230934186793e-08, "logits/chosen": -3.015625, "logits/rejected": -2.921875, "logps/chosen": -696.0, "logps/rejected": -1120.0, "loss": 0.1483, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.34375, "rewards/rejected": -9.75, "step": 13790 }, { "epoch": 1.7637548725158156, "grad_norm": 8.674885967333253, "learning_rate": 2.094421150576381e-08, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -716.0, "logps/rejected": -1120.0, "loss": 0.1353, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.3125, "rewards/rejected": -9.75, "step": 13800 }, { "epoch": 1.765032909451083, "grad_norm": 10.827112291973835, "learning_rate": 2.072133193462286e-08, "logits/chosen": -3.09375, "logits/rejected": -2.875, "logps/chosen": -732.0, "logps/rejected": -1168.0, "loss": 0.1453, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 4.71875, "rewards/rejected": -10.1875, "step": 13810 }, { "epoch": 1.7663109463863504, "grad_norm": 11.82887235352997, "learning_rate": 2.0499593329886293e-08, "logits/chosen": -3.0625, "logits/rejected": -2.734375, "logps/chosen": -704.0, "logps/rejected": -1176.0, "loss": 0.1395, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.46875, "rewards/margins": 4.6875, "rewards/rejected": -10.125, "step": 13820 }, { "epoch": 1.7675889833216178, "grad_norm": 12.534723631836071, "learning_rate": 2.0278996794998616e-08, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -748.0, "logps/rejected": -1152.0, "loss": 0.1487, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.78125, "rewards/margins": 4.21875, "rewards/rejected": -10.0, "step": 13830 }, { "epoch": 1.7688670202568855, "grad_norm": 17.467097863848547, "learning_rate": 2.0059543427721177e-08, "logits/chosen": -2.9375, "logits/rejected": -2.828125, "logps/chosen": -716.0, "logps/rejected": -1112.0, "loss": 0.1671, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.09375, "rewards/rejected": -9.625, "step": 13840 }, { "epoch": 1.7701450571921529, "grad_norm": 9.588420485753964, "learning_rate": 1.9841234320126356e-08, "logits/chosen": -3.078125, "logits/rejected": -2.796875, "logps/chosen": -740.0, "logps/rejected": -1184.0, "loss": 0.1525, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.6875, "rewards/margins": 4.5, "rewards/rejected": -10.1875, "step": 13850 }, { "epoch": 1.7714230941274203, "grad_norm": 14.998857660958935, "learning_rate": 1.9624070558592444e-08, "logits/chosen": -3.09375, "logits/rejected": -2.703125, "logps/chosen": -692.0, "logps/rejected": -1176.0, "loss": 0.1446, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.875, "rewards/rejected": -10.3125, "step": 13860 }, { "epoch": 1.7727011310626877, "grad_norm": 11.498974088453098, "learning_rate": 1.9408053223798044e-08, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -704.0, "logps/rejected": -1152.0, "loss": 0.1605, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.5, "rewards/margins": 4.5, "rewards/rejected": -10.0, "step": 13870 }, { "epoch": 1.7739791679979553, "grad_norm": 11.34559480835344, "learning_rate": 1.9193183390716768e-08, "logits/chosen": -3.046875, "logits/rejected": -2.734375, "logps/chosen": -716.0, "logps/rejected": -1160.0, "loss": 0.1512, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.59375, "rewards/rejected": -10.1875, "step": 13880 }, { "epoch": 1.7752572049332227, "grad_norm": 15.385336092658148, "learning_rate": 1.8979462128611957e-08, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -680.0, "logps/rejected": -1152.0, "loss": 0.1687, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.78125, "rewards/rejected": -10.1875, "step": 13890 }, { "epoch": 1.77653524186849, "grad_norm": 9.563306457518955, "learning_rate": 1.8766890501031142e-08, "logits/chosen": -3.140625, "logits/rejected": -2.765625, "logps/chosen": -720.0, "logps/rejected": -1168.0, "loss": 0.1571, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.625, "rewards/rejected": -10.1875, "step": 13900 }, { "epoch": 1.7778132788037575, "grad_norm": 11.95208502989548, "learning_rate": 1.8555469565801062e-08, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1152.0, "loss": 0.1475, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.4375, "rewards/margins": 4.5, "rewards/rejected": -9.9375, "step": 13910 }, { "epoch": 1.779091315739025, "grad_norm": 12.686984192311087, "learning_rate": 1.834520037502213e-08, "logits/chosen": -3.046875, "logits/rejected": -2.65625, "logps/chosen": -688.0, "logps/rejected": -1128.0, "loss": 0.1642, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.375, "rewards/rejected": -10.0, "step": 13920 }, { "epoch": 1.7803693526742923, "grad_norm": 9.574801828792555, "learning_rate": 1.8136083975063193e-08, "logits/chosen": -3.0625, "logits/rejected": -2.859375, "logps/chosen": -732.0, "logps/rejected": -1152.0, "loss": 0.1481, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.5625, "rewards/margins": 4.46875, "rewards/rejected": -10.0625, "step": 13930 }, { "epoch": 1.7816473896095597, "grad_norm": 11.876562802028962, "learning_rate": 1.7928121406556696e-08, "logits/chosen": -3.0, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1088.0, "loss": 0.1597, "rewards/accuracies": 0.875, "rewards/chosen": -5.375, "rewards/margins": 4.21875, "rewards/rejected": -9.5625, "step": 13940 }, { "epoch": 1.782925426544827, "grad_norm": 10.726637400908194, "learning_rate": 1.7721313704392977e-08, "logits/chosen": -3.046875, "logits/rejected": -2.734375, "logps/chosen": -704.0, "logps/rejected": -1160.0, "loss": 0.1529, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.75, "rewards/rejected": -10.25, "step": 13950 }, { "epoch": 1.7842034634800945, "grad_norm": 13.423853508050184, "learning_rate": 1.751566189771561e-08, "logits/chosen": -2.984375, "logits/rejected": -2.796875, "logps/chosen": -684.0, "logps/rejected": -1112.0, "loss": 0.143, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.21875, "rewards/margins": 4.53125, "rewards/rejected": -9.75, "step": 13960 }, { "epoch": 1.785481500415362, "grad_norm": 10.942400408166316, "learning_rate": 1.7311167009915833e-08, "logits/chosen": -3.046875, "logits/rejected": -2.859375, "logps/chosen": -696.0, "logps/rejected": -1152.0, "loss": 0.1715, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.3125, "rewards/margins": 4.625, "rewards/rejected": -9.9375, "step": 13970 }, { "epoch": 1.7867595373506293, "grad_norm": 12.61138052341031, "learning_rate": 1.7107830058627758e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -696.0, "logps/rejected": -1104.0, "loss": 0.1483, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 4.1875, "rewards/rejected": -9.625, "step": 13980 }, { "epoch": 1.7880375742858967, "grad_norm": 9.296731970428597, "learning_rate": 1.6905652055723334e-08, "logits/chosen": -3.046875, "logits/rejected": -2.921875, "logps/chosen": -696.0, "logps/rejected": -1088.0, "loss": 0.1677, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.25, "rewards/rejected": -9.5625, "step": 13990 }, { "epoch": 1.7893156112211643, "grad_norm": 17.27131332653722, "learning_rate": 1.6704634007307068e-08, "logits/chosen": -3.0625, "logits/rejected": -2.890625, "logps/chosen": -708.0, "logps/rejected": -1144.0, "loss": 0.1597, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.53125, "rewards/margins": 4.46875, "rewards/rejected": -10.0, "step": 14000 }, { "epoch": 1.7905936481564317, "grad_norm": 13.643812551024203, "learning_rate": 1.6504776913711132e-08, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1080.0, "loss": 0.1521, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5, "rewards/margins": 4.03125, "rewards/rejected": -9.5, "step": 14010 }, { "epoch": 1.7918716850916991, "grad_norm": 9.285428705838198, "learning_rate": 1.6306081769490443e-08, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -680.0, "logps/rejected": -1136.0, "loss": 0.1317, "rewards/accuracies": 0.9375, "rewards/chosen": -5.40625, "rewards/margins": 4.4375, "rewards/rejected": -9.8125, "step": 14020 }, { "epoch": 