zephyr-7b-gpo-v1-i3 / trainer_state.json
lole25's picture
Model save
d26c8c4 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3750,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.3333333333333334e-08,
"logits/chosen": -1.4607182741165161,
"logits/rejected": -1.0577633380889893,
"logps/chosen": -377.1839599609375,
"logps/rejected": -1292.140625,
"loss": 0.3828,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 1.3333333333333336e-07,
"logits/chosen": -1.422508716583252,
"logits/rejected": -1.0050468444824219,
"logps/chosen": -603.676025390625,
"logps/rejected": -1303.1278076171875,
"loss": 0.3726,
"rewards/accuracies": 0.3888888955116272,
"rewards/chosen": -0.00046346496674232185,
"rewards/margins": -0.0005589782958850265,
"rewards/rejected": 9.551318362355232e-05,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-07,
"logits/chosen": -1.469663381576538,
"logits/rejected": -1.0887094736099243,
"logps/chosen": -544.6007080078125,
"logps/rejected": -1043.440673828125,
"loss": 0.4111,
"rewards/accuracies": 0.550000011920929,
"rewards/chosen": 0.0005788623820990324,
"rewards/margins": 0.001082524424418807,
"rewards/rejected": -0.0005036621587350965,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.0000000000000003e-07,
"logits/chosen": -1.6168861389160156,
"logits/rejected": -1.055768609046936,
"logps/chosen": -653.9005737304688,
"logps/rejected": -1314.4986572265625,
"loss": 0.3835,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -8.300555782625452e-05,
"rewards/margins": 0.0015174217987805605,
"rewards/rejected": -0.001600427320227027,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333335e-07,
"logits/chosen": -1.3431918621063232,
"logits/rejected": -0.6968151926994324,
"logps/chosen": -665.5634765625,
"logps/rejected": -1349.5455322265625,
"loss": 0.3624,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.0007652758504264057,
"rewards/margins": 0.004389578010886908,
"rewards/rejected": -0.005154854152351618,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-07,
"logits/chosen": -1.36802077293396,
"logits/rejected": -0.7910489439964294,
"logps/chosen": -665.3347778320312,
"logps/rejected": -1172.180419921875,
"loss": 0.3291,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.0013909535482525826,
"rewards/margins": 0.00601952476426959,
"rewards/rejected": -0.0074104792438447475,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-07,
"logits/chosen": -1.6208984851837158,
"logits/rejected": -0.9291224479675293,
"logps/chosen": -660.3453369140625,
"logps/rejected": -1401.126220703125,
"loss": 0.3343,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.00016163568943738937,
"rewards/margins": 0.020959021523594856,
"rewards/rejected": -0.02079738676548004,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 9.333333333333334e-07,
"logits/chosen": -1.5982778072357178,
"logits/rejected": -0.6397973299026489,
"logps/chosen": -600.2814331054688,
"logps/rejected": -1338.4075927734375,
"loss": 0.3381,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.0026822402141988277,
"rewards/margins": 0.034768685698509216,
"rewards/rejected": -0.032086439430713654,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.066666666666667e-06,
"logits/chosen": -1.6199333667755127,
"logits/rejected": -0.8920964002609253,
"logps/chosen": -589.6707763671875,
"logps/rejected": -1347.08837890625,
"loss": 0.3062,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.005661836825311184,
"rewards/margins": 0.04629923403263092,
"rewards/rejected": -0.05196107178926468,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"logits/chosen": -1.1941367387771606,
"logits/rejected": -0.7822949290275574,
"logps/chosen": -618.1773681640625,
"logps/rejected": -1214.458251953125,
"loss": 0.3073,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.02980031445622444,
"rewards/margins": 0.08387573063373566,
"rewards/rejected": -0.1136760488152504,
"step": 90
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333334e-06,
"logits/chosen": -1.1643040180206299,
"logits/rejected": -0.3697466552257538,
"logps/chosen": -628.5645141601562,
"logps/rejected": -1501.861572265625,
"loss": 0.2341,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.11866208165884018,
"rewards/margins": 0.1647382378578186,
"rewards/rejected": -0.283400297164917,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.4666666666666669e-06,
"logits/chosen": -1.1232091188430786,
"logits/rejected": -0.17076462507247925,
"logps/chosen": -800.4198608398438,
"logps/rejected": -1697.21875,
"loss": 0.1819,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.23575441539287567,
"rewards/margins": 0.2496374547481537,
"rewards/rejected": -0.48539191484451294,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 1.6000000000000001e-06,
"logits/chosen": -1.1497989892959595,
"logits/rejected": -0.4379865527153015,
"logps/chosen": -914.0667114257812,
"logps/rejected": -1844.274169921875,
"loss": 0.2595,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.29556211829185486,
"rewards/margins": 0.2584637701511383,
"rewards/rejected": -0.5540258288383484,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 1.7333333333333336e-06,
"logits/chosen": -1.2714927196502686,
"logits/rejected": -0.7818469405174255,
"logps/chosen": -809.2772216796875,
"logps/rejected": -1702.701171875,
"loss": 0.252,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.20852358639240265,
"rewards/margins": 0.2036806344985962,
"rewards/rejected": -0.4122042655944824,
"step": 130
},
{
"epoch": 0.04,
"learning_rate": 1.8666666666666669e-06,
"logits/chosen": -1.4557206630706787,
"logits/rejected": -0.5710101127624512,
"logps/chosen": -727.69189453125,
"logps/rejected": -1672.314208984375,
"loss": 0.1792,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2208588570356369,
"rewards/margins": 0.21561995148658752,
"rewards/rejected": -0.436478853225708,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.0000000000000003e-06,
"logits/chosen": -1.2810709476470947,
"logits/rejected": -0.3112773001194,
"logps/chosen": -960.1036376953125,
"logps/rejected": -1976.63671875,
"loss": 0.2275,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3756103217601776,
"rewards/margins": 0.28143054246902466,
"rewards/rejected": -0.6570408344268799,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 2.133333333333334e-06,
"logits/chosen": -1.1071354150772095,
"logits/rejected": -0.2948054373264313,
"logps/chosen": -707.2420654296875,
"logps/rejected": -1617.572509765625,
"loss": 0.3027,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.2601652145385742,
"rewards/margins": 0.25246208906173706,
"rewards/rejected": -0.5126273036003113,
"step": 160
},
{
"epoch": 0.05,
"learning_rate": 2.266666666666667e-06,
"logits/chosen": -1.1747493743896484,
"logits/rejected": -0.516975998878479,
"logps/chosen": -671.3464965820312,
"logps/rejected": -1495.285400390625,
"loss": 0.2483,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.1535535305738449,
"rewards/margins": 0.13311822712421417,
"rewards/rejected": -0.2866717278957367,
"step": 170
},
{
"epoch": 0.05,
"learning_rate": 2.4000000000000003e-06,
"logits/chosen": -1.5080617666244507,
"logits/rejected": -0.1595776528120041,
"logps/chosen": -766.1541137695312,
"logps/rejected": -1505.566162109375,
"loss": 0.1896,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1698867678642273,
"rewards/margins": 0.1528700590133667,
"rewards/rejected": -0.322756826877594,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 2.5333333333333338e-06,
"logits/chosen": -1.4777127504348755,
"logits/rejected": -0.1578340083360672,
"logps/chosen": -883.9136962890625,
"logps/rejected": -1729.061279296875,
"loss": 0.2441,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.2274136245250702,
"rewards/margins": 0.1767813265323639,
"rewards/rejected": -0.4041949212551117,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 2.666666666666667e-06,
"logits/chosen": -1.5583598613739014,
"logits/rejected": -0.15428844094276428,
"logps/chosen": -954.4393310546875,
"logps/rejected": -1927.611328125,
"loss": 0.1691,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2696084976196289,
"rewards/margins": 0.28209298849105835,
"rewards/rejected": -0.5517014861106873,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 2.8000000000000003e-06,
"logits/chosen": -1.148453950881958,
"logits/rejected": 0.03259178251028061,
"logps/chosen": -963.2941284179688,
"logps/rejected": -1878.1683349609375,
"loss": 0.2653,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.29248175024986267,
"rewards/margins": 0.2701197564601898,
"rewards/rejected": -0.5626015067100525,
"step": 210
},
{
"epoch": 0.06,
"learning_rate": 2.9333333333333338e-06,
"logits/chosen": -1.4708116054534912,
"logits/rejected": -0.4165799617767334,
"logps/chosen": -663.6636352539062,
"logps/rejected": -1678.5081787109375,
"loss": 0.1756,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1571148931980133,
"rewards/margins": 0.24838630855083466,
"rewards/rejected": -0.40550118684768677,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 3.066666666666667e-06,
"logits/chosen": -1.17020583152771,
"logits/rejected": -0.47575148940086365,
"logps/chosen": -920.1783447265625,
"logps/rejected": -1948.2427978515625,
"loss": 0.1913,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2500799894332886,
"rewards/margins": 0.24250411987304688,
"rewards/rejected": -0.49258413910865784,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 3.2000000000000003e-06,
"logits/chosen": -1.615384817123413,
"logits/rejected": -0.5452179312705994,
"logps/chosen": -855.1349487304688,
"logps/rejected": -1652.406494140625,
"loss": 0.1915,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.20125408470630646,
"rewards/margins": 0.23299658298492432,
"rewards/rejected": -0.4342506527900696,
"step": 240
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": -1.231403112411499,
"logits/rejected": 0.04080945998430252,
"logps/chosen": -830.4852294921875,
"logps/rejected": -1568.180908203125,
"loss": 0.2531,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.16112910211086273,
"rewards/margins": 0.1832246333360672,
"rewards/rejected": -0.34435373544692993,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 3.4666666666666672e-06,
"logits/chosen": -1.4119511842727661,
"logits/rejected": 0.3003128170967102,
"logps/chosen": -697.4551391601562,
"logps/rejected": -1790.427978515625,
"loss": 0.1495,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12422380596399307,
"rewards/margins": 0.2803717255592346,
"rewards/rejected": -0.4045955538749695,
"step": 260
},
{
"epoch": 0.07,
"learning_rate": 3.6000000000000003e-06,
"logits/chosen": -1.4090235233306885,
"logits/rejected": 0.20056810975074768,
"logps/chosen": -781.8931884765625,
"logps/rejected": -1546.812744140625,
"loss": 0.2132,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.18441830575466156,
"rewards/margins": 0.21885094046592712,
"rewards/rejected": -0.40326929092407227,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 3.7333333333333337e-06,
"logits/chosen": -1.3472883701324463,
"logits/rejected": -0.16090384125709534,
"logps/chosen": -674.9337158203125,
"logps/rejected": -1524.169677734375,