1.7931497220269665, "grad_norm": 11.831120919568061, "learning_rate": 1.6108549563417756e-08, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -672.0, "logps/rejected": -1152.0, "loss": 0.1536, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.15625, "rewards/margins": 4.9375, "rewards/rejected": -10.0625, "step": 14030 }, { "epoch": 1.7944277589622342, "grad_norm": 11.774157531421034, "learning_rate": 1.5912181278478597e-08, "logits/chosen": -3.0, "logits/rejected": -2.78125, "logps/chosen": -692.0, "logps/rejected": -1136.0, "loss": 0.1561, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 14040 }, { "epoch": 1.7957057958975016, "grad_norm": 12.111032807950124, "learning_rate": 1.5716977891866352e-08, "logits/chosen": -3.0, "logits/rejected": -2.75, "logps/chosen": -700.0, "logps/rejected": -1128.0, "loss": 0.1532, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.375, "rewards/margins": 4.28125, "rewards/rejected": -9.6875, "step": 14050 }, { "epoch": 1.796983832832769, "grad_norm": 12.49771071125514, "learning_rate": 1.5522940374977618e-08, "logits/chosen": -3.078125, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1160.0, "loss": 0.1382, "rewards/accuracies": 0.981249988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.90625, "rewards/rejected": -10.125, "step": 14060 }, { "epoch": 1.7982618697680364, "grad_norm": 15.091897925618236, "learning_rate": 1.5330069693407165e-08, "logits/chosen": -3.125, "logits/rejected": -2.96875, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1538, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.625, "rewards/margins": 4.34375, "rewards/rejected": -10.0, "step": 14070 }, { "epoch": 1.7995399067033038, "grad_norm": 15.012264420287543, "learning_rate": 1.5138366806943302e-08, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1160.0, "loss": 0.1699, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.5, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 14080 }, { "epoch": 1.8008179436385712, "grad_norm": 12.879249742583966, "learning_rate": 1.4947832669562844e-08, "logits/chosen": -3.15625, "logits/rejected": -2.765625, "logps/chosen": -704.0, "logps/rejected": -1120.0, "loss": 0.1372, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.34375, "rewards/rejected": -9.875, "step": 14090 }, { "epoch": 1.8020959805738386, "grad_norm": 10.117385918315096, "learning_rate": 1.4758468229426612e-08, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -712.0, "logps/rejected": -1152.0, "loss": 0.156, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.59375, "rewards/rejected": -10.125, "step": 14100 }, { "epoch": 1.803374017509106, "grad_norm": 10.434715411629035, "learning_rate": 1.4570274428874679e-08, "logits/chosen": -3.03125, "logits/rejected": -2.921875, "logps/chosen": -724.0, "logps/rejected": -1160.0, "loss": 0.1402, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5, "rewards/margins": 4.6875, "rewards/rejected": -10.1875, "step": 14110 }, { "epoch": 1.8046520544443734, "grad_norm": 10.257712949256526, "learning_rate": 1.438325220442152e-08, "logits/chosen": -3.046875, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -1120.0, "loss": 0.1672, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.3125, "rewards/rejected": -9.75, "step": 14120 }, { "epoch": 1.8059300913796408, "grad_norm": 8.785574136152228, "learning_rate": 1.4197402486751581e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -704.0, "logps/rejected": -1160.0, "loss": 0.162, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.75, "rewards/rejected": -10.0625, "step": 14130 }, { "epoch": 1.8072081283149082, "grad_norm": 12.447729841189695, "learning_rate": 1.401272620071442e-08, "logits/chosen": -3.046875, "logits/rejected": -2.84375, "logps/chosen": -684.0, "logps/rejected": -1160.0, "loss": 0.1643, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.34375, "rewards/margins": 4.875, "rewards/rejected": -10.1875, "step": 14140 }, { "epoch": 1.8084861652501756, "grad_norm": 10.070325070925708, "learning_rate": 1.3829224265320211e-08, "logits/chosen": -3.046875, "logits/rejected": -2.78125, "logps/chosen": -680.0, "logps/rejected": -1120.0, "loss": 0.1333, "rewards/accuracies": 0.9375, "rewards/chosen": -5.1875, "rewards/margins": 4.71875, "rewards/rejected": -9.875, "step": 14150 }, { "epoch": 1.8097642021854432, "grad_norm": 12.903234521283592, "learning_rate": 1.3646897593735273e-08, "logits/chosen": -3.09375, "logits/rejected": -2.671875, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1624, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 14160 }, { "epoch": 1.8110422391207106, "grad_norm": 12.686883856670724, "learning_rate": 1.346574709327733e-08, "logits/chosen": -3.078125, "logits/rejected": -2.84375, "logps/chosen": -716.0, "logps/rejected": -1152.0, "loss": 0.1534, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.5625, "rewards/rejected": -10.25, "step": 14170 }, { "epoch": 1.812320276055978, "grad_norm": 14.22861593842286, "learning_rate": 1.3285773665411143e-08, "logits/chosen": -3.0, "logits/rejected": -2.59375, "logps/chosen": -692.0, "logps/rejected": -1184.0, "loss": 0.1636, "rewards/accuracies": 0.9375, "rewards/chosen": -5.4375, "rewards/margins": 5.125, "rewards/rejected": -10.5625, "step": 14180 }, { "epoch": 1.8135983129912454, "grad_norm": 9.789271909311166, "learning_rate": 1.3106978205743968e-08, "logits/chosen": -2.953125, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1674, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.25, "rewards/rejected": -9.6875, "step": 14190 }, { "epoch": 1.814876349926513, "grad_norm": 12.328876632604587, "learning_rate": 1.2929361604021028e-08, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -696.0, "logps/rejected": -1160.0, "loss": 0.1405, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.53125, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 14200 }, { "epoch": 1.8161543868617804, "grad_norm": 7.992402341168219, "learning_rate": 1.275292474412129e-08, "logits/chosen": -3.0, "logits/rejected": -2.703125, "logps/chosen": -676.0, "logps/rejected": -1104.0, "loss": 0.1485, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.25, "rewards/margins": 4.5, "rewards/rejected": -9.75, "step": 14210 }, { "epoch": 1.8174324237970478, "grad_norm": 11.691821378605601, "learning_rate": 1.2577668504052813e-08, "logits/chosen": -3.125, "logits/rejected": -2.796875, "logps/chosen": -700.0, "logps/rejected": -1208.0, "loss": 0.1412, "rewards/accuracies": 0.96875, "rewards/chosen": -5.375, "rewards/margins": 4.9375, "rewards/rejected": -10.3125, "step": 14220 }, { "epoch": 1.8187104607323152, "grad_norm": 12.155431721595809, "learning_rate": 1.2403593755948572e-08, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -680.0, "logps/rejected": -1152.0, "loss": 0.1481, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.75, "rewards/rejected": -10.0625, "step": 14230 }, { "epoch": 1.8199884976675826, "grad_norm": 9.91908769920145, "learning_rate": 1.2230701366061969e-08, "logits/chosen": -3.0, "logits/rejected": -2.890625, "logps/chosen": -728.0, "logps/rejected": -1136.0, "loss": 0.1344, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.34375, "rewards/rejected": -9.875, "step": 14240 }, { "epoch": 1.82126653460285, "grad_norm": 9.339881834605144, "learning_rate": 1.2058992194762724e-08, "logits/chosen": -3.046875, "logits/rejected": -2.828125, "logps/chosen": -696.0, "logps/rejected": -1120.0, "loss": 0.1335, "rewards/accuracies": 0.9375, "rewards/chosen": -5.25, "rewards/margins": 4.40625, "rewards/rejected": -9.6875, "step": 14250 }, { "epoch": 1.8225445715381174, "grad_norm": 11.10352886253188, "learning_rate": 1.1888467096532406e-08, "logits/chosen": -3.078125, "logits/rejected": -2.96875, "logps/chosen": -736.0, "logps/rejected": -1088.0, "loss": 0.1716, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.75, "rewards/margins": 3.703125, "rewards/rejected": -9.4375, "step": 14260 }, { "epoch": 