"loss": 0.2165,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.11200114339590073,
"rewards/margins": 0.18906202912330627,
"rewards/rejected": -0.3010631799697876,
"step": 280
},
{
"epoch": 0.08,
"learning_rate": 3.866666666666667e-06,
"logits/chosen": -1.5704516172409058,
"logits/rejected": -0.3292531967163086,
"logps/chosen": -733.7753295898438,
"logps/rejected": -1638.066162109375,
"loss": 0.2036,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0726594403386116,
"rewards/margins": 0.25331053137779236,
"rewards/rejected": -0.32596996426582336,
"step": 290
},
{
"epoch": 0.08,
"learning_rate": 4.000000000000001e-06,
"logits/chosen": -1.4241251945495605,
"logits/rejected": -0.6696543097496033,
"logps/chosen": -756.4315795898438,
"logps/rejected": -1572.5772705078125,
"loss": 0.1903,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.16283096373081207,
"rewards/margins": 0.22412030398845673,
"rewards/rejected": -0.3869512677192688,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 4.133333333333333e-06,
"logits/chosen": -1.368127465248108,
"logits/rejected": -0.3331999182701111,
"logps/chosen": -967.3336791992188,
"logps/rejected": -1911.6470947265625,
"loss": 0.1603,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2711753249168396,
"rewards/margins": 0.3143623471260071,
"rewards/rejected": -0.5855377316474915,
"step": 310
},
{
"epoch": 0.09,
"learning_rate": 4.266666666666668e-06,
"logits/chosen": -1.71210515499115,
"logits/rejected": 0.0729786604642868,
"logps/chosen": -1014.9069213867188,
"logps/rejected": -1880.586669921875,
"loss": 0.1866,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.31470808386802673,
"rewards/margins": 0.3131619393825531,
"rewards/rejected": -0.6278700828552246,
"step": 320
},
{
"epoch": 0.09,
"learning_rate": 4.4e-06,
"logits/chosen": -1.2539806365966797,
"logits/rejected": -0.792563796043396,
"logps/chosen": -697.0578002929688,
"logps/rejected": -1576.9892578125,
"loss": 0.1946,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15875259041786194,
"rewards/margins": 0.22502513229846954,
"rewards/rejected": -0.3837777078151703,
"step": 330
},
{
"epoch": 0.09,
"learning_rate": 4.533333333333334e-06,
"logits/chosen": -1.3388795852661133,
"logits/rejected": -0.407992422580719,
"logps/chosen": -761.8697509765625,
"logps/rejected": -1639.280517578125,
"loss": 1.8252,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.2346605360507965,
"rewards/margins": 0.4083401560783386,
"rewards/rejected": -0.6430006623268127,
"step": 340
},
{
"epoch": 0.09,
"learning_rate": 4.666666666666667e-06,
"logits/chosen": -1.370415449142456,
"logits/rejected": -0.36617863178253174,
"logps/chosen": -647.6007080078125,
"logps/rejected": -1585.802001953125,
"loss": 0.2431,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1556210219860077,
"rewards/margins": 0.14714348316192627,
"rewards/rejected": -0.30276453495025635,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 4.800000000000001e-06,
"logits/chosen": -1.4223321676254272,
"logits/rejected": -0.7678893804550171,
"logps/chosen": -862.3582763671875,
"logps/rejected": -1554.5501708984375,
"loss": 0.3499,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.15241160988807678,
"rewards/margins": 0.07054910808801651,
"rewards/rejected": -0.22296074032783508,
"step": 360
},
{
"epoch": 0.1,
"learning_rate": 4.933333333333334e-06,
"logits/chosen": -1.6838983297348022,
"logits/rejected": -0.8386624455451965,
"logps/chosen": -854.6920776367188,
"logps/rejected": -1489.2125244140625,
"loss": 0.3149,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": -0.12944355607032776,
"rewards/margins": 0.06882871687412262,
"rewards/rejected": -0.19827227294445038,
"step": 370
},
{
"epoch": 0.1,
"learning_rate": 4.999972922944898e-06,
"logits/chosen": -1.6282720565795898,
"logits/rejected": -1.0739606618881226,
"logps/chosen": -714.4268188476562,
"logps/rejected": -1637.6285400390625,
"loss": 0.1918,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07471625506877899,
"rewards/margins": 0.17545118927955627,
"rewards/rejected": -0.2501674294471741,
"step": 380
},
{
"epoch": 0.1,
"learning_rate": 4.999756310023261e-06,
"logits/chosen": -1.5494694709777832,
"logits/rejected": -0.6611472964286804,
"logps/chosen": -666.4691162109375,
"logps/rejected": -1401.627685546875,
"loss": 0.2341,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.024204634130001068,
"rewards/margins": 0.2136712521314621,
"rewards/rejected": -0.23787586390972137,
"step": 390
},
{
"epoch": 0.11,
"learning_rate": 4.999323102948655e-06,
"logits/chosen": -1.4422892332077026,
"logits/rejected": -0.2696318030357361,
"logps/chosen": -627.1585083007812,
"logps/rejected": -1432.5504150390625,
"loss": 0.2485,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.0546126663684845,
"rewards/margins": 0.1983325034379959,
"rewards/rejected": -0.2529451549053192,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 4.998673339256785e-06,
"logits/chosen": -1.4705862998962402,
"logits/rejected": -0.8628055453300476,
"logps/chosen": -736.0491333007812,
"logps/rejected": -1735.3140869140625,
"loss": 0.1994,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1055409163236618,
"rewards/margins": 0.21688199043273926,
"rewards/rejected": -0.32242292165756226,
"step": 410
},
{
"epoch": 0.11,
"learning_rate": 4.997807075247147e-06,
"logits/chosen": -1.3044060468673706,
"logits/rejected": -0.5804620981216431,
"logps/chosen": -709.5520629882812,
"logps/rejected": -1824.0869140625,
"loss": 0.1758,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.13376520574092865,
"rewards/margins": 0.34297293424606323,
"rewards/rejected": -0.4767381548881531,
"step": 420
},
{
"epoch": 0.11,
"learning_rate": 4.996724385978142e-06,
"logits/chosen": -1.5715945959091187,
"logits/rejected": -0.49768322706222534,
"logps/chosen": -752.0654907226562,
"logps/rejected": -1800.0816650390625,
"loss": 0.1937,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12571558356285095,
"rewards/margins": 0.3276565372943878,
"rewards/rejected": -0.45337215065956116,
"step": 430
},
{
"epoch": 0.12,
"learning_rate": 4.995425365260585e-06,
"logits/chosen": -1.4646804332733154,
"logits/rejected": -0.6573309898376465,
"logps/chosen": -604.1141357421875,
"logps/rejected": -1509.289306640625,
"loss": 0.1685,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.08397423475980759,
"rewards/margins": 0.22961445152759552,
"rewards/rejected": -0.3135886788368225,
"step": 440
},
{
"epoch": 0.12,
"learning_rate": 4.993910125649561e-06,
"logits/chosen": -1.58323073387146,
"logits/rejected": -0.9330165982246399,
"logps/chosen": -575.3523559570312,
"logps/rejected": -1430.528564453125,
"loss": 0.1947,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.0727187842130661,
"rewards/margins": 0.2145281583070755,
"rewards/rejected": -0.2872469425201416,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 4.992178798434684e-06,
"logits/chosen": -1.5033951997756958,
"logits/rejected": -0.5900696516036987,
"logps/chosen": -897.4099731445312,
"logps/rejected": -1658.7633056640625,
"loss": 0.2771,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17770914733409882,
"rewards/margins": 0.19083838164806366,
"rewards/rejected": -0.3685474991798401,
"step": 460
},
{
"epoch": 0.13,
"learning_rate": 4.990231533628719e-06,
"logits/chosen": -1.6505463123321533,
"logits/rejected": -0.9378656148910522,
"logps/chosen": -690.3605346679688,
"logps/rejected": -1701.1597900390625,
"loss": 0.2146,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13155733048915863,
"rewards/margins": 0.2767486572265625,
"rewards/rejected": -0.40830597281455994,
"step": 470
},
{
"epoch": 0.13,
"learning_rate": 4.988068499954578e-06,
"logits/chosen": -1.555983543395996,
"logits/rejected": -0.4679872393608093,
"logps/chosen": -698.6531982421875,
"logps/rejected": -1681.360595703125,
"loss": 0.2585,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12316081672906876,
"rewards/margins": 0.31332293152809143,
"rewards/rejected": -0.436483770608902,
"step": 480
},
{
"epoch": 0.13,
"learning_rate": 4.985689884830711e-06,
"logits/chosen": -1.5021615028381348,
"logits/rejected": -0.6322463750839233,
"logps/chosen": -743.8492431640625,
"logps/rejected": -1656.78515625,
"loss": 0.1861,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09808465093374252,
"rewards/margins": 0.18296462297439575,
"rewards/rejected": -0.28104931116104126,
"step": 490
},
{
"epoch": 0.13,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": -1.537832260131836,
"logits/rejected": -0.5591806173324585,
"logps/chosen": -771.895751953125,
"logps/rejected": -1593.089111328125,
"loss": 0.2529,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11263638734817505,
"rewards/margins": 0.1450023353099823,
"rewards/rejected": -0.25763875246047974,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 4.980286753286196e-06,
"logits/chosen": -1.4696406126022339,
"logits/rejected": -0.6640992164611816,
"logps/chosen": -686.0765380859375,
"logps/rejected": -1460.5648193359375,
"loss": 0.1911,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12830252945423126,
"rewards/margins": 0.16823554039001465,
"rewards/rejected": -0.2965380549430847,
"step": 510
},
{
"epoch": 0.14,
"learning_rate": 4.97726270502586e-06,
"logits/chosen": -1.4664279222488403,
"logits/rejected": -0.7313605546951294,
"logps/chosen": -818.5562744140625,
"logps/rejected": -1775.9241943359375,
"loss": 0.1947,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.20023846626281738,
"rewards/margins": 0.24244621396064758,
"rewards/rejected": -0.44268471002578735,
"step": 520
},
{
"epoch": 0.14,
"learning_rate": 4.974024011595864e-06,
"logits/chosen": -1.5485690832138062,
"logits/rejected": -0.6978067755699158,
"logps/chosen": -942.9921875,
"logps/rejected": -1794.7066650390625,
"loss": 0.3032,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.21240389347076416,
"rewards/margins": 0.2622910141944885,
"rewards/rejected": -0.4746948778629303,
"step": 530
},
{
"epoch": 0.14,
"learning_rate": 4.970570953616383e-06,
"logits/chosen": -1.4141438007354736,
"logits/rejected": -0.4546588957309723,
"logps/chosen": -668.2946166992188,
"logps/rejected": -1650.0318603515625,
"loss": 0.2265,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1157810240983963,
"rewards/margins": 0.2432054728269577,
"rewards/rejected": -0.3589865267276764,
"step": 540
},
{
"epoch": 0.15,
"learning_rate": 4.966903830281449e-06,
"logits/chosen": -1.8105099201202393,
"logits/rejected": -0.36172086000442505,
"logps/chosen": -677.1749267578125,
"logps/rejected": -1432.0328369140625,
"loss": 0.2364,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.14466652274131775,
"rewards/margins": 0.17044545710086823,
"rewards/rejected": -0.3151119649410248,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 4.9630229593330226e-06,
"logits/chosen": -1.544588327407837,
"logits/rejected": -0.8351410031318665,
"logps/chosen": -834.5636596679688,
"logps/rejected": -1683.013671875,
"loss": 0.2298,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.18364231288433075,
"rewards/margins": 0.259293794631958,
"rewards/rejected": -0.44293609261512756,
"step": 560
},
{
"epoch": 0.15,
"learning_rate": 4.958928677033465e-06,