1.8238226084733848, "grad_norm": 14.93846998373106, "learning_rate": 1.1719126919960187e-08, "logits/chosen": -3.078125, "logits/rejected": -2.828125, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1538, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.375, "rewards/margins": 4.5625, "rewards/rejected": -9.9375, "step": 14270 }, { "epoch": 1.8251006454086522, "grad_norm": 10.202931139450149, "learning_rate": 1.1550972507738815e-08, "logits/chosen": -3.09375, "logits/rejected": -2.859375, "logps/chosen": -732.0, "logps/rejected": -1144.0, "loss": 0.1596, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5625, "rewards/margins": 4.46875, "rewards/rejected": -10.0625, "step": 14280 }, { "epoch": 1.8263786823439196, "grad_norm": 8.27863731117087, "learning_rate": 1.1384004696660148e-08, "logits/chosen": -3.015625, "logits/rejected": -2.703125, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.1565, "rewards/accuracies": 0.9375, "rewards/chosen": -5.5, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 14290 }, { "epoch": 1.827656719279187, "grad_norm": 12.30629947950159, "learning_rate": 1.1218224317611159e-08, "logits/chosen": -3.03125, "logits/rejected": -2.703125, "logps/chosen": -692.0, "logps/rejected": -1168.0, "loss": 0.1425, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.625, "rewards/rejected": -10.1875, "step": 14300 }, { "epoch": 1.8289347562144544, "grad_norm": 6.7009280288802735, "learning_rate": 1.1053632195569845e-08, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -704.0, "logps/rejected": -1168.0, "loss": 0.1359, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.53125, "rewards/margins": 4.75, "rewards/rejected": -10.3125, "step": 14310 }, { "epoch": 1.830212793149722, "grad_norm": 14.503511364519225, "learning_rate": 1.0890229149600943e-08, "logits/chosen": -3.078125, "logits/rejected": -2.8125, "logps/chosen": -740.0, "logps/rejected": -1184.0, "loss": 0.1624, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.84375, "rewards/margins": 4.34375, "rewards/rejected": -10.1875, "step": 14320 }, { "epoch": 1.8314908300849895, "grad_norm": 13.359834100545635, "learning_rate": 1.072801599285203e-08, "logits/chosen": -2.984375, "logits/rejected": -2.921875, "logps/chosen": -696.0, "logps/rejected": -1208.0, "loss": 0.1619, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.40625, "rewards/margins": 5.125, "rewards/rejected": -10.5625, "step": 14330 }, { "epoch": 1.8327688670202569, "grad_norm": 9.795144374083144, "learning_rate": 1.0566993532549363e-08, "logits/chosen": -3.015625, "logits/rejected": -2.78125, "logps/chosen": -720.0, "logps/rejected": -1152.0, "loss": 0.1464, "rewards/accuracies": 0.96875, "rewards/chosen": -5.625, "rewards/margins": 4.5, "rewards/rejected": -10.125, "step": 14340 }, { "epoch": 1.8340469039555243, "grad_norm": 11.966682565809306, "learning_rate": 1.0407162569993855e-08, "logits/chosen": -3.125, "logits/rejected": -2.703125, "logps/chosen": -724.0, "logps/rejected": -1160.0, "loss": 0.1441, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.71875, "rewards/margins": 4.46875, "rewards/rejected": -10.1875, "step": 14350 }, { "epoch": 1.835324940890792, "grad_norm": 11.976332678053666, "learning_rate": 1.024852390055722e-08, "logits/chosen": -2.921875, "logits/rejected": -2.78125, "logps/chosen": -708.0, "logps/rejected": -1112.0, "loss": 0.1437, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.40625, "rewards/margins": 4.40625, "rewards/rejected": -9.8125, "step": 14360 }, { "epoch": 1.8366029778260593, "grad_norm": 11.073959106533652, "learning_rate": 1.0091078313677892e-08, "logits/chosen": -2.8125, "logits/rejected": -2.734375, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.1351, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.28125, "rewards/margins": 4.53125, "rewards/rejected": -9.8125, "step": 14370 }, { "epoch": 1.8378810147613267, "grad_norm": 13.967643551885132, "learning_rate": 9.93482659285716e-09, "logits/chosen": -3.0625, "logits/rejected": -2.875, "logps/chosen": -708.0, "logps/rejected": -1080.0, "loss": 0.1459, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.3125, "rewards/margins": 4.125, "rewards/rejected": -9.4375, "step": 14380 }, { "epoch": 1.839159051696594, "grad_norm": 12.418189122707291, "learning_rate": 9.779769515655179e-09, "logits/chosen": -3.140625, "logits/rejected": -2.859375, "logps/chosen": -676.0, "logps/rejected": -1120.0, "loss": 0.1549, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.1875, "rewards/margins": 4.59375, "rewards/rejected": -9.8125, "step": 14390 }, { "epoch": 1.8404370886318615, "grad_norm": 10.408033015979928, "learning_rate": 9.625907853687165e-09, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -708.0, "logps/rejected": -1112.0, "loss": 0.1621, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.28125, "rewards/rejected": -9.8125, "step": 14400 }, { "epoch": 1.841715125567129, "grad_norm": 9.263369906030954, "learning_rate": 9.473242372619621e-09, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -724.0, "logps/rejected": -1144.0, "loss": 0.1467, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.5625, "rewards/margins": 4.34375, "rewards/rejected": -9.875, "step": 14410 }, { "epoch": 1.8429931625023963, "grad_norm": 12.739122397711153, "learning_rate": 9.321773832166397e-09, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -716.0, "logps/rejected": -1168.0, "loss": 0.1606, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.4375, "rewards/margins": 4.4375, "rewards/rejected": -9.875, "step": 14420 }, { "epoch": 1.8442711994376637, "grad_norm": 11.839371678365538, "learning_rate": 9.171502986085072e-09, "logits/chosen": -2.953125, "logits/rejected": -2.875, "logps/chosen": -732.0, "logps/rejected": -1152.0, "loss": 0.1584, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.71875, "rewards/margins": 4.5, "rewards/rejected": -10.1875, "step": 14430 }, { "epoch": 1.8455492363729311, "grad_norm": 12.856022816191095, "learning_rate": 9.022430582172919e-09, "logits/chosen": -3.0625, "logits/rejected": -2.71875, "logps/chosen": -700.0, "logps/rejected": -1104.0, "loss": 0.1533, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.1875, "rewards/rejected": -9.6875, "step": 14440 }, { "epoch": 1.8468272733081985, "grad_norm": 8.926987752379693, "learning_rate": 8.874557362263529e-09, "logits/chosen": -3.0625, "logits/rejected": -2.75, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1439, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.40625, "rewards/rejected": -9.8125, "step": 14450 }, { "epoch": 1.848105310243466, "grad_norm": 12.864364750595266, "learning_rate": 8.727884062222885e-09, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -708.0, "logps/rejected": -1136.0, "loss": 0.1666, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 14460 }, { "epoch": 1.8493833471787333, "grad_norm": 12.647684825072547, "learning_rate": 8.582411411945768e-09, "logits/chosen": -3.015625, "logits/rejected": -2.734375, "logps/chosen": -724.0, "logps/rejected": -1104.0, "loss": 0.1537, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.65625, "rewards/margins": 4.03125, "rewards/rejected": -9.6875, "step": 14470 }, { "epoch": 1.850661384114001, "grad_norm": 11.822844846768222, "learning_rate": 8.438140135352211e-09, "logits/chosen": -2.9375, "logits/rejected": -2.546875, "logps/chosen": -704.0, "logps/rejected": -1160.0, "loss": 0.1717, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.4375, "rewards/margins": 4.75, "rewards/rejected": -10.1875, "step": 14480 }, { "epoch": 1.8519394210492683, "grad_norm": 13.036845289997391, "learning_rate": 8.295070950383647e-09, "logits/chosen": -3.140625, "logits/rejected": -2.875, "logps/chosen": -712.0, "logps/rejected": -1096.0, "loss": 0.1613, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.53125, "rewards/margins": 3.921875, "rewards/rejected": -9.4375, "step": 14490 }, { "epoch": 1.8532174579845357, "grad_norm": 9.44512506950682, "learning_rate": 8.153204568999733e-09, "logits/chosen": -3.1875, "logits/rejected": -2.921875, "logps/chosen": -664.0, "logps/rejected": -1144.0, "loss": 0.1337, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.125, "rewards/margins": 4.875, "rewards/rejected": -10.0, "step": 14500 }, { "epoch": 1.8544954949198031, "grad_norm": 10.391187249363702, "learning_rate": 8.012541697174414e-09, "logits/chosen": -2.953125, "logits/rejected": -2.703125, "logps/chosen": -700.0, "logps/rejected": -1080.0, "loss": 0.1514, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.34375, "rewards/margins": 4.125, "rewards/rejected": -9.4375, "step": 14510 }, { "epoch": 1.8557735318550708, "grad_norm": 14.04567191399965, "learning_rate": 7.873083034892537e-09, "logits/chosen": -3.03125, "logits/rejected": -2.75, "logps/chosen": -728.0, "logps/rejected": -1176.0, "loss": 0.1443, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.65625, "rewards/rejected": -10.1875, "step": 14520 }, { "epoch": 1.8570515687903382, "grad_norm": 13.633018900248839, "learning_rate": 7.734829276146576e-09, "logits/chosen": -2.921875, "logits/rejected": -2.609375, "logps/chosen": -692.0, "logps/rejected": -1144.0, "loss": 0.1552, "rewards/accuracies": 0.9375, "rewards/chosen": -5.34375, "rewards/margins": 4.46875, "rewards/rejected": -9.8125, "step": 14530 }, { "epoch": 1.8583296057256056, "grad_norm": 14.81761718347149, "learning_rate": 7.59778110893286e-09, "logits/chosen": -3.015625, "logits/rejected": -2.859375, "logps/chosen": -704.0, "logps/rejected": -1088.0, "loss": 0.1421, "rewards/accuracies": 0.90625, "rewards/chosen": -5.4375, "rewards/margins": 4.0, "rewards/rejected": -9.4375, "step": 14540 }, { "epoch": 1.859607642660873, "grad_norm": 8.88382465270999, "learning_rate": 7.461939215248403e-09, "logits/chosen": -3.046875, "logits/rejected": -2.734375, "logps/chosen": -704.0, "logps/rejected": -1120.0, "loss": 0.1575, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5, "rewards/margins": 4.28125, "rewards/rejected": -9.8125, "step": 14550 }, { "epoch": 1.8608856795961404, "grad_norm": 10.324327735645847, "learning_rate": 7.327304271087298e-09, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -724.0, "logps/rejected": -1096.0, "loss": 0.1674, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.78125, "rewards/margins": 3.890625, "rewards/rejected": -9.6875, "step": 14560 }, { "epoch": 1.8621637165314078, "grad_norm": 11.893562246279691, "learning_rate": 7.1938769464375e-09, "logits/chosen": -3.140625, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1192.0, "loss": 0.1663, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.375, "rewards/margins": 5.03125, "rewards/rejected": -10.375, "step": 14570 }, { "epoch": 1.8634417534666752, "grad_norm": 12.310902619295284, "learning_rate": 7.061657905277518e-09, "logits/chosen": -3.09375, "logits/rejected": -2.71875, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1505, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.4375, "rewards/rejected": -9.875, "step": 14580 }, { "epoch": 1.8647197904019426, "grad_norm": 12.099849619715219, "learning_rate": 6.9306478055729254e-09, "logits/chosen": -3.046875, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1104.0, "loss": 0.1593, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.5, "rewards/margins": 4.125, "rewards/rejected": -9.625, "step": 14590 }, { "epoch": 1.86599782733721, "grad_norm": 16.938882516568054, "learning_rate": 6.800847299273327e-09, "logits/chosen": -3.03125, "logits/rejected": -2.796875, "logps/chosen": -728.0, "logps/rejected": -1144.0, "loss": 0.1653, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.375, "rewards/rejected": -9.9375, "step": 14600 }, { "epoch": 1.8672758642724774, "grad_norm": 11.587208367743546, "learning_rate": 6.672257032308864e-09, "logits/chosen": -2.9375, "logits/rejected": -2.8125, "logps/chosen": -700.0, "logps/rejected": -1080.0, "loss": 0.1547, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.34375, "rewards/margins": 4.0, "rewards/rejected": -9.375, "step": 14610 }, { "epoch": 1.8685539012077448, "grad_norm": 11.619646400256674, "learning_rate": 6.544877644587193e-09, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -716.0, "logps/rejected": -1104.0, "loss": 0.1465, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.6875, "rewards/margins": 3.984375, "rewards/rejected": -9.6875, "step": 14620 }, { "epoch": 1.8698319381430122, "grad_norm": 12.217282596762695, "learning_rate": 6.418709769990316e-09, "logits/chosen": -2.921875, "logits/rejected": -2.78125, "logps/chosen": -688.0, "logps/rejected": -1096.0, "loss": 0.1638, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.375, "rewards/margins": 4.125, "rewards/rejected": -9.5, "step": 14630 }, { "epoch": 1.8711099750782798, "grad_norm": 10.678100699980412, "learning_rate": 6.293754036371141e-09, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -736.0, "logps/rejected": -1144.0, "loss": 0.1413, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.59375, "rewards/rejected": -10.0, "step": 14640 }, { "epoch": 1.8723880120135472, "grad_norm": 16.210974014943545, "learning_rate": 6.170011065550818e-09, "logits/chosen": -2.953125, "logits/rejected": -2.78125, "logps/chosen": -668.0, "logps/rejected": -1088.0, "loss": 0.1387, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.25, "rewards/margins": 4.4375, "rewards/rejected": -9.6875, "step": 14650 }, { "epoch": 1.8736660489488146, "grad_norm": 10.127056266258883, "learning_rate": 6.047481473315102e-09, "logits/chosen": -3.03125, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1112.0, "loss": 0.1491, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.125, "rewards/rejected": -9.5625, "step": 14660 }, { "epoch": 1.874944085884082, "grad_norm": 11.71800474936891, "learning_rate": 5.926165869411859e-09, "logits/chosen": -2.984375, "logits/rejected": -2.71875, "logps/chosen": -672.0, "logps/rejected": -1096.0, "loss": 0.1671, "rewards/accuracies": 0.9375, "rewards/chosen": -5.09375, "rewards/margins": 4.5, "rewards/rejected": -9.625, "step": 14670 }, { "epoch": 1.8762221228193496, "grad_norm": 10.427549178650835, "learning_rate": 5.8060648575475326e-09, "logits/chosen": -2.9375, "logits/rejected": -2.625, "logps/chosen": -692.0, "logps/rejected": -1096.0, "loss": 0.1589, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.40625, "rewards/rejected": -9.75, "step": 14680 }, { "epoch": 1.877500159754617, "grad_norm": 9.623257167478432, "learning_rate": 5.68717903538446e-09, "logits/chosen": -3.0, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1144.0, "loss": 0.1331, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.625, "rewards/rejected": -10.0625, "step": 14690 }, { "epoch": 1.8787781966898844, "grad_norm": 10.184536050278489, "learning_rate": 5.569508994537814e-09, "logits/chosen": -2.90625, "logits/rejected": -2.765625, "logps/chosen": -728.0, "logps/rejected": -1104.0, "loss": 0.144, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.59375, "rewards/margins": 4.15625, "rewards/rejected": -9.75, "step": 14700 }, { "epoch": 1.8800562336251518, "grad_norm": 11.21262756503898, "learning_rate": 5.453055320572525e-09, "logits/chosen": -2.921875, "logits/rejected": -2.734375, "logps/chosen": -732.0, "logps/rejected": -1144.0, "loss": 0.1438, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.53125, "rewards/rejected": -9.9375, "step": 14710 }, { "epoch": 1.8813342705604192, "grad_norm": 8.276067538916859, "learning_rate": 5.337818593000615e-09, "logits/chosen": -3.0625, "logits/rejected": -2.84375, "logps/chosen": -676.0, "logps/rejected": -1120.0, "loss": 0.1583, "rewards/accuracies": 0.9375, "rewards/chosen": -5.21875, "rewards/margins": 4.65625, "rewards/rejected": -9.8125, "step": 14720 }, { "epoch": 1.8826123074956866, "grad_norm": 13.181049739902091, "learning_rate": 5.22379938527806e-09, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -704.0, "logps/rejected": -1128.0, "loss": 