"logits/chosen": -1.6264123916625977,
"logits/rejected": -0.5471200942993164,
"logps/chosen": -858.4713134765625,
"logps/rejected": -1681.949951171875,
"loss": 0.2358,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.2060559242963791,
"rewards/margins": 0.24328169226646423,
"rewards/rejected": -0.4493376612663269,
"step": 570
},
{
"epoch": 0.15,
"learning_rate": 4.954621338136399e-06,
"logits/chosen": -1.4879562854766846,
"logits/rejected": -0.6239393949508667,
"logps/chosen": -916.1363525390625,
"logps/rejected": -1731.7279052734375,
"loss": 0.2434,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1614031344652176,
"rewards/margins": 0.23414401710033417,
"rewards/rejected": -0.39554715156555176,
"step": 580
},
{
"epoch": 0.16,
"learning_rate": 4.95010131585597e-06,
"logits/chosen": -1.220314383506775,
"logits/rejected": -0.44166284799575806,
"logps/chosen": -772.2953491210938,
"logps/rejected": -1575.3052978515625,
"loss": 0.2662,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08703169226646423,
"rewards/margins": 0.2424904853105545,
"rewards/rejected": -0.32952219247817993,
"step": 590
},
{
"epoch": 0.16,
"learning_rate": 4.9453690018345144e-06,
"logits/chosen": -1.4573420286178589,
"logits/rejected": -0.5127624273300171,
"logps/chosen": -504.53057861328125,
"logps/rejected": -1400.8695068359375,
"loss": 0.1911,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.05916530638933182,
"rewards/margins": 0.23245540261268616,
"rewards/rejected": -0.291620671749115,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 4.940424806108619e-06,
"logits/chosen": -1.6605002880096436,
"logits/rejected": -0.6169177293777466,
"logps/chosen": -778.3709716796875,
"logps/rejected": -1566.796630859375,
"loss": 0.1824,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08246553689241409,
"rewards/margins": 0.2251826822757721,
"rewards/rejected": -0.307648241519928,
"step": 610
},
{
"epoch": 0.17,
"learning_rate": 4.935269157073597e-06,
"logits/chosen": -1.6027615070343018,
"logits/rejected": -1.1422998905181885,
"logps/chosen": -592.0419921875,
"logps/rejected": -1658.290283203125,
"loss": 0.2057,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.048401519656181335,
"rewards/margins": 0.334246426820755,
"rewards/rejected": -0.38264790177345276,
"step": 620
},
{
"epoch": 0.17,
"learning_rate": 4.9299025014463665e-06,
"logits/chosen": -1.4095045328140259,
"logits/rejected": -0.1883077323436737,
"logps/chosen": -607.0613403320312,
"logps/rejected": -1580.00341796875,
"loss": 0.1792,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07415771484375,
"rewards/margins": 0.2879538834095001,
"rewards/rejected": -0.3621116280555725,
"step": 630
},
{
"epoch": 0.17,
"learning_rate": 4.924325304226745e-06,
"logits/chosen": -1.6297693252563477,
"logits/rejected": -0.49017101526260376,
"logps/chosen": -792.6488037109375,
"logps/rejected": -1665.556396484375,
"loss": 0.1603,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09000878036022186,
"rewards/margins": 0.2915512025356293,
"rewards/rejected": -0.38155999779701233,
"step": 640
},
{
"epoch": 0.17,
"learning_rate": 4.91853804865716e-06,
"logits/chosen": -1.1054975986480713,
"logits/rejected": -0.42201828956604004,
"logps/chosen": -713.8890991210938,
"logps/rejected": -1576.742919921875,
"loss": 0.1704,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11752257496118546,
"rewards/margins": 0.2575463354587555,
"rewards/rejected": -0.37506890296936035,
"step": 650
},
{
"epoch": 0.18,
"learning_rate": 4.912541236180779e-06,
"logits/chosen": -1.3571802377700806,
"logits/rejected": -0.6059755086898804,
"logps/chosen": -884.0383911132812,
"logps/rejected": -1731.5504150390625,
"loss": 0.2647,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.24475233256816864,
"rewards/margins": 0.20648574829101562,
"rewards/rejected": -0.45123806595802307,
"step": 660
},
{
"epoch": 0.18,
"learning_rate": 4.9063353863980565e-06,
"logits/chosen": -1.3680832386016846,
"logits/rejected": -0.6090233325958252,
"logps/chosen": -807.2792358398438,
"logps/rejected": -1725.8609619140625,
"loss": 0.1917,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.19080020487308502,
"rewards/margins": 0.28758734464645386,
"rewards/rejected": -0.4783875346183777,
"step": 670
},
{
"epoch": 0.18,
"learning_rate": 4.899921037021719e-06,
"logits/chosen": -1.7092090845108032,
"logits/rejected": -0.9691111445426941,
"logps/chosen": -696.3843994140625,
"logps/rejected": -1516.789794921875,
"loss": 0.1585,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16210845112800598,
"rewards/margins": 0.24107725918293,
"rewards/rejected": -0.4031856954097748,
"step": 680
},
{
"epoch": 0.18,
"learning_rate": 4.893298743830168e-06,
"logits/chosen": -1.375356674194336,
"logits/rejected": -0.36455339193344116,
"logps/chosen": -759.2677612304688,
"logps/rejected": -1726.4449462890625,
"loss": 0.1731,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.196892648935318,
"rewards/margins": 0.2657999098300934,
"rewards/rejected": -0.46269258856773376,
"step": 690
},
{
"epoch": 0.19,
"learning_rate": 4.88646908061933e-06,
"logits/chosen": -1.4417951107025146,
"logits/rejected": -0.3565208911895752,
"logps/chosen": -741.1653442382812,
"logps/rejected": -1683.4515380859375,
"loss": 0.2447,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.18826603889465332,
"rewards/margins": 0.2816776633262634,
"rewards/rejected": -0.46994370222091675,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 4.879432639152935e-06,
"logits/chosen": -1.37065851688385,
"logits/rejected": -0.2773549556732178,
"logps/chosen": -838.9853515625,
"logps/rejected": -1707.4296875,
"loss": 0.2116,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12412204593420029,
"rewards/margins": 0.23440977931022644,
"rewards/rejected": -0.3585318624973297,
"step": 710
},
{
"epoch": 0.19,
"learning_rate": 4.8721900291112415e-06,
"logits/chosen": -1.580479621887207,
"logits/rejected": -0.6059103608131409,
"logps/chosen": -790.3890380859375,
"logps/rejected": -1548.9473876953125,
"loss": 0.2076,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.11692629009485245,
"rewards/margins": 0.20617632567882538,
"rewards/rejected": -0.3231026232242584,
"step": 720
},
{
"epoch": 0.19,
"learning_rate": 4.864741878038218e-06,
"logits/chosen": -1.4246976375579834,
"logits/rejected": -0.6282674670219421,
"logps/chosen": -739.7086791992188,
"logps/rejected": -1689.615234375,
"loss": 0.1781,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11477355659008026,
"rewards/margins": 0.28999191522598267,
"rewards/rejected": -0.40476545691490173,
"step": 730
},
{
"epoch": 0.2,
"learning_rate": 4.857088831287158e-06,
"logits/chosen": -1.603228211402893,
"logits/rejected": -0.13681410253047943,
"logps/chosen": -734.5335693359375,
"logps/rejected": -1756.3695068359375,
"loss": 0.171,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16050152480602264,
"rewards/margins": 0.3059554696083069,
"rewards/rejected": -0.4664570391178131,
"step": 740
},
{
"epoch": 0.2,
"learning_rate": 4.849231551964771e-06,
"logits/chosen": -1.3526674509048462,
"logits/rejected": -0.1356133222579956,
"logps/chosen": -775.5394897460938,
"logps/rejected": -1612.2784423828125,
"loss": 0.2257,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.16132906079292297,
"rewards/margins": 0.2226000726222992,
"rewards/rejected": -0.38392913341522217,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 4.841170720873723e-06,
"logits/chosen": -1.466218113899231,
"logits/rejected": -0.2698180377483368,
"logps/chosen": -699.693359375,
"logps/rejected": -1555.100341796875,
"loss": 0.2182,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13787570595741272,
"rewards/margins": 0.24788126349449158,
"rewards/rejected": -0.3857569694519043,
"step": 760
},
{
"epoch": 0.21,
"learning_rate": 4.832907036453647e-06,
"logits/chosen": -1.4457144737243652,
"logits/rejected": -0.35442107915878296,
"logps/chosen": -820.4182739257812,
"logps/rejected": -1727.771484375,
"loss": 0.1765,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12979528307914734,
"rewards/margins": 0.2619260847568512,
"rewards/rejected": -0.39172136783599854,
"step": 770
},
{
"epoch": 0.21,
"learning_rate": 4.824441214720629e-06,
"logits/chosen": -1.2430192232131958,
"logits/rejected": -0.31560835242271423,
"logps/chosen": -599.2308959960938,
"logps/rejected": -1673.4691162109375,
"loss": 0.1464,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11397616565227509,
"rewards/margins": 0.341776579618454,
"rewards/rejected": -0.45575276017189026,
"step": 780
},
{
"epoch": 0.21,
"learning_rate": 4.815773989205165e-06,
"logits/chosen": -1.5526068210601807,
"logits/rejected": -0.6952439546585083,
"logps/chosen": -760.5252685546875,
"logps/rejected": -1961.7568359375,
"loss": 0.1258,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16641755402088165,
"rewards/margins": 0.44562679529190063,
"rewards/rejected": -0.6120442748069763,
"step": 790
},
{
"epoch": 0.21,
"learning_rate": 4.806906110888606e-06,
"logits/chosen": -1.4445289373397827,
"logits/rejected": -0.6352332234382629,
"logps/chosen": -702.0392456054688,
"logps/rejected": -1625.85302734375,
"loss": 0.1976,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.14837577939033508,
"rewards/margins": 0.2570492923259735,
"rewards/rejected": -0.4054250717163086,
"step": 800
},
{
"epoch": 0.22,
"learning_rate": 4.7978383481380865e-06,
"logits/chosen": -1.404865026473999,
"logits/rejected": -0.4630287289619446,
"logps/chosen": -645.7227172851562,
"logps/rejected": -1415.0540771484375,
"loss": 0.1912,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.13058145344257355,
"rewards/margins": 0.18823085725307465,
"rewards/rejected": -0.3188122808933258,
"step": 810
},
{
"epoch": 0.22,
"learning_rate": 4.788571486639948e-06,
"logits/chosen": -1.32173752784729,
"logits/rejected": -0.5882034301757812,
"logps/chosen": -841.7330932617188,
"logps/rejected": -1966.875732421875,
"loss": 0.1613,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14226606488227844,
"rewards/margins": 0.33964481949806213,
"rewards/rejected": -0.4819108545780182,
"step": 820
},
{
"epoch": 0.22,
"learning_rate": 4.779106329331665e-06,
"logits/chosen": -1.5085209608078003,
"logits/rejected": -0.3688226044178009,
"logps/chosen": -730.3353271484375,
"logps/rejected": -1609.4810791015625,
"loss": 0.1961,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14409688115119934,
"rewards/margins": 0.268598735332489,
"rewards/rejected": -0.41269564628601074,
"step": 830
},
{
"epoch": 0.22,
"learning_rate": 4.769443696332272e-06,
"logits/chosen": -1.30078125,
"logits/rejected": -0.17256946861743927,
"logps/chosen": -978.0304565429688,
"logps/rejected": -2078.25830078125,
"loss": 0.1996,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.25719624757766724,
"rewards/margins": 0.36804524064064026,
"rewards/rejected": -0.6252414584159851,
"step": 840
},
{
"epoch": 0.23,
"learning_rate": 4.759584424871302e-06,
"logits/chosen": -1.2128162384033203,
"logits/rejected": -0.2545672059059143,
"logps/chosen": -833.3894653320312,
"logps/rejected": -1909.220703125,
"loss": 0.1221,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.23837868869304657,
"rewards/margins": 0.341843843460083,