0.1453, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.375, "rewards/margins": 4.375, "rewards/rejected": -9.75, "step": 14730 }, { "epoch": 1.883890344430954, "grad_norm": 13.959442989381273, "learning_rate": 5.1109982648021296e-09, "logits/chosen": -3.03125, "logits/rejected": -2.640625, "logps/chosen": -724.0, "logps/rejected": -1120.0, "loss": 0.1455, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.6875, "rewards/margins": 4.0625, "rewards/rejected": -9.75, "step": 14740 }, { "epoch": 1.8851683813662214, "grad_norm": 13.495500872308062, "learning_rate": 4.999415792908495e-09, "logits/chosen": -3.078125, "logits/rejected": -2.859375, "logps/chosen": -704.0, "logps/rejected": -1120.0, "loss": 0.1599, "rewards/accuracies": 0.9375, "rewards/chosen": -5.40625, "rewards/margins": 4.375, "rewards/rejected": -9.75, "step": 14750 }, { "epoch": 1.8864464183014888, "grad_norm": 14.023588667374646, "learning_rate": 4.889052524868348e-09, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1312, "rewards/accuracies": 0.96875, "rewards/chosen": -5.3125, "rewards/margins": 4.5, "rewards/rejected": -9.8125, "step": 14760 }, { "epoch": 1.8877244552367562, "grad_norm": 7.947339683870802, "learning_rate": 4.779909009885841e-09, "logits/chosen": -3.125, "logits/rejected": -2.78125, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1537, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.46875, "rewards/rejected": -10.0, "step": 14770 }, { "epoch": 1.8890024921720237, "grad_norm": 10.536940544419787, "learning_rate": 4.6719857910950955e-09, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -680.0, "logps/rejected": -1208.0, "loss": 0.1319, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.375, "rewards/margins": 5.1875, "rewards/rejected": -10.5625, "step": 14780 }, { "epoch": 1.890280529107291, "grad_norm": 11.818957062933153, "learning_rate": 4.565283405557757e-09, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -736.0, "logps/rejected": -1168.0, "loss": 0.139, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.625, "rewards/margins": 4.46875, "rewards/rejected": -10.125, "step": 14790 }, { "epoch": 1.8915585660425587, "grad_norm": 9.286164362041376, "learning_rate": 4.4598023842601085e-09, "logits/chosen": -2.984375, "logits/rejected": -2.734375, "logps/chosen": -704.0, "logps/rejected": -1128.0, "loss": 0.1468, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.3125, "rewards/rejected": -9.75, "step": 14800 }, { "epoch": 1.892836602977826, "grad_norm": 11.296676357308892, "learning_rate": 4.3555432521105996e-09, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -728.0, "logps/rejected": -1160.0, "loss": 0.1618, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 14810 }, { "epoch": 1.8941146399130935, "grad_norm": 13.00937492775704, "learning_rate": 4.2525065279370995e-09, "logits/chosen": -3.0, "logits/rejected": -2.828125, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.164, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.28125, "rewards/rejected": -9.75, "step": 14820 }, { "epoch": 1.8953926768483609, "grad_norm": 10.15410344253207, "learning_rate": 4.1506927244844004e-09, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -716.0, "logps/rejected": -1128.0, "loss": 0.1349, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.40625, "rewards/rejected": -9.875, "step": 14830 }, { "epoch": 1.8966707137836285, "grad_norm": 9.119767621505575, "learning_rate": 4.050102348411605e-09, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -688.0, "logps/rejected": -1120.0, "loss": 0.1548, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.46875, "rewards/rejected": -9.75, "step": 14840 }, { "epoch": 1.897948750718896, "grad_norm": 14.469398940553326, "learning_rate": 3.9507359002897145e-09, "logits/chosen": -3.171875, "logits/rejected": -2.953125, "logps/chosen": -716.0, "logps/rejected": -1136.0, "loss": 0.1709, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.65625, "rewards/margins": 4.25, "rewards/rejected": -9.9375, "step": 14850 }, { "epoch": 1.8992267876541633, "grad_norm": 12.231871395300658, "learning_rate": 3.852593874598936e-09, "logits/chosen": -2.953125, "logits/rejected": -2.734375, "logps/chosen": -748.0, "logps/rejected": -1128.0, "loss": 0.1719, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.875, "rewards/margins": 3.921875, "rewards/rejected": -9.8125, "step": 14860 }, { "epoch": 1.9005048245894307, "grad_norm": 12.073682499643972, "learning_rate": 3.755676759726434e-09, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -696.0, "logps/rejected": -1176.0, "loss": 0.1477, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.5, "rewards/margins": 4.8125, "rewards/rejected": -10.3125, "step": 14870 }, { "epoch": 1.901782861524698, "grad_norm": 12.282004370609691, "learning_rate": 3.6599850379638032e-09, "logits/chosen": -3.015625, "logits/rejected": -2.796875, "logps/chosen": -676.0, "logps/rejected": -1120.0, "loss": 0.1394, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.21875, "rewards/margins": 4.5, "rewards/rejected": -9.75, "step": 14880 }, { "epoch": 1.9030608984599655, "grad_norm": 10.207448871691478, "learning_rate": 3.565519185504684e-09, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -696.0, "logps/rejected": -1160.0, "loss": 0.1445, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.5, "rewards/rejected": -10.0, "step": 14890 }, { "epoch": 1.904338935395233, "grad_norm": 11.856070963380358, "learning_rate": 3.4722796724423466e-09, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -700.0, "logps/rejected": -1136.0, "loss": 0.1632, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.375, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 14900 }, { "epoch": 1.9056169723305003, "grad_norm": 14.22172415536922, "learning_rate": 3.380266962767386e-09, "logits/chosen": -3.078125, "logits/rejected": -2.65625, "logps/chosen": -692.0, "logps/rejected": -1192.0, "loss": 0.1582, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.96875, "rewards/rejected": -10.375, "step": 14910 }, { "epoch": 1.9068950092657677, "grad_norm": 13.602757040527935, "learning_rate": 3.289481514365533e-09, "logits/chosen": -3.015625, "logits/rejected": -2.6875, "logps/chosen": -708.0, "logps/rejected": -1192.0, "loss": 0.1476, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.34375, "rewards/margins": 5.09375, "rewards/rejected": -10.375, "step": 14920 }, { "epoch": 1.9081730462010351, "grad_norm": 17.815448615194278, "learning_rate": 3.199923779015068e-09, "logits/chosen": -3.03125, "logits/rejected": -2.734375, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1674, "rewards/accuracies": 0.887499988079071, "rewards/chosen": -5.5, "rewards/margins": 4.46875, "rewards/rejected": -10.0, "step": 14930 }, { "epoch": 1.9094510831363025, "grad_norm": 9.167336092535807, "learning_rate": 3.111594202384937e-09, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1128.0, "loss": 0.1453, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.375, "rewards/margins": 4.4375, "rewards/rejected": -9.8125, "step": 14940 }, { "epoch": 1.91072912007157, "grad_norm": 15.831129982327335, "learning_rate": 3.0244932240323095e-09, "logits/chosen": -3.015625, "logits/rejected": -2.765625, "logps/chosen": -712.0, "logps/rejected": -1152.0, "loss": 0.1571, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.53125, "rewards/rejected": -10.0, "step": 14950 }, { "epoch": 1.9120071570068375, "grad_norm": 10.519254389733671, "learning_rate": 2.938621277400355e-09, "logits/chosen": -3.09375, "logits/rejected": -2.875, "logps/chosen": -708.0, "logps/rejected": -1096.0, "loss": 0.1486, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.15625, "rewards/rejected": -9.625, "step": 14960 }, { "epoch": 1.913285193942105, "grad_norm": 12.955231124203314, "learning_rate": 2.8539787898163302e-09, "logits/chosen": -3.078125, "logits/rejected": -2.6875, "logps/chosen": -720.0, "logps/rejected": -1184.0, "loss": 