"rewards/rejected": -0.580222487449646,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 4.749529369216246e-06,
"logits/chosen": -1.505275845527649,
"logits/rejected": -0.8203606605529785,
"logps/chosen": -819.6467895507812,
"logps/rejected": -1751.1910400390625,
"loss": 0.2127,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19994454085826874,
"rewards/margins": 0.27609339356422424,
"rewards/rejected": -0.47603797912597656,
"step": 860
},
{
"epoch": 0.23,
"learning_rate": 4.7392794005985324e-06,
"logits/chosen": -1.4275624752044678,
"logits/rejected": -0.5869510769844055,
"logps/chosen": -676.9829711914062,
"logps/rejected": -1566.9033203125,
"loss": 0.1924,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.13645704090595245,
"rewards/margins": 0.26502570509910583,
"rewards/rejected": -0.4014827311038971,
"step": 870
},
{
"epoch": 0.23,
"learning_rate": 4.7288354071380415e-06,
"logits/chosen": -1.2436352968215942,
"logits/rejected": 0.3669028580188751,
"logps/chosen": -650.0794067382812,
"logps/rejected": -1452.456298828125,
"loss": 0.1965,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1296919882297516,
"rewards/margins": 0.22590501606464386,
"rewards/rejected": -0.35559698939323425,
"step": 880
},
{
"epoch": 0.24,
"learning_rate": 4.7181982937661485e-06,
"logits/chosen": -1.2913326025009155,
"logits/rejected": -0.41281813383102417,
"logps/chosen": -856.18896484375,
"logps/rejected": -1820.381591796875,
"loss": 0.1497,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.22433075308799744,
"rewards/margins": 0.2952847480773926,
"rewards/rejected": -0.5196155309677124,
"step": 890
},
{
"epoch": 0.24,
"learning_rate": 4.707368982147318e-06,
"logits/chosen": -1.3868077993392944,
"logits/rejected": -0.7847863435745239,
"logps/chosen": -884.7540283203125,
"logps/rejected": -1902.289306640625,
"loss": 0.1453,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.24957780539989471,
"rewards/margins": 0.3298245370388031,
"rewards/rejected": -0.5794023275375366,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 4.696348410599244e-06,
"logits/chosen": -1.3086035251617432,
"logits/rejected": -0.8314957618713379,
"logps/chosen": -689.6898803710938,
"logps/rejected": -1640.506103515625,
"loss": 0.1847,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.20904815196990967,
"rewards/margins": 0.30142712593078613,
"rewards/rejected": -0.5104752779006958,
"step": 910
},
{
"epoch": 0.25,
"learning_rate": 4.685137534011549e-06,
"logits/chosen": -1.565288782119751,
"logits/rejected": -0.27172648906707764,
"logps/chosen": -844.7750244140625,
"logps/rejected": -1699.747314453125,
"loss": 0.2055,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.2513663172721863,
"rewards/margins": 0.26970070600509644,
"rewards/rejected": -0.5210670232772827,
"step": 920
},
{
"epoch": 0.25,
"learning_rate": 4.673737323763048e-06,
"logits/chosen": -1.2228829860687256,
"logits/rejected": -0.5248723030090332,
"logps/chosen": -664.1156616210938,
"logps/rejected": -1562.0504150390625,
"loss": 0.1669,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.162824347615242,
"rewards/margins": 0.25564128160476685,
"rewards/rejected": -0.41846561431884766,
"step": 930
},
{
"epoch": 0.25,
"learning_rate": 4.662148767637578e-06,
"logits/chosen": -1.2507654428482056,
"logits/rejected": -0.5844524502754211,
"logps/chosen": -839.5402221679688,
"logps/rejected": -1746.392333984375,
"loss": 0.2229,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.18832561373710632,
"rewards/margins": 0.2621099650859833,
"rewards/rejected": -0.4504355788230896,
"step": 940
},
{
"epoch": 0.25,
"learning_rate": 4.650372869738415e-06,
"logits/chosen": -1.55803644657135,
"logits/rejected": -0.7119165658950806,
"logps/chosen": -736.268310546875,
"logps/rejected": -1563.0191650390625,
"loss": 0.2164,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10577349364757538,
"rewards/margins": 0.2629134953022003,
"rewards/rejected": -0.3686870038509369,
"step": 950
},
{
"epoch": 0.26,
"learning_rate": 4.638410650401267e-06,
"logits/chosen": -1.3386160135269165,
"logits/rejected": -0.11453273147344589,
"logps/chosen": -630.7725830078125,
"logps/rejected": -1448.709228515625,
"loss": 0.2449,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09674180299043655,
"rewards/margins": 0.18708564341068268,
"rewards/rejected": -0.28382742404937744,
"step": 960
},
{
"epoch": 0.26,
"learning_rate": 4.626263146105875e-06,
"logits/chosen": -1.6188023090362549,
"logits/rejected": -0.530141294002533,
"logps/chosen": -792.281005859375,
"logps/rejected": -1616.726806640625,
"loss": 0.2387,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.16821546852588654,
"rewards/margins": 0.20957235991954803,
"rewards/rejected": -0.37778785824775696,
"step": 970
},
{
"epoch": 0.26,
"learning_rate": 4.613931409386196e-06,
"logits/chosen": -1.4189367294311523,
"logits/rejected": -0.2773023247718811,
"logps/chosen": -763.1712646484375,
"logps/rejected": -1585.35791015625,
"loss": 0.2073,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15055999159812927,
"rewards/margins": 0.23061557114124298,
"rewards/rejected": -0.38117554783821106,
"step": 980
},
{
"epoch": 0.26,
"learning_rate": 4.601416508739211e-06,
"logits/chosen": -1.4606117010116577,
"logits/rejected": -0.6397336721420288,
"logps/chosen": -748.8417358398438,
"logps/rejected": -1670.744140625,
"loss": 0.1761,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14542202651500702,
"rewards/margins": 0.2664222717285156,
"rewards/rejected": -0.4118443429470062,
"step": 990
},
{
"epoch": 0.27,
"learning_rate": 4.588719528532342e-06,
"logits/chosen": -1.324894666671753,
"logits/rejected": -0.27193373441696167,
"logps/chosen": -764.4343872070312,
"logps/rejected": -1688.8870849609375,
"loss": 0.2086,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.13419455289840698,
"rewards/margins": 0.2759809195995331,
"rewards/rejected": -0.41017547249794006,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 4.575841568909494e-06,
"logits/chosen": -1.6580641269683838,
"logits/rejected": -0.3782033622264862,
"logps/chosen": -732.3780517578125,
"logps/rejected": -1696.1322021484375,
"loss": 0.1872,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10972636938095093,
"rewards/margins": 0.29507070779800415,
"rewards/rejected": -0.4047970771789551,
"step": 1010
},
{
"epoch": 0.27,
"learning_rate": 4.562783745695738e-06,
"logits/chosen": -1.503248691558838,
"logits/rejected": 0.1558968424797058,
"logps/chosen": -716.7626953125,
"logps/rejected": -1444.2918701171875,
"loss": 0.2014,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09443630278110504,
"rewards/margins": 0.21701212227344513,
"rewards/rejected": -0.31144842505455017,
"step": 1020
},
{
"epoch": 0.27,
"learning_rate": 4.549547190300622e-06,
"logits/chosen": -1.3067868947982788,
"logits/rejected": -0.4560883641242981,
"logps/chosen": -790.8909912109375,
"logps/rejected": -1757.387451171875,
"loss": 0.224,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1645580530166626,
"rewards/margins": 0.325479656457901,
"rewards/rejected": -0.490037739276886,
"step": 1030
},
{
"epoch": 0.28,
"learning_rate": 4.536133049620143e-06,
"logits/chosen": -1.140446662902832,
"logits/rejected": -0.34088680148124695,
"logps/chosen": -722.1959228515625,
"logps/rejected": -1831.537353515625,
"loss": 0.1268,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12402956187725067,
"rewards/margins": 0.37322643399238586,
"rewards/rejected": -0.49725598096847534,
"step": 1040
},
{
"epoch": 0.28,
"learning_rate": 4.522542485937369e-06,
"logits/chosen": -1.475990891456604,
"logits/rejected": -0.5005327463150024,
"logps/chosen": -783.8773193359375,
"logps/rejected": -1766.515869140625,
"loss": 0.176,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15318922698497772,
"rewards/margins": 0.31895914673805237,
"rewards/rejected": -0.4721483290195465,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 4.508776676821739e-06,
"logits/chosen": -1.4537649154663086,
"logits/rejected": -0.6859838366508484,
"logps/chosen": -839.9093627929688,
"logps/rejected": -1558.96435546875,
"loss": 0.1925,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.18507587909698486,
"rewards/margins": 0.21864044666290283,
"rewards/rejected": -0.4037163257598877,
"step": 1060
},
{
"epoch": 0.29,
"learning_rate": 4.494836815027022e-06,
"logits/chosen": -1.3298923969268799,
"logits/rejected": 0.0534161701798439,
"logps/chosen": -754.2385864257812,
"logps/rejected": -1705.55078125,
"loss": 0.232,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1495281457901001,
"rewards/margins": 0.2592945694923401,
"rewards/rejected": -0.4088227152824402,
"step": 1070
},
{
"epoch": 0.29,
"learning_rate": 4.4807241083879774e-06,
"logits/chosen": -1.3468838930130005,
"logits/rejected": -0.2606663107872009,
"logps/chosen": -726.0997314453125,
"logps/rejected": -1622.6500244140625,
"loss": 0.203,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13949835300445557,
"rewards/margins": 0.3188321590423584,
"rewards/rejected": -0.45833054184913635,
"step": 1080
},
{
"epoch": 0.29,
"learning_rate": 4.466439779715696e-06,
"logits/chosen": -1.3387072086334229,
"logits/rejected": -0.5179058313369751,
"logps/chosen": -702.0155029296875,
"logps/rejected": -1574.4173583984375,
"loss": 0.1881,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08557260781526566,
"rewards/margins": 0.29005369544029236,
"rewards/rejected": -0.3756263256072998,
"step": 1090
},
{
"epoch": 0.29,
"learning_rate": 4.451985066691649e-06,
"logits/chosen": -1.3348770141601562,
"logits/rejected": -0.6792385578155518,
"logps/chosen": -666.2745361328125,
"logps/rejected": -1547.283203125,
"loss": 0.1755,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09519227594137192,
"rewards/margins": 0.2866414785385132,
"rewards/rejected": -0.3818337321281433,
"step": 1100
},
{
"epoch": 0.3,
"learning_rate": 4.437361221760449e-06,
"logits/chosen": -1.5723166465759277,
"logits/rejected": -0.519290566444397,
"logps/chosen": -685.1051025390625,
"logps/rejected": -1876.2593994140625,
"loss": 0.1298,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10364224761724472,
"rewards/margins": 0.37660035490989685,
"rewards/rejected": -0.48024263978004456,
"step": 1110
},
{
"epoch": 0.3,
"learning_rate": 4.422569512021332e-06,
"logits/chosen": -1.4255059957504272,
"logits/rejected": 0.1319495439529419,
"logps/chosen": -696.9476318359375,
"logps/rejected": -1619.9892578125,
"loss": 0.2105,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0946447104215622,
"rewards/margins": 0.2710058093070984,
"rewards/rejected": -0.3656505048274994,
"step": 1120
},
{
"epoch": 0.3,
"learning_rate": 4.407611219118363e-06,
"logits/chosen": -1.3705596923828125,
"logits/rejected": -0.5301432609558105,
"logps/chosen": -698.9526977539062,
"logps/rejected": -1700.2493896484375,
"loss": 0.164,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1371428668498993,
"rewards/margins": 0.29292064905166626,
"rewards/rejected": -0.43006348609924316,
"step": 1130
},
{
"epoch": 0.3,
"learning_rate": 4.3924876391293915e-06,
"logits/chosen": -1.3947621583938599,
"logits/rejected": -0.8983534574508667,
"logps/chosen": -651.66162109375,
"logps/rejected": -1657.3372802734375,
"loss": 0.1776,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11623096466064453,
"rewards/margins": 0.2917022407054901,
"rewards/rejected": -0.40793323516845703,
"step": 1140
},
{
"epoch": 0.31,
"learning_rate": 4.377200082453748e-06,
"logits/chosen": -1.3964405059814453,
"logits/rejected": -0.6084726452827454,
"logps/chosen": -692.3182983398438,
"logps/rejected": -1706.6009521484375,
"loss": 0.4044,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12358059734106064,
"rewards/margins": 0.3641052842140198,
"rewards/rejected": -0.4876858592033386,
"step": 1150
},
{
"epoch": 0.31,
"learning_rate": 4.361749873698707e-06,
"logits/chosen": -1.4753860235214233,
"logits/rejected": -0.7048450708389282,
"logps/chosen": -698.3816528320312,
"logps/rejected": -1705.034423828125,
"loss": 0.124,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08291604369878769,
"rewards/margins": 0.2811528742313385,
"rewards/rejected": -0.3640689253807068,
"step": 1160
},
{
"epoch": 0.31,
"learning_rate": 4.346138351564711e-06,
"logits/chosen": -1.4870970249176025,
"logits/rejected": -0.9103308916091919,
"logps/chosen": -671.7784423828125,
"logps/rejected": -1529.8997802734375,
"loss": 0.2008,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06387507170438766,
"rewards/margins": 0.2635475695133209,
"rewards/rejected": -0.3274226784706116,
"step": 1170
},
{
"epoch": 0.31,
"learning_rate": 4.330366868729376e-06,
"logits/chosen": -1.421744465827942,
"logits/rejected": -0.6302933692932129,
"logps/chosen": -724.4271240234375,
"logps/rejected": -1651.8883056640625,
"loss": 0.1881,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1005069762468338,
"rewards/margins": 0.2800332307815552,
"rewards/rejected": -0.3805401921272278,
"step": 1180
},
{
"epoch": 0.32,
"learning_rate": 4.3144367917302964e-06,
"logits/chosen": -1.408508062362671,
"logits/rejected": -0.9906333088874817,
"logps/chosen": -622.6488647460938,
"logps/rejected": -1618.7296142578125,
"loss": 0.1708,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08803695440292358,
"rewards/margins": 0.27410316467285156,
"rewards/rejected": -0.36214011907577515,
"step": 1190
},
{
"epoch": 0.32,
"learning_rate": 4.2983495008466285e-06,
"logits/chosen": -1.4352304935455322,
"logits/rejected": -0.9279989004135132,
"logps/chosen": -665.1091918945312,
"logps/rejected": -1825.142333984375,
"loss": 0.2047,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10474482923746109,
"rewards/margins": 0.35141628980636597,
"rewards/rejected": -0.45616111159324646,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 4.2821063899795015e-06,
"logits/chosen": -1.563280463218689,
"logits/rejected": -0.555804431438446,
"logps/chosen": -646.5196533203125,
"logps/rejected": -1623.5345458984375,
"loss": 0.1522,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.055479682981967926,
"rewards/margins": 0.325848251581192,
"rewards/rejected": -0.38132789731025696,
"step": 1210
},
{
"epoch": 0.33,
"learning_rate": 4.265708866531238e-06,
"logits/chosen": -1.6181457042694092,
"logits/rejected": -0.8855217695236206,
"logps/chosen": -649.41455078125,
"logps/rejected": -1402.5155029296875,
"loss": 0.1827,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.051223646849393845,
"rewards/margins": 0.2268896847963333,
"rewards/rejected": -0.27811330556869507,
"step": 1220
},
{
"epoch": 0.33,
"learning_rate": 4.249158351283414e-06,
"logits/chosen": -1.4349555969238281,
"logits/rejected": -0.6732171773910522,
"logps/chosen": -621.8024291992188,
"logps/rejected": -1621.55908203125,
"loss": 0.1949,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06152881309390068,
"rewards/margins": 0.23424020409584045,
"rewards/rejected": -0.2957690358161926,
"step": 1230
},
{
"epoch": 0.33,
"learning_rate": 4.232456278273743e-06,
"logits/chosen": -1.477827548980713,
"logits/rejected": -1.140575647354126,
"logps/chosen": -703.6815185546875,
"logps/rejected": -1337.854248046875,
"loss": 0.2197,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": -0.10893256962299347,
"rewards/margins": 0.16516511142253876,
"rewards/rejected": -0.2740976810455322,
"step": 1240
},
{
"epoch": 0.33,
"learning_rate": 4.215604094671835e-06,
"logits/chosen": -1.6464779376983643,
"logits/rejected": -0.990412712097168,
"logps/chosen": -644.1171264648438,
"logps/rejected": -1499.906005859375,
"loss": 0.2661,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.10046522319316864,
"rewards/margins": 0.22357575595378876,
"rewards/rejected": -0.3240409791469574,
"step": 1250
},
{
"epoch": 0.34,
"learning_rate": 4.198603260653792e-06,
"logits/chosen": -1.5124647617340088,
"logits/rejected": -0.45678481459617615,
"logps/chosen": -716.1838989257812,
"logps/rejected": -1518.812255859375,
"loss": 0.165,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05440264195203781,
"rewards/margins": 0.2616801857948303,
"rewards/rejected": -0.3160828649997711,
"step": 1260
},
{
"epoch": 0.34,
"learning_rate": 4.181455249275701e-06,
"logits/chosen": -1.2498810291290283,
"logits/rejected": -0.8803867101669312,
"logps/chosen": -659.9884643554688,
"logps/rejected": -1640.7095947265625,
"loss": 0.2331,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.08121126145124435,
"rewards/margins": 0.27540498971939087,
"rewards/rejected": -0.35661619901657104,
"step": 1270
},
{
"epoch": 0.34,
"learning_rate": 4.1641615463459926e-06,
"logits/chosen": -1.5280659198760986,
"logits/rejected": -0.723167359828949,
"logps/chosen": -665.0728759765625,
"logps/rejected": -1953.181884765625,
"loss": 0.15,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.034947071224451065,
"rewards/margins": 0.38297995924949646,
"rewards/rejected": -0.417927086353302,
"step": 1280
},
{
"epoch": 0.34,
"learning_rate": 4.146723650296701e-06,
"logits/chosen": -1.5791549682617188,
"logits/rejected": -0.7244309186935425,
"logps/chosen": -609.816650390625,
"logps/rejected": -1450.82080078125,
"loss": 0.1426,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.05617643520236015,
"rewards/margins": 0.2224070131778717,
"rewards/rejected": -0.2785834074020386,
"step": 1290
},
{
"epoch": 0.35,
"learning_rate": 4.129143072053639e-06,
"logits/chosen": -1.7151685953140259,
"logits/rejected": -0.39171552658081055,
"logps/chosen": -769.5672607421875,
"logps/rejected": -1687.483154296875,
"loss": 0.1444,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.071159228682518,
"rewards/margins": 0.29008185863494873,
"rewards/rejected": -0.36124110221862793,
"step": 1300
},
{
"epoch": 0.35,
"learning_rate": 4.111421334905468e-06,
"logits/chosen": -1.49336838722229,
"logits/rejected": -0.5438629984855652,
"logps/chosen": -848.5822143554688,
"logps/rejected": -1763.63671875,
"loss": 0.2436,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10583722591400146,
"rewards/margins": 0.27904340624809265,
"rewards/rejected": -0.38488060235977173,
"step": 1310
},
{
"epoch": 0.35,
"learning_rate": 4.093559974371725e-06,
"logits/chosen": -1.359708547592163,
"logits/rejected": -0.6944249272346497,
"logps/chosen": -919.8367919921875,
"logps/rejected": -1900.1605224609375,
"loss": 0.1501,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1508992612361908,
"rewards/margins": 0.33068108558654785,
"rewards/rejected": -0.48158034682273865,
"step": 1320
},
{
"epoch": 0.35,
"learning_rate": 4.075560538069767e-06,
"logits/chosen": -1.4843828678131104,
"logits/rejected": -0.6843305826187134,
"logps/chosen": -726.9734497070312,
"logps/rejected": -1652.866943359375,
"loss": 0.1679,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08053059130907059,
"rewards/margins": 0.38363346457481384,
"rewards/rejected": -0.4641640782356262,
"step": 1330
},
{
"epoch": 0.36,
"learning_rate": 4.05742458558068e-06,
"logits/chosen": -1.5596168041229248,
"logits/rejected": -0.7530814409255981,
"logps/chosen": -646.64697265625,
"logps/rejected": -1448.167724609375,
"loss": 0.1954,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.07448308169841766,
"rewards/margins": 0.23775526881217957,
"rewards/rejected": -0.3122383654117584,
"step": 1340
},
{
"epoch": 0.36,
"learning_rate": 4.039153688314146e-06,
"logits/chosen": -1.5747339725494385,
"logits/rejected": -0.38576704263687134,
"logps/chosen": -659.3181762695312,
"logps/rejected": -1664.3912353515625,
"loss": 0.198,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.04063820466399193,
"rewards/margins": 0.32130542397499084,
"rewards/rejected": -0.36194363236427307,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 4.020749429372286e-06,
"logits/chosen": -1.5057035684585571,
"logits/rejected": -0.5099014043807983,
"logps/chosen": -759.4317626953125,
"logps/rejected": -1881.2880859375,
"loss": 0.1902,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08269429206848145,
"rewards/margins": 0.3579484820365906,
"rewards/rejected": -0.440642774105072,
"step": 1360
},
{
"epoch": 0.37,
"learning_rate": 4.002213403412492e-06,
"logits/chosen": -1.440815806388855,
"logits/rejected": -0.31881892681121826,
"logps/chosen": -595.3434448242188,
"logps/rejected": -1587.8076171875,
"loss": 0.2562,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07667971402406693,
"rewards/margins": 0.292102575302124,
"rewards/rejected": -0.36878231167793274,
"step": 1370
},
{
"epoch": 0.37,
"learning_rate": 3.983547216509254e-06,
"logits/chosen": -1.7196152210235596,
"logits/rejected": -0.5447965264320374,
"logps/chosen": -607.3961181640625,
"logps/rejected": -1578.577880859375,
"loss": 0.1834,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04431430250406265,
"rewards/margins": 0.25382199883461,
"rewards/rejected": -0.29813629388809204,
"step": 1380
},
{
"epoch": 0.37,
"learning_rate": 3.964752486015001e-06,
"logits/chosen": -1.5760372877120972,
"logits/rejected": -0.7092293500900269,
"logps/chosen": -662.7578125,
"logps/rejected": -1628.713623046875,
"loss": 0.1779,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06009943410754204,
"rewards/margins": 0.26038116216659546,
"rewards/rejected": -0.320480614900589,
"step": 1390
},
{
"epoch": 0.37,
"learning_rate": 3.945830840419966e-06,
"logits/chosen": -1.5873006582260132,
"logits/rejected": -0.4512631893157959,
"logps/chosen": -656.304931640625,
"logps/rejected": -1755.419677734375,
"loss": 0.1451,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.07909057289361954,
"rewards/margins": 0.36252719163894653,
"rewards/rejected": -0.44161778688430786,
"step": 1400
},
{
"epoch": 0.38,
"learning_rate": 3.92678391921108e-06,
"logits/chosen": -1.5479198694229126,
"logits/rejected": -0.4876307547092438,
"logps/chosen": -590.7452392578125,
"logps/rejected": -1592.341064453125,
"loss": 0.1647,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07574008405208588,
"rewards/margins": 0.30887579917907715,
"rewards/rejected": -0.38461586833000183,
"step": 1410
},
{
"epoch": 0.38,
"learning_rate": 3.907613372729916e-06,
"logits/chosen": -1.3529613018035889,
"logits/rejected": -0.5033882260322571,
"logps/chosen": -667.5601806640625,
"logps/rejected": -1620.945556640625,
"loss": 0.1869,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10878247022628784,
"rewards/margins": 0.27404552698135376,
"rewards/rejected": -0.3828279972076416,
"step": 1420
},
{
"epoch": 0.38,