0.1818, "rewards/accuracies": 0.96875, "rewards/chosen": -5.5, "rewards/margins": 4.625, "rewards/rejected": -10.125, "step": 14970 }, { "epoch": 1.9145632308773723, "grad_norm": 14.75785834916094, "learning_rate": 2.7705661824891922e-09, "logits/chosen": -3.109375, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1136.0, "loss": 0.1616, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.53125, "rewards/rejected": -9.9375, "step": 14980 }, { "epoch": 1.9158412678126397, "grad_norm": 11.390718009084507, "learning_rate": 2.6883838705076265e-09, "logits/chosen": -2.90625, "logits/rejected": -2.75, "logps/chosen": -736.0, "logps/rejected": -1168.0, "loss": 0.1499, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.75, "rewards/margins": 4.1875, "rewards/rejected": -9.9375, "step": 14990 }, { "epoch": 1.9171193047479074, "grad_norm": 9.648512751461917, "learning_rate": 2.607432262837966e-09, "logits/chosen": -2.921875, "logits/rejected": -2.84375, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1498, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.375, "rewards/margins": 4.625, "rewards/rejected": -10.0, "step": 15000 }, { "epoch": 1.9183973416831748, "grad_norm": 10.772529811188837, "learning_rate": 2.5277117623221654e-09, "logits/chosen": -3.0625, "logits/rejected": -2.828125, "logps/chosen": -676.0, "logps/rejected": -1080.0, "loss": 0.1565, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.375, "rewards/margins": 4.15625, "rewards/rejected": -9.5, "step": 15010 }, { "epoch": 1.9196753786184422, "grad_norm": 11.968154707014914, "learning_rate": 2.4492227656757736e-09, "logits/chosen": -3.09375, "logits/rejected": -2.828125, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1557, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.53125, "rewards/margins": 4.21875, "rewards/rejected": -9.75, "step": 15020 }, { "epoch": 1.9209534155537096, "grad_norm": 12.502294203547288, "learning_rate": 2.3719656634859642e-09, "logits/chosen": -3.015625, "logits/rejected": -2.90625, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1561, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.375, "rewards/rejected": -9.8125, "step": 15030 }, { "epoch": 1.922231452488977, "grad_norm": 8.959430299711036, "learning_rate": 2.2959408402096202e-09, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -692.0, "logps/rejected": -1136.0, "loss": 0.1575, "rewards/accuracies": 0.9375, "rewards/chosen": -5.34375, "rewards/margins": 4.46875, "rewards/rejected": -9.8125, "step": 15040 }, { "epoch": 1.9235094894242444, "grad_norm": 10.710421147055664, "learning_rate": 2.2211486741713635e-09, "logits/chosen": -3.078125, "logits/rejected": -2.890625, "logps/chosen": -672.0, "logps/rejected": -1152.0, "loss": 0.1467, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.3125, "rewards/margins": 4.78125, "rewards/rejected": -10.0625, "step": 15050 }, { "epoch": 1.9247875263595118, "grad_norm": 11.312570556276544, "learning_rate": 2.1475895375617226e-09, "logits/chosen": -3.09375, "logits/rejected": -2.75, "logps/chosen": -712.0, "logps/rejected": -1160.0, "loss": 0.1636, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.5, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 15060 }, { "epoch": 1.9260655632947792, "grad_norm": 12.268507744807593, "learning_rate": 2.0752637964352727e-09, "logits/chosen": -3.078125, "logits/rejected": -2.8125, "logps/chosen": -720.0, "logps/rejected": -1160.0, "loss": 0.1621, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.53125, "rewards/rejected": -10.0625, "step": 15070 }, { "epoch": 1.9273436002300466, "grad_norm": 8.086827719122715, "learning_rate": 2.0041718107088046e-09, "logits/chosen": -3.015625, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1184.0, "loss": 0.1201, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.8125, "rewards/rejected": -10.3125, "step": 15080 }, { "epoch": 1.928621637165314, "grad_norm": 10.64959916901656, "learning_rate": 1.9343139341595204e-09, "logits/chosen": -3.0, "logits/rejected": -2.796875, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1603, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.4375, "rewards/rejected": -9.9375, "step": 15090 }, { "epoch": 1.9298996741005814, "grad_norm": 11.142977135987163, "learning_rate": 1.865690514423257e-09, "logits/chosen": -3.140625, "logits/rejected": -2.921875, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.1431, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.5, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 15100 }, { "epoch": 1.9311777110358488, "grad_norm": 13.70615553224399, "learning_rate": 1.7983018929929039e-09, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -736.0, "logps/rejected": -1144.0, "loss": 0.1518, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.625, "rewards/margins": 4.34375, "rewards/rejected": -9.9375, "step": 15110 }, { "epoch": 1.9324557479711164, "grad_norm": 17.93121904701502, "learning_rate": 1.7321484052164325e-09, "logits/chosen": -3.09375, "logits/rejected": -2.8125, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1596, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.34375, "rewards/rejected": -9.875, "step": 15120 }, { "epoch": 1.9337337849063838, "grad_norm": 11.71594907844557, "learning_rate": 1.6672303802954812e-09, "logits/chosen": -3.015625, "logits/rejected": -2.65625, "logps/chosen": -732.0, "logps/rejected": -1152.0, "loss": 0.1679, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.65625, "rewards/margins": 4.28125, "rewards/rejected": -9.9375, "step": 15130 }, { "epoch": 1.9350118218416512, "grad_norm": 9.520084659517018, "learning_rate": 1.603548141283606e-09, "logits/chosen": -3.03125, "logits/rejected": -2.71875, "logps/chosen": -704.0, "logps/rejected": -1104.0, "loss": 0.1493, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.59375, "rewards/margins": 4.0, "rewards/rejected": -9.625, "step": 15140 }, { "epoch": 1.9362898587769186, "grad_norm": 7.489687357486243, "learning_rate": 1.5411020050846434e-09, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1541, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.375, "rewards/margins": 4.53125, "rewards/rejected": -9.9375, "step": 15150 }, { "epoch": 1.9375678957121862, "grad_norm": 14.027073581679955, "learning_rate": 1.4798922824512116e-09, "logits/chosen": -3.09375, "logits/rejected": -2.75, "logps/chosen": -720.0, "logps/rejected": -1152.0, "loss": 0.154, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5625, "rewards/margins": 4.40625, "rewards/rejected": -10.0, "step": 15160 }, { "epoch": 1.9388459326474536, "grad_norm": 8.37642421142221, "learning_rate": 1.4199192779831282e-09, "logits/chosen": -3.03125, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1144.0, "loss": 0.1431, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.4375, "rewards/margins": 4.5, "rewards/rejected": -9.9375, "step": 15170 }, { "epoch": 1.940123969582721, "grad_norm": 10.76061449217435, "learning_rate": 1.3611832901258558e-09, "logits/chosen": -3.0, "logits/rejected": -2.65625, "logps/chosen": -716.0, "logps/rejected": -1168.0, "loss": 0.146, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.59375, "rewards/margins": 4.6875, "rewards/rejected": -10.25, "step": 15180 }, { "epoch": 1.9414020065179884, "grad_norm": 13.39651566636036, "learning_rate": 1.3036846111690869e-09, "logits/chosen": -2.984375, "logits/rejected": -2.78125, "logps/chosen": -704.0, "logps/rejected": -1112.0, "loss": 0.1502, "rewards/accuracies": 0.8687499761581421, "rewards/chosen": -5.34375, "rewards/margins": 4.375, "rewards/rejected": -9.6875, "step": 15190 }, { "epoch": 1.9426800434532558, "grad_norm": 16.455139746642296, "learning_rate": 1.247423527245217e-09, "logits/chosen": -3.078125, "logits/rejected": -2.875, "logps/chosen": -676.0, "logps/rejected": -1160.0, "loss": 0.1469, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.25, "rewards/margins": 