"learning_rate": 3.888320862029699e-06,
"logits/chosen": -1.7743628025054932,
"logits/rejected": -0.5587132573127747,
"logps/chosen": -699.76318359375,
"logps/rejected": -1693.8353271484375,
"loss": 0.2028,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07925678789615631,
"rewards/margins": 0.30291515588760376,
"rewards/rejected": -0.38217195868492126,
"step": 1430
},
{
"epoch": 0.38,
"learning_rate": 3.868908058731376e-06,
"logits/chosen": -1.607465386390686,
"logits/rejected": -0.8050976991653442,
"logps/chosen": -772.9467163085938,
"logps/rejected": -1667.0992431640625,
"loss": 0.1693,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.05015742778778076,
"rewards/margins": 0.27269458770751953,
"rewards/rejected": -0.32285207509994507,
"step": 1440
},
{
"epoch": 0.39,
"learning_rate": 3.849376644878783e-06,
"logits/chosen": -1.4109113216400146,
"logits/rejected": -0.8917434811592102,
"logps/chosen": -581.6261596679688,
"logps/rejected": -1721.5927734375,
"loss": 0.1552,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.02441004104912281,
"rewards/margins": 0.3415711224079132,
"rewards/rejected": -0.36598116159439087,
"step": 1450
},
{
"epoch": 0.39,
"learning_rate": 3.829728312792895e-06,
"logits/chosen": -1.7239364385604858,
"logits/rejected": -0.6429109573364258,
"logps/chosen": -607.516357421875,
"logps/rejected": -1509.8909912109375,
"loss": 0.2134,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.0776829943060875,
"rewards/margins": 0.28966718912124634,
"rewards/rejected": -0.36735019087791443,
"step": 1460
},
{
"epoch": 0.39,
"learning_rate": 3.8099647649251984e-06,
"logits/chosen": -1.3290464878082275,
"logits/rejected": -0.5565693974494934,
"logps/chosen": -806.9801025390625,
"logps/rejected": -1813.9703369140625,
"loss": 0.1661,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13428995013237,
"rewards/margins": 0.3366580903530121,
"rewards/rejected": -0.4709479808807373,
"step": 1470
},
{
"epoch": 0.39,
"learning_rate": 3.790087713710179e-06,
"logits/chosen": -1.167667031288147,
"logits/rejected": -0.08988530933856964,
"logps/chosen": -961.8670654296875,
"logps/rejected": -1891.443115234375,
"loss": 0.2308,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2256636619567871,
"rewards/margins": 0.27781441807746887,
"rewards/rejected": -0.5034780502319336,
"step": 1480
},
{
"epoch": 0.4,
"learning_rate": 3.770098881416945e-06,
"logits/chosen": -1.445077896118164,
"logits/rejected": -0.7122173309326172,
"logps/chosen": -851.2200927734375,
"logps/rejected": -1791.288818359375,
"loss": 0.194,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18502262234687805,
"rewards/margins": 0.28681907057762146,
"rewards/rejected": -0.4718416631221771,
"step": 1490
},
{
"epoch": 0.4,
"learning_rate": 3.7500000000000005e-06,
"logits/chosen": -1.6007522344589233,
"logits/rejected": -0.3576156497001648,
"logps/chosen": -716.0091552734375,
"logps/rejected": -1709.4010009765625,
"loss": 0.1703,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12502431869506836,
"rewards/margins": 0.3000204563140869,
"rewards/rejected": -0.4250447750091553,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 3.7297928109491765e-06,
"logits/chosen": -1.4100292921066284,
"logits/rejected": -0.429446280002594,
"logps/chosen": -616.4102172851562,
"logps/rejected": -1615.930908203125,
"loss": 0.1452,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10738323628902435,
"rewards/margins": 0.33697018027305603,
"rewards/rejected": -0.4443534016609192,
"step": 1510
},
{
"epoch": 0.41,
"learning_rate": 3.7094790651387414e-06,
"logits/chosen": -1.6581932306289673,
"logits/rejected": -0.812160849571228,
"logps/chosen": -681.979736328125,
"logps/rejected": -1556.7125244140625,
"loss": 0.1798,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13339491188526154,
"rewards/margins": 0.25554531812667847,
"rewards/rejected": -0.3889401853084564,
"step": 1520
},
{
"epoch": 0.41,
"learning_rate": 3.689060522675689e-06,
"logits/chosen": -1.450721025466919,
"logits/rejected": -0.5784316062927246,
"logps/chosen": -761.6971435546875,
"logps/rejected": -1722.798583984375,
"loss": 0.2113,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1309405416250229,
"rewards/margins": 0.2867937386035919,
"rewards/rejected": -0.417734295129776,
"step": 1530
},
{
"epoch": 0.41,
"learning_rate": 3.668538952747236e-06,
"logits/chosen": -1.659879446029663,
"logits/rejected": -0.39928197860717773,
"logps/chosen": -810.2103881835938,
"logps/rejected": -1897.2626953125,
"loss": 0.1285,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12557505071163177,
"rewards/margins": 0.390308141708374,
"rewards/rejected": -0.515883207321167,
"step": 1540
},
{
"epoch": 0.41,
"learning_rate": 3.6479161334675294e-06,
"logits/chosen": -1.4575773477554321,
"logits/rejected": -0.07413512468338013,
"logps/chosen": -820.2180786132812,
"logps/rejected": -1623.4227294921875,
"loss": 0.2012,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.13560107350349426,
"rewards/margins": 0.23566505312919617,
"rewards/rejected": -0.37126606702804565,
"step": 1550
},
{
"epoch": 0.42,
"learning_rate": 3.627193851723577e-06,
"logits/chosen": -1.3455625772476196,
"logits/rejected": -0.9724159240722656,
"logps/chosen": -803.569091796875,
"logps/rejected": -1693.400390625,
"loss": 0.2043,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13027095794677734,
"rewards/margins": 0.29757434129714966,
"rewards/rejected": -0.4278453290462494,
"step": 1560
},
{
"epoch": 0.42,
"learning_rate": 3.6063739030204226e-06,
"logits/chosen": -1.5993826389312744,
"logits/rejected": -0.9014598727226257,
"logps/chosen": -670.8797607421875,
"logps/rejected": -1565.011474609375,
"loss": 0.1955,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.13589544594287872,
"rewards/margins": 0.2834901213645935,
"rewards/rejected": -0.4193855822086334,
"step": 1570
},
{
"epoch": 0.42,
"learning_rate": 3.5854580913255706e-06,
"logits/chosen": -1.425690770149231,
"logits/rejected": -0.18355034291744232,
"logps/chosen": -754.1428833007812,
"logps/rejected": -1659.813720703125,
"loss": 0.2144,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17976424098014832,
"rewards/margins": 0.26915082335472107,
"rewards/rejected": -0.4489150941371918,
"step": 1580
},
{
"epoch": 0.42,
"learning_rate": 3.564448228912682e-06,
"logits/chosen": -1.5854262113571167,
"logits/rejected": -0.6857632398605347,
"logps/chosen": -745.6060791015625,
"logps/rejected": -1763.920166015625,
"loss": 0.1558,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12223930656909943,
"rewards/margins": 0.31619927287101746,
"rewards/rejected": -0.4384385943412781,
"step": 1590
},
{
"epoch": 0.43,
"learning_rate": 3.543346136204545e-06,
"logits/chosen": -1.2614418268203735,
"logits/rejected": -0.3045172095298767,
"logps/chosen": -896.9788818359375,
"logps/rejected": -1856.7265625,
"loss": 0.1537,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1813734471797943,
"rewards/margins": 0.2908100187778473,
"rewards/rejected": -0.4721834659576416,
"step": 1600
},
{
"epoch": 0.43,
"learning_rate": 3.522153641615345e-06,
"logits/chosen": -1.4344582557678223,
"logits/rejected": -0.6459347009658813,
"logps/chosen": -729.3267822265625,
"logps/rejected": -1761.7333984375,
"loss": 0.1404,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12001194804906845,
"rewards/margins": 0.3599693179130554,
"rewards/rejected": -0.47998133301734924,
"step": 1610
},
{
"epoch": 0.43,
"learning_rate": 3.5008725813922383e-06,
"logits/chosen": -1.4531586170196533,
"logits/rejected": -0.5490698218345642,
"logps/chosen": -856.87255859375,
"logps/rejected": -1625.8597412109375,
"loss": 0.1619,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10974371433258057,
"rewards/margins": 0.3060295283794403,
"rewards/rejected": -0.4157732427120209,
"step": 1620
},
{
"epoch": 0.43,
"learning_rate": 3.4795047994562463e-06,
"logits/chosen": -1.4780563116073608,
"logits/rejected": -0.4708196222782135,
"logps/chosen": -758.4917602539062,
"logps/rejected": -1597.0938720703125,
"loss": 0.2097,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12127669155597687,
"rewards/margins": 0.24855844676494598,
"rewards/rejected": -0.36983510851860046,
"step": 1630
},
{
"epoch": 0.44,
"learning_rate": 3.458052147242494e-06,
"logits/chosen": -1.7074508666992188,
"logits/rejected": -0.8495294451713562,
"logps/chosen": -809.2350463867188,
"logps/rejected": -1472.79443359375,
"loss": 0.186,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.10575731843709946,
"rewards/margins": 0.18694370985031128,
"rewards/rejected": -0.29270103573799133,
"step": 1640
},
{
"epoch": 0.44,
"learning_rate": 3.436516483539781e-06,
"logits/chosen": -1.6757071018218994,
"logits/rejected": -0.4471355378627777,
"logps/chosen": -738.4520874023438,
"logps/rejected": -1533.0238037109375,
"loss": 0.1949,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09687335044145584,
"rewards/margins": 0.2056708037853241,
"rewards/rejected": -0.30254414677619934,
"step": 1650
},
{
"epoch": 0.44,
"learning_rate": 3.4148996743295305e-06,
"logits/chosen": -1.704223394393921,
"logits/rejected": -0.5917268991470337,
"logps/chosen": -850.6807861328125,
"logps/rejected": -1609.865478515625,
"loss": 0.2066,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.05562291666865349,
"rewards/margins": 0.2252390831708908,
"rewards/rejected": -0.2808619737625122,
"step": 1660
},
{
"epoch": 0.45,
"learning_rate": 3.3932035926241103e-06,
"logits/chosen": -1.5309293270111084,
"logits/rejected": -1.2061899900436401,
"logps/chosen": -625.7445068359375,
"logps/rejected": -1734.443603515625,
"loss": 0.1488,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05109323933720589,
"rewards/margins": 0.34637781977653503,
"rewards/rejected": -0.3974711000919342,
"step": 1670
},
{
"epoch": 0.45,
"learning_rate": 3.3714301183045382e-06,
"logits/chosen": -1.503655195236206,
"logits/rejected": -0.6576262712478638,
"logps/chosen": -691.9463500976562,
"logps/rejected": -1632.8353271484375,
"loss": 0.1711,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08215099573135376,
"rewards/margins": 0.270992249250412,
"rewards/rejected": -0.35314327478408813,
"step": 1680
},
{
"epoch": 0.45,
"learning_rate": 3.349581137957604e-06,
"logits/chosen": -1.7165721654891968,
"logits/rejected": -0.6651977896690369,
"logps/chosen": -709.5531005859375,
"logps/rejected": -1576.0364990234375,
"loss": 0.2304,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08810341358184814,
"rewards/margins": 0.2739308476448059,
"rewards/rejected": -0.36203423142433167,
"step": 1690
},
{
"epoch": 0.45,
"learning_rate": 3.3276585447123957e-06,
"logits/chosen": -1.232627034187317,
"logits/rejected": -0.1004166379570961,
"logps/chosen": -676.5740966796875,
"logps/rejected": -1472.2635498046875,
"loss": 0.2152,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10900652408599854,
"rewards/margins": 0.22487536072731018,
"rewards/rejected": -0.3338818848133087,
"step": 1700
},
{
"epoch": 0.46,
"learning_rate": 3.3056642380762783e-06,
"logits/chosen": -1.5267202854156494,
"logits/rejected": -0.49788492918014526,
"logps/chosen": -719.8569946289062,
"logps/rejected": -1706.168212890625,
"loss": 0.1489,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04703357070684433,
"rewards/margins": 0.3542724847793579,
"rewards/rejected": -0.40130606293678284,
"step": 1710
},
{
"epoch": 0.46,
"learning_rate": 3.2836001237702993e-06,
"logits/chosen": -1.6021124124526978,
"logits/rejected": -0.44486141204833984,
"logps/chosen": -737.7840576171875,
"logps/rejected": -1778.464111328125,
"loss": 0.1286,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.05765662342309952,
"rewards/margins": 0.3132147192955017,
"rewards/rejected": -0.370871365070343,
"step": 1720
},
{
"epoch": 0.46,
"learning_rate": 3.2614681135640696e-06,
"logits/chosen": -1.5471503734588623,
"logits/rejected": -0.8976337313652039,
"logps/chosen": -721.9874267578125,
"logps/rejected": -1760.84765625,
"loss": 0.161,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0736970454454422,
"rewards/margins": 0.24696488678455353,
"rewards/rejected": -0.3206619620323181,
"step": 1730
},
{
"epoch": 0.46,
"learning_rate": 3.2392701251101172e-06,
"logits/chosen": -1.4463790655136108,
"logits/rejected": -0.8106206059455872,
"logps/chosen": -668.1365356445312,
"logps/rejected": -1573.209716796875,
"loss": 0.141,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08170486986637115,
"rewards/margins": 0.264334499835968,
"rewards/rejected": -0.34603938460350037,
"step": 1740
},
{
"epoch": 0.47,
"learning_rate": 3.217008081777726e-06,
"logits/chosen": -1.5568472146987915,
"logits/rejected": -0.6958287954330444,
"logps/chosen": -861.89013671875,
"logps/rejected": -1898.191162109375,
"loss": 0.1875,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18643470108509064,
"rewards/margins": 0.3724428713321686,
"rewards/rejected": -0.5588775277137756,
"step": 1750
},
{
"epoch": 0.47,
"learning_rate": 3.1946839124862873e-06,
"logits/chosen": -1.314772367477417,
"logits/rejected": -0.18543431162834167,
"logps/chosen": -708.21923828125,
"logps/rejected": -1773.858642578125,
"loss": 0.1311,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12136778980493546,
"rewards/margins": 0.3536931574344635,
"rewards/rejected": -0.4750608801841736,
"step": 1760
},
{
"epoch": 0.47,
"learning_rate": 3.1722995515381644e-06,
"logits/chosen": -1.386833667755127,
"logits/rejected": -0.44682741165161133,
"logps/chosen": -727.7785034179688,
"logps/rejected": -1677.187744140625,
"loss": 0.2002,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10554149001836777,
"rewards/margins": 0.328396201133728,
"rewards/rejected": -0.4339376389980316,
"step": 1770
},
{
"epoch": 0.47,
"learning_rate": 3.149856938451094e-06,
"logits/chosen": -1.6613432168960571,
"logits/rejected": 0.04711759090423584,
"logps/chosen": -814.40380859375,
"logps/rejected": -1579.5726318359375,
"loss": 0.1541,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12790891528129578,
"rewards/margins": 0.259776771068573,
"rewards/rejected": -0.387685626745224,
"step": 1780
},
{
"epoch": 0.48,
"learning_rate": 3.127358017790132e-06,
"logits/chosen": -1.352160096168518,
"logits/rejected": 0.36388832330703735,
"logps/chosen": -850.2373046875,
"logps/rejected": -1793.983154296875,
"loss": 0.1351,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15040907263755798,
"rewards/margins": 0.26560917496681213,
"rewards/rejected": -0.4160182476043701,
"step": 1790
},
{
"epoch": 0.48,
"learning_rate": 3.1048047389991693e-06,
"logits/chosen": -1.3986834287643433,
"logits/rejected": -0.2482280433177948,
"logps/chosen": -609.4581298828125,
"logps/rejected": -1647.325439453125,
"loss": 0.1148,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.05277906730771065,
"rewards/margins": 0.2984154522418976,
"rewards/rejected": -0.35119450092315674,
"step": 1800
},
{
"epoch": 0.48,
"learning_rate": 3.082199056232015e-06,
"logits/chosen": -1.453552007675171,
"logits/rejected": -1.1569862365722656,
"logps/chosen": -593.3402709960938,
"logps/rejected": -1446.66943359375,
"loss": 0.2248,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06173533946275711,
"rewards/margins": 0.21113066375255585,
"rewards/rejected": -0.27286598086357117,
"step": 1810
},
{
"epoch": 0.49,
"learning_rate": 3.059542928183079e-06,
"logits/chosen": -1.123428463935852,
"logits/rejected": -0.1749972403049469,
"logps/chosen": -752.8836669921875,
"logps/rejected": -1819.65234375,
"loss": 0.1707,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07862022519111633,
"rewards/margins": 0.3714417815208435,
"rewards/rejected": -0.4500620365142822,
"step": 1820
},
{
"epoch": 0.49,
"learning_rate": 3.0368383179176584e-06,
"logits/chosen": -1.3128252029418945,
"logits/rejected": -0.5668991208076477,
"logps/chosen": -749.9387817382812,
"logps/rejected": -1737.239501953125,
"loss": 0.1744,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1024591326713562,
"rewards/margins": 0.33585435152053833,
"rewards/rejected": -0.43831348419189453,
"step": 1830
},
{
"epoch": 0.49,
"learning_rate": 3.0140871927018466e-06,
"logits/chosen": -1.5394883155822754,
"logits/rejected": -0.7085919380187988,
"logps/chosen": -835.1018676757812,
"logps/rejected": -1881.120361328125,
"loss": 0.1541,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16260461509227753,
"rewards/margins": 0.349585622549057,
"rewards/rejected": -0.5121902227401733,
"step": 1840
},
{
"epoch": 0.49,
"learning_rate": 2.9912915238320755e-06,
"logits/chosen": -1.323406457901001,
"logits/rejected": -0.5690110921859741,
"logps/chosen": -649.0794677734375,
"logps/rejected": -1694.8863525390625,
"loss": 0.1603,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.10500358045101166,
"rewards/margins": 0.30655449628829956,
"rewards/rejected": -0.41155806183815,
"step": 1850
},
{
"epoch": 0.5,
"learning_rate": 2.9684532864643123e-06,
"logits/chosen": -1.6879428625106812,
"logits/rejected": -0.8558928370475769,
"logps/chosen": -687.2265014648438,
"logps/rejected": -1546.723876953125,
"loss": 0.1733,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14953789114952087,
"rewards/margins": 0.27107977867126465,
"rewards/rejected": -0.4206176698207855,
"step": 1860
},
{
"epoch": 0.5,
"learning_rate": 2.945574459442917e-06,
"logits/chosen": -1.5647737979888916,
"logits/rejected": -1.020084261894226,
"logps/chosen": -731.6143798828125,
"logps/rejected": -1668.8765869140625,
"loss": 0.1614,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15018759667873383,
"rewards/margins": 0.3018895983695984,
"rewards/rejected": -0.4520772099494934,
"step": 1870
},
{
"epoch": 0.5,
"learning_rate": 2.922657025129185e-06,
"logits/chosen": -1.4925051927566528,
"logits/rejected": -0.517613410949707,
"logps/chosen": -759.2529296875,
"logps/rejected": -1762.791748046875,
"loss": 0.2552,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16193100810050964,
"rewards/margins": 0.29394176602363586,
"rewards/rejected": -0.4558727741241455,
"step": 1880
},
{
"epoch": 0.5,
"learning_rate": 2.8997029692295875e-06,
"logits/chosen": -1.5928837060928345,
"logits/rejected": -0.7866376042366028,
"logps/chosen": -774.6558837890625,
"logps/rejected": -1785.3062744140625,
"loss": 0.1509,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1210324764251709,
"rewards/margins": 0.31033051013946533,
"rewards/rejected": -0.43136295676231384,
"step": 1890
},
{
"epoch": 0.51,
"learning_rate": 2.876714280623708e-06,
"logits/chosen": -1.347715139389038,
"logits/rejected": -0.3572938144207001,
"logps/chosen": -739.8132934570312,
"logps/rejected": -1673.948486328125,
"loss": 0.1395,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0940694585442543,
"rewards/margins": 0.3482546806335449,
"rewards/rejected": -0.44232410192489624,
"step": 1900
},
{
"epoch": 0.51,
"learning_rate": 2.8536929511919227e-06,
"logits/chosen": -1.5348753929138184,
"logits/rejected": -0.0369141586124897,
"logps/chosen": -672.0413818359375,
"logps/rejected": -1597.8165283203125,
"loss": 0.1693,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09073442965745926,
"rewards/margins": 0.2980082929134369,
"rewards/rejected": -0.38874274492263794,
"step": 1910
},
{
"epoch": 0.51,
"learning_rate": 2.8306409756428067e-06,
"logits/chosen": -1.3153806924819946,
"logits/rejected": -0.5097673535346985,
"logps/chosen": -708.4708251953125,
"logps/rejected": -1756.6849365234375,
"loss": 0.1513,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13321314752101898,
"rewards/margins": 0.31301358342170715,
"rewards/rejected": -0.44622668623924255,
"step": 1920
},
{
"epoch": 0.51,
"learning_rate": 2.807560351340302e-06,
"logits/chosen": -1.4458258152008057,
"logits/rejected": -0.42417654395103455,
"logps/chosen": -671.51611328125,
"logps/rejected": -1660.485107421875,
"loss": 0.1559,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11455672979354858,
"rewards/margins": 0.3709821105003357,
"rewards/rejected": -0.48553887009620667,
"step": 1930
},
{
"epoch": 0.52,
"learning_rate": 2.7844530781306544e-06,
"logits/chosen": -1.3651716709136963,
"logits/rejected": -0.4336569905281067,
"logps/chosen": -840.4568481445312,
"logps/rejected": -2070.00146484375,
"loss": 0.1597,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19497820734977722,
"rewards/margins": 0.396168977022171,
"rewards/rejected": -0.5911471247673035,
"step": 1940
},
{
"epoch": 0.52,
"learning_rate": 2.761321158169134e-06,
"logits/chosen": -1.573979377746582,
"logits/rejected": -0.7406612038612366,
"logps/chosen": -825.8062744140625,
"logps/rejected": -1895.0921630859375,
"loss": 0.1527,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.156173974275589,
"rewards/margins": 0.32018476724624634,
"rewards/rejected": -0.4763587415218353,
"step": 1950
},
{
"epoch": 0.52,
"learning_rate": 2.738166595746554e-06,
"logits/chosen": -1.6485137939453125,
"logits/rejected": -0.31914329528808594,
"logps/chosen": -834.5794677734375,
"logps/rejected": -1904.170654296875,
"loss": 0.1606,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.15797320008277893,
"rewards/margins": 0.32314831018447876,
"rewards/rejected": -0.4811214506626129,
"step": 1960
},
{
"epoch": 0.53,
"learning_rate": 2.7149913971156105e-06,
"logits/chosen": -1.5849123001098633,
"logits/rejected": -0.718601644039154,
"logps/chosen": -758.3035278320312,
"logps/rejected": -1674.5123291015625,
"loss": 0.1629,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.15117041766643524,
"rewards/margins": 0.3155497610569,
"rewards/rejected": -0.4667201042175293,
"step": 1970
},
{
"epoch": 0.53,
"learning_rate": 2.6917975703170466e-06,
"logits/chosen": -1.2325398921966553,
"logits/rejected": -0.8877838850021362,
"logps/chosen": -891.3978271484375,
"logps/rejected": -1764.385498046875,
"loss": 0.149,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2210262268781662,
"rewards/margins": 0.3151123523712158,
"rewards/rejected": -0.536138653755188,
"step": 1980
},
{
"epoch": 0.53,
"learning_rate": 2.668587125005663e-06,
"logits/chosen": -1.5326184034347534,
"logits/rejected": -0.6966463923454285,
"logps/chosen": -733.0716552734375,
"logps/rejected": -1570.1944580078125,
"loss": 0.1168,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14229288697242737,
"rewards/margins": 0.30633196234703064,
"rewards/rejected": -0.44862478971481323,
"step": 1990
},
{
"epoch": 0.53,
"learning_rate": 2.6453620722761897e-06,
"logits/chosen": -1.5661652088165283,