4.875, "rewards/rejected": -10.125, "step": 15200 }, { "epoch": 1.9439580803885232, "grad_norm": 13.102136764965527, "learning_rate": 1.192400318328013e-09, "logits/chosen": -3.03125, "logits/rejected": -2.921875, "logps/chosen": -712.0, "logps/rejected": -1128.0, "loss": 0.135, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.3125, "rewards/rejected": -9.8125, "step": 15210 }, { "epoch": 1.9452361173237906, "grad_norm": 14.701240584022592, "learning_rate": 1.138615258231168e-09, "logits/chosen": -3.015625, "logits/rejected": -2.828125, "logps/chosen": -756.0, "logps/rejected": -1160.0, "loss": 0.1822, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.90625, "rewards/margins": 4.09375, "rewards/rejected": -10.0, "step": 15220 }, { "epoch": 1.946514154259058, "grad_norm": 10.082463775879928, "learning_rate": 1.086068614606861e-09, "logits/chosen": -3.03125, "logits/rejected": -2.671875, "logps/chosen": -712.0, "logps/rejected": -1168.0, "loss": 0.1408, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.625, "rewards/margins": 4.34375, "rewards/rejected": -10.0, "step": 15230 }, { "epoch": 1.9477921911943255, "grad_norm": 11.289323866928568, "learning_rate": 1.0347606489445881e-09, "logits/chosen": -2.96875, "logits/rejected": -2.765625, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1522, "rewards/accuracies": 0.90625, "rewards/chosen": -5.71875, "rewards/margins": 4.09375, "rewards/rejected": -9.8125, "step": 15240 }, { "epoch": 1.9490702281295929, "grad_norm": 10.055014544286356, "learning_rate": 9.846916165696938e-10, "logits/chosen": -3.015625, "logits/rejected": -2.84375, "logps/chosen": -732.0, "logps/rejected": -1144.0, "loss": 0.1408, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.28125, "rewards/rejected": -9.8125, "step": 15250 }, { "epoch": 1.9503482650648603, "grad_norm": 15.40463618483445, "learning_rate": 9.358617666422875e-10, "logits/chosen": -3.078125, "logits/rejected": -2.734375, "logps/chosen": -712.0, "logps/rejected": -1144.0, "loss": 0.1509, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.4375, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 15260 }, { "epoch": 1.9516263020001277, "grad_norm": 9.301535906865904, "learning_rate": 8.882713421557997e-10, "logits/chosen": -3.109375, "logits/rejected": -2.84375, "logps/chosen": -716.0, "logps/rejected": -1128.0, "loss": 0.1533, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.78125, "rewards/margins": 4.09375, "rewards/rejected": -9.875, "step": 15270 }, { "epoch": 1.9529043389353953, "grad_norm": 11.068625667949027, "learning_rate": 8.419205799359009e-10, "logits/chosen": -3.15625, "logits/rejected": -2.953125, "logps/chosen": -680.0, "logps/rejected": -1112.0, "loss": 0.1533, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.34375, "rewards/margins": 4.40625, "rewards/rejected": -9.75, "step": 15280 }, { "epoch": 1.9541823758706627, "grad_norm": 10.016924649285729, "learning_rate": 7.968097106393345e-10, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -680.0, "logps/rejected": -1168.0, "loss": 0.1705, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.3125, "rewards/margins": 5.0, "rewards/rejected": -10.3125, "step": 15290 }, { "epoch": 1.95546041280593, "grad_norm": 13.01372115011469, "learning_rate": 7.529389587526413e-10, "logits/chosen": -3.140625, "logits/rejected": -2.8125, "logps/chosen": -704.0, "logps/rejected": -1176.0, "loss": 0.1491, "rewards/accuracies": 0.96875, "rewards/chosen": -5.5, "rewards/margins": 4.75, "rewards/rejected": -10.25, "step": 15300 }, { "epoch": 1.9567384497411975, "grad_norm": 10.0163137391518, "learning_rate": 7.103085425912147e-10, "logits/chosen": -2.9375, "logits/rejected": -2.671875, "logps/chosen": -668.0, "logps/rejected": -1120.0, "loss": 0.1503, "rewards/accuracies": 0.96875, "rewards/chosen": -5.125, "rewards/margins": 4.78125, "rewards/rejected": -9.875, "step": 15310 }, { "epoch": 1.958016486676465, "grad_norm": 16.423064610005266, "learning_rate": 6.689186742980524e-10, "logits/chosen": -3.0, "logits/rejected": -2.859375, "logps/chosen": -688.0, "logps/rejected": -1120.0, "loss": 0.1521, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.4375, "rewards/rejected": -9.8125, "step": 15320 }, { "epoch": 1.9592945236117325, "grad_norm": 13.916020179023214, "learning_rate": 6.287695598428399e-10, "logits/chosen": -3.03125, "logits/rejected": -2.8125, "logps/chosen": -712.0, "logps/rejected": -1160.0, "loss": 0.1542, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.5625, "rewards/rejected": -10.0, "step": 15330 }, { "epoch": 1.960572560547, "grad_norm": 9.588268395911374, "learning_rate": 5.898613990208412e-10, "logits/chosen": -2.984375, "logits/rejected": -2.765625, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1403, "rewards/accuracies": 0.9375, "rewards/chosen": -5.625, "rewards/margins": 4.125, "rewards/rejected": -9.75, "step": 15340 }, { "epoch": 1.9618505974822673, "grad_norm": 13.784001571321943, "learning_rate": 5.521943854518984e-10, "logits/chosen": -2.921875, "logits/rejected": -2.78125, "logps/chosen": -728.0, "logps/rejected": -1128.0, "loss": 0.1405, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.5625, "rewards/margins": 4.3125, "rewards/rejected": -9.875, "step": 15350 }, { "epoch": 1.9631286344175347, "grad_norm": 9.205782297316095, "learning_rate": 5.157687065795446e-10, "logits/chosen": -3.125, "logits/rejected": -2.828125, "logps/chosen": -708.0, "logps/rejected": -1128.0, "loss": 0.1602, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 15360 }, { "epoch": 1.9644066713528021, "grad_norm": 13.261409322792284, "learning_rate": 4.805845436699763e-10, "logits/chosen": -2.953125, "logits/rejected": -2.71875, "logps/chosen": -712.0, "logps/rejected": -1144.0, "loss": 0.1592, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.375, "rewards/rejected": -9.8125, "step": 15370 }, { "epoch": 1.9656847082880695, "grad_norm": 13.788978001801002, "learning_rate": 4.4664207181119295e-10, "logits/chosen": -3.046875, "logits/rejected": -2.8125, "logps/chosen": -700.0, "logps/rejected": -1168.0, "loss": 0.1638, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.4375, "rewards/margins": 4.6875, "rewards/rejected": -10.125, "step": 15380 }, { "epoch": 1.966962745223337, "grad_norm": 14.237004738621629, "learning_rate": 4.1394145991219267e-10, "logits/chosen": -3.09375, "logits/rejected": -2.90625, "logps/chosen": -704.0, "logps/rejected": -1136.0, "loss": 0.1578, "rewards/accuracies": 0.9125000238418579, "rewards/chosen": -5.4375, "rewards/margins": 4.5, "rewards/rejected": -9.9375, "step": 15390 }, { "epoch": 1.9682407821586043, "grad_norm": 10.140151106144453, "learning_rate": 3.8248287070200003e-10, "logits/chosen": -2.9375, "logits/rejected": -2.65625, "logps/chosen": -708.0, "logps/rejected": -1120.0, "loss": 0.1583, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.5, "rewards/margins": 4.34375, "rewards/rejected": -9.8125, "step": 15400 }, { "epoch": 1.9695188190938717, "grad_norm": 14.432458275606095, "learning_rate": 3.5226646072894497e-10, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -692.0, "logps/rejected": -1128.0, "loss": 0.153, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.28125, "rewards/margins": 4.5625, "rewards/rejected": -9.8125, "step": 15410 }, { "epoch": 1.9707968560291391, "grad_norm": 11.184958679299635, "learning_rate": 3.232923803598575e-10, "logits/chosen": -3.09375, "logits/rejected": -2.796875, "logps/chosen": -672.0, "logps/rejected": -1104.0, "loss": 0.1508, "rewards/accuracies": 0.96875, "rewards/chosen": -5.28125, "rewards/margins": 4.4375, "rewards/rejected": -9.6875, "step": 15420 }, { "epoch": 1.9720748929644065, "grad_norm": 11.322731302817918, "learning_rate": 2.955607737793464e-10, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1588, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.40625, "rewards/margins": 4.46875, "rewards/rejected": -9.875, "step": 15430 }, { "epoch": 1.9733529298996741, "grad_norm": 9.0128125279139, "learning_rate": 2.690717789890773e-10, "logits/chosen": -3.0625, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1136.0, "loss": 0.1289, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.40625, "rewards/rejected": -9.9375, "step": 15440 }, { "epoch": 1.9746309668349415, "grad_norm": 7.468856109935237, "learning_rate": 2.4382552780696787e-10, "logits/chosen": -3.0625, "logits/rejected": -2.859375, "logps/chosen": -684.0, "logps/rejected": -1120.0, "loss": 0.1648, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.25, "rewards/margins": 4.4375, "rewards/rejected": -9.6875, "step": 15450 }, { "epoch": 1.975909003770209, "grad_norm": 14.22098149523964, "learning_rate": 2.1982214586679927e-10, "logits/chosen": -3.03125, "logits/rejected": -2.84375, "logps/chosen": -712.0, "logps/rejected": -1168.0, "loss": 0.1495, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.40625, "rewards/margins": 4.59375, "rewards/rejected": -10.0, "step": 15460 }, { "epoch": 1.9771870407054764, "grad_norm": 12.531352602719545, "learning_rate": 1.9706175261724467e-10, "logits/chosen": -3.109375, "logits/rejected": -2.859375, "logps/chosen": -716.0, "logps/rejected": -1120.0, "loss": 0.1508, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.5, "rewards/margins": 4.3125, "rewards/rejected": -9.8125, "step": 15470 }, { "epoch": 1.978465077640744, "grad_norm": 12.129161571300925, "learning_rate": 1.7554446132159172e-10, "logits/chosen": -3.0625, "logits/rejected": -2.8125, "logps/chosen": -720.0, "logps/rejected": -1128.0, "loss": 0.1534, "rewards/accuracies": 0.9375, "rewards/chosen": -5.46875, "rewards/margins": 4.3125, "rewards/rejected": -9.8125, "step": 15480 }, { "epoch": 1.9797431145760114, "grad_norm": 10.829943019749386, "learning_rate": 1.5527037905699315e-10, "logits/chosen": -3.03125, "logits/rejected": -2.765625, "logps/chosen": -696.0, "logps/rejected": -1176.0, "loss": 0.1454, "rewards/accuracies": 0.956250011920929, "rewards/chosen": -5.46875, "rewards/margins": 4.84375, "rewards/rejected": -10.3125, "step": 15490 }, { "epoch": 1.9810211515112788, "grad_norm": 10.449380680038518, "learning_rate": 1.362396067139393e-10, "logits/chosen": -2.96875, "logits/rejected": -2.734375, "logps/chosen": -704.0, "logps/rejected": -1152.0, "loss": 0.1609, "rewards/accuracies": 0.949999988079071, "rewards/chosen": -5.375, "rewards/margins": 4.6875, "rewards/rejected": -10.0625, "step": 15500 }, { "epoch": 1.9822991884465462, "grad_norm": 10.15870473477103, "learning_rate": 1.1845223899586975e-10, "logits/chosen": -3.046875, "logits/rejected": -2.734375, "logps/chosen": -720.0, "logps/rejected": -1184.0, "loss": 0.1576, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.46875, "rewards/margins": 4.78125, "rewards/rejected": -10.25, "step": 15510 }, { "epoch": 1.9835772253818136, "grad_norm": 12.471330269650029, "learning_rate": 1.019083644185903e-10, "logits/chosen": -3.0625, "logits/rejected": -2.796875, "logps/chosen": -688.0, "logps/rejected": -1112.0, "loss": 0.1533, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.375, "rewards/margins": 4.28125, "rewards/rejected": -9.625, "step": 15520 }, { "epoch": 1.984855262317081, "grad_norm": 10.493625321839762, "learning_rate": 8.660806530991216e-11, "logits/chosen": -3.0, "logits/rejected": -2.59375, "logps/chosen": -684.0, "logps/rejected": -1120.0, "loss": 0.1667, "rewards/accuracies": 0.9375, "rewards/chosen": -5.3125, "rewards/margins": 4.375, "rewards/rejected": -9.6875, "step": 15530 }, { "epoch": 1.9861332992523484, "grad_norm": 17.377116192588073, "learning_rate": 7.255141780918018e-11, "logits/chosen": -3.0, "logits/rejected": -2.84375, "logps/chosen": -768.0, "logps/rejected": -1152.0, "loss": 0.1665, "rewards/accuracies": 0.893750011920929, "rewards/chosen": -5.8125, "rewards/margins": 4.09375, "rewards/rejected": -9.9375, "step": 15540 }, { "epoch": 1.9874113361876158, "grad_norm": 13.187550536514495, "learning_rate": 5.973849186685643e-11, "logits/chosen": -2.96875, "logits/rejected": -2.8125, "logps/chosen": -780.0, "logps/rejected": -1152.0, "loss": 0.1823, "rewards/accuracies": 0.90625, "rewards/chosen": -6.03125, "rewards/margins": 4.09375, "rewards/rejected": -10.125, "step": 15550 }, { "epoch": 1.9886893731228832, "grad_norm": 11.526002181186742, "learning_rate": 4.816935124435373e-11, "logits/chosen": -3.046875, "logits/rejected": -2.6875, "logps/chosen": -696.0, "logps/rejected": -1216.0, "loss": 0.1346, "rewards/accuracies": 0.9437500238418579, "rewards/chosen": -5.34375, "rewards/margins": 5.0, "rewards/rejected": -10.375, "step": 15560 }, { "epoch": 1.9899674100581506, "grad_norm": 11.545572776265107, "learning_rate": 3.7844053513536035e-11, "logits/chosen": -3.03125, "logits/rejected": -2.875, "logps/chosen": -716.0, "logps/rejected": -1120.0, "loss": 0.1742, "rewards/accuracies": 0.925000011920929, "rewards/chosen": -5.53125, "rewards/margins": 4.375, "rewards/rejected": -9.9375, "step": 15570 }, { "epoch": 1.991245446993418, "grad_norm": 13.213488743746863, "learning_rate": 2.8762650056468607e-11, "logits/chosen": -3.0, "logits/rejected": -2.828125, "logps/chosen": -704.0, "logps/rejected": -1144.0, "loss": 0.1423, "rewards/accuracies": 0.9750000238418579, "rewards/chosen": -5.40625, "rewards/margins": 4.5625, "rewards/rejected": -10.0, "step": 15580 }, { "epoch": 1.9925234839286854, "grad_norm": 9.895766976087293, "learning_rate": 2.0925186065223756e-11, "logits/chosen": -3.078125, "logits/rejected": -2.765625, "logps/chosen": -672.0, "logps/rejected": -1136.0, "loss": 0.1705, "rewards/accuracies": 0.9375, "rewards/chosen": -5.34375, "rewards/margins": 4.6875, "rewards/rejected": -10.0625, "step": 15590 }, { "epoch": 1.993801520863953, "grad_norm": 10.674668242208217, "learning_rate": 1.433170054163102e-11, "logits/chosen": -3.015625, "logits/rejected": -2.65625, "logps/chosen": -696.0, "logps/rejected": -1120.0, "loss": 0.1578, "rewards/accuracies": 0.8999999761581421, "rewards/chosen": -5.46875, "rewards/margins": 4.28125, "rewards/rejected": -9.75, "step": 15600 }, { "epoch": 1.9950795577992204, "grad_norm": 9.530695253521662, "learning_rate": 8.982226296999629e-12, "logits/chosen": -3.125, "logits/rejected": -2.828125, "logps/chosen": -720.0, "logps/rejected": -1136.0, "loss": 0.1513, "rewards/accuracies": 0.8812500238418579, "rewards/chosen": -5.59375, "rewards/margins": 4.3125, "rewards/rejected": -9.875, "step": 15610 }, { "epoch": 1.9963575947344878, "grad_norm": 20.579796060334516, "learning_rate": 4.876789952146243e-12, "logits/chosen": -3.09375, "logits/rejected": -2.78125, "logps/chosen": -720.0, "logps/rejected": -1152.0, "loss": 0.1676, "rewards/accuracies": 0.9312499761581421, "rewards/chosen": -5.53125, "rewards/margins": 4.46875, "rewards/rejected": -10.0, "step": 15620 }, { "epoch": 1.9976356316697552, "grad_norm": 9.627868429654509, "learning_rate": 2.015411937061895e-12, "logits/chosen": -3.03125, "logits/rejected": -2.890625, "logps/chosen": -672.0, "logps/rejected": -1128.0, "loss": 0.1395, "rewards/accuracies": 0.9624999761581421, "rewards/chosen": -5.21875, "rewards/margins": 4.71875, "rewards/rejected": -9.9375, "step": 15630 }, { "epoch": 1.9989136686050228, "grad_norm": 10.344279691753776, "learning_rate": 3.981064908842313e-13, "logits/chosen": -3.046875, "logits/rejected": -2.75, "logps/chosen": -716.0, "logps/rejected": -1160.0, "loss": 0.1478, "rewards/accuracies": 0.918749988079071, "rewards/chosen": -5.53125, "rewards/margins": 4.5625, "rewards/rejected": -10.0625, "step": 15640 }, { "epoch": 1.9999360981532366, "step": 15648, "total_flos": 0.0, "train_loss": 0.0, "train_runtime": 42.011, "train_samples_per_second": 23839.372, "train_steps_per_second": 186.237 } ], "logging_steps": 10, "max_steps": 7824, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 10000, "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": 8, "trial_name": null, "trial_params": null }