"logits/rejected": -0.7556155920028687,
"logps/chosen": -820.5909423828125,
"logps/rejected": -1868.76171875,
"loss": 0.1276,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.21517057716846466,
"rewards/margins": 0.35130250453948975,
"rewards/rejected": -0.566473126411438,
"step": 2000
},
{
"epoch": 0.54,
"learning_rate": 2.6221244244890336e-06,
"logits/chosen": -1.37876558303833,
"logits/rejected": -0.1346241682767868,
"logps/chosen": -989.1995239257812,
"logps/rejected": -1853.1685791015625,
"loss": 0.1738,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2625274062156677,
"rewards/margins": 0.27383849024772644,
"rewards/rejected": -0.5363659858703613,
"step": 2010
},
{
"epoch": 0.54,
"learning_rate": 2.5988761950959133e-06,
"logits/chosen": -1.4494779109954834,
"logits/rejected": -0.5709508061408997,
"logps/chosen": -801.075927734375,
"logps/rejected": -1952.412841796875,
"loss": 0.1589,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.18583470582962036,
"rewards/margins": 0.39585351943969727,
"rewards/rejected": -0.5816881656646729,
"step": 2020
},
{
"epoch": 0.54,
"learning_rate": 2.575619398465402e-06,
"logits/chosen": -1.5121077299118042,
"logits/rejected": -0.7577626705169678,
"logps/chosen": -630.1747436523438,
"logps/rejected": -1437.3203125,
"loss": 0.2045,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12697990238666534,
"rewards/margins": 0.22890718281269073,
"rewards/rejected": -0.35588711500167847,
"step": 2030
},
{
"epoch": 0.54,
"learning_rate": 2.5523560497083927e-06,
"logits/chosen": -1.165709137916565,
"logits/rejected": -0.6048153638839722,
"logps/chosen": -719.2890625,
"logps/rejected": -1616.2135009765625,
"loss": 0.1598,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11988051235675812,
"rewards/margins": 0.26138466596603394,
"rewards/rejected": -0.38126516342163086,
"step": 2040
},
{
"epoch": 0.55,
"learning_rate": 2.5290881645034932e-06,
"logits/chosen": -1.6587855815887451,
"logits/rejected": -0.6381639838218689,
"logps/chosen": -765.1217041015625,
"logps/rejected": -1829.5943603515625,
"loss": 0.1613,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16699892282485962,
"rewards/margins": 0.37708956003189087,
"rewards/rejected": -0.5440884828567505,
"step": 2050
},
{
"epoch": 0.55,
"learning_rate": 2.5058177589223766e-06,
"logits/chosen": -1.379034399986267,
"logits/rejected": -0.6342862248420715,
"logps/chosen": -677.946533203125,
"logps/rejected": -1782.3040771484375,
"loss": 0.1529,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12249922752380371,
"rewards/margins": 0.374000608921051,
"rewards/rejected": -0.49649983644485474,
"step": 2060
},
{
"epoch": 0.55,
"learning_rate": 2.482546849255096e-06,
"logits/chosen": -1.2877978086471558,
"logits/rejected": -0.628576397895813,
"logps/chosen": -799.4703369140625,
"logps/rejected": -1640.319580078125,
"loss": 0.1346,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15292036533355713,
"rewards/margins": 0.2804059386253357,
"rewards/rejected": -0.4333263039588928,
"step": 2070
},
{
"epoch": 0.55,
"learning_rate": 2.4592774518353858e-06,
"logits/chosen": -1.4938738346099854,
"logits/rejected": -0.6422006487846375,
"logps/chosen": -752.1285400390625,
"logps/rejected": -1773.2620849609375,
"loss": 0.1651,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16756217181682587,
"rewards/margins": 0.3254481554031372,
"rewards/rejected": -0.49301037192344666,
"step": 2080
},
{
"epoch": 0.56,
"learning_rate": 2.436011582865945e-06,
"logits/chosen": -1.5365828275680542,
"logits/rejected": -0.6787657141685486,
"logps/chosen": -893.0496215820312,
"logps/rejected": -1954.225341796875,
"loss": 0.1604,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2155316323041916,
"rewards/margins": 0.3428398668766022,
"rewards/rejected": -0.5583714842796326,
"step": 2090
},
{
"epoch": 0.56,
"learning_rate": 2.4127512582437486e-06,
"logits/chosen": -1.1174455881118774,
"logits/rejected": 0.26067572832107544,
"logps/chosen": -777.9120483398438,
"logps/rejected": -1707.302001953125,
"loss": 0.1845,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.17711251974105835,
"rewards/margins": 0.2716800570487976,
"rewards/rejected": -0.44879254698753357,
"step": 2100
},
{
"epoch": 0.56,
"learning_rate": 2.3894984933853734e-06,
"logits/chosen": -1.5128055810928345,
"logits/rejected": -0.5743480920791626,
"logps/chosen": -856.6569213867188,
"logps/rejected": -1764.4976806640625,
"loss": 0.1695,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1948443353176117,
"rewards/margins": 0.25961554050445557,
"rewards/rejected": -0.4544598460197449,
"step": 2110
},
{
"epoch": 0.57,
"learning_rate": 2.366255303052377e-06,
"logits/chosen": -1.5328229665756226,
"logits/rejected": -0.41585612297058105,
"logps/chosen": -865.1114501953125,
"logps/rejected": -1857.373046875,
"loss": 0.1726,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16477537155151367,
"rewards/margins": 0.2972187399864197,
"rewards/rejected": -0.46199408173561096,
"step": 2120
},
{
"epoch": 0.57,
"learning_rate": 2.3430237011767166e-06,
"logits/chosen": -1.5353227853775024,
"logits/rejected": 0.3234299123287201,
"logps/chosen": -791.5120239257812,
"logps/rejected": -2002.2496337890625,
"loss": 0.1074,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.15296730399131775,
"rewards/margins": 0.41708940267562866,
"rewards/rejected": -0.5700567960739136,
"step": 2130
},
{
"epoch": 0.57,
"learning_rate": 2.319805700686257e-06,
"logits/chosen": -1.2240071296691895,
"logits/rejected": -0.7209154963493347,
"logps/chosen": -790.4075927734375,
"logps/rejected": -1767.944580078125,
"loss": 0.1377,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12179949134588242,
"rewards/margins": 0.326397180557251,
"rewards/rejected": -0.4481966495513916,
"step": 2140
},
{
"epoch": 0.57,
"learning_rate": 2.296603313330355e-06,
"logits/chosen": -1.6259534358978271,
"logits/rejected": -0.29775649309158325,
"logps/chosen": -746.3842163085938,
"logps/rejected": -1520.571533203125,
"loss": 0.1711,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.13973134756088257,
"rewards/margins": 0.26066604256629944,
"rewards/rejected": -0.400397390127182,
"step": 2150
},
{
"epoch": 0.58,
"learning_rate": 2.2734185495055503e-06,
"logits/chosen": -1.2232530117034912,
"logits/rejected": -0.18962112069129944,
"logps/chosen": -576.0113525390625,
"logps/rejected": -1693.240966796875,
"loss": 0.1384,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13496878743171692,
"rewards/margins": 0.3316168785095215,
"rewards/rejected": -0.466585636138916,
"step": 2160
},
{
"epoch": 0.58,
"learning_rate": 2.250253418081373e-06,
"logits/chosen": -1.4692437648773193,
"logits/rejected": 0.04818035289645195,
"logps/chosen": -899.2703857421875,
"logps/rejected": -1808.1484375,
"loss": 0.2009,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21512703597545624,
"rewards/margins": 0.30080264806747437,
"rewards/rejected": -0.5159296989440918,
"step": 2170
},
{
"epoch": 0.58,
"learning_rate": 2.22710992622628e-06,
"logits/chosen": -1.5507208108901978,
"logits/rejected": 0.1027313843369484,
"logps/chosen": -805.1966552734375,
"logps/rejected": -1811.1793212890625,
"loss": 0.1454,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13587817549705505,
"rewards/margins": 0.39650699496269226,
"rewards/rejected": -0.5323852300643921,
"step": 2180
},
{
"epoch": 0.58,
"learning_rate": 2.2039900792337477e-06,
"logits/chosen": -1.395817756652832,
"logits/rejected": -0.45912352204322815,
"logps/chosen": -768.6142578125,
"logps/rejected": -1828.7073974609375,
"loss": 0.1529,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1312076300382614,
"rewards/margins": 0.3623288571834564,
"rewards/rejected": -0.4935365319252014,
"step": 2190
},
{
"epoch": 0.59,
"learning_rate": 2.1808958803485134e-06,
"logits/chosen": -1.5901384353637695,
"logits/rejected": -0.8347817659378052,
"logps/chosen": -523.8443603515625,
"logps/rejected": -1578.4923095703125,
"loss": 0.1123,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07345361262559891,
"rewards/margins": 0.34433066844940186,
"rewards/rejected": -0.41778427362442017,
"step": 2200
},
{
"epoch": 0.59,
"learning_rate": 2.157829330593008e-06,
"logits/chosen": -1.5194613933563232,
"logits/rejected": 0.044505536556243896,
"logps/chosen": -846.5181884765625,
"logps/rejected": -1842.278564453125,
"loss": 0.1109,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.18009065091609955,
"rewards/margins": 0.37484800815582275,
"rewards/rejected": -0.5549386739730835,
"step": 2210
},
{
"epoch": 0.59,
"learning_rate": 2.134792428593971e-06,
"logits/chosen": -1.517464280128479,
"logits/rejected": -0.3259011209011078,
"logps/chosen": -696.5281982421875,
"logps/rejected": -1474.2158203125,
"loss": 0.1964,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.13613803684711456,
"rewards/margins": 0.25182804465293884,
"rewards/rejected": -0.3879660964012146,
"step": 2220
},
{
"epoch": 0.59,
"learning_rate": 2.1117871704092818e-06,
"logits/chosen": -1.5486299991607666,
"logits/rejected": -0.5334731936454773,
"logps/chosen": -767.1392822265625,
"logps/rejected": -1842.3082275390625,
"loss": 0.121,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1175505518913269,
"rewards/margins": 0.3775568902492523,
"rewards/rejected": -0.4951074719429016,
"step": 2230
},
{
"epoch": 0.6,
"learning_rate": 2.0888155493550027e-06,
"logits/chosen": -1.5350594520568848,
"logits/rejected": -0.2820148468017578,
"logps/chosen": -626.4788208007812,
"logps/rejected": -1641.033447265625,
"loss": 0.1776,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07961525768041611,
"rewards/margins": 0.3609825670719147,
"rewards/rejected": -0.4405978322029114,
"step": 2240
},
{
"epoch": 0.6,
"learning_rate": 2.0658795558326745e-06,
"logits/chosen": -1.458655595779419,
"logits/rejected": -0.2539186477661133,
"logps/chosen": -764.8986206054688,
"logps/rejected": -1802.0277099609375,
"loss": 0.1842,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09169100970029831,
"rewards/margins": 0.3511391580104828,
"rewards/rejected": -0.4428301751613617,
"step": 2250
},
{
"epoch": 0.6,
"learning_rate": 2.0429811771568468e-06,
"logits/chosen": -1.5172240734100342,
"logits/rejected": -0.4895601272583008,
"logps/chosen": -801.1700439453125,
"logps/rejected": -1634.53955078125,
"loss": 0.165,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1120598167181015,
"rewards/margins": 0.27336305379867554,
"rewards/rejected": -0.38542285561561584,
"step": 2260
},
{
"epoch": 0.61,
"learning_rate": 2.0201223973828917e-06,
"logits/chosen": -1.627150535583496,
"logits/rejected": -0.6142350435256958,
"logps/chosen": -660.062744140625,
"logps/rejected": -1752.922119140625,
"loss": 0.1921,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08810608088970184,
"rewards/margins": 0.36186718940734863,
"rewards/rejected": -0.44997328519821167,
"step": 2270
},
{
"epoch": 0.61,
"learning_rate": 1.997305197135089e-06,
"logits/chosen": -1.5025413036346436,
"logits/rejected": -0.59123295545578,
"logps/chosen": -817.2398681640625,
"logps/rejected": -1722.5006103515625,
"loss": 0.1627,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10656937211751938,
"rewards/margins": 0.2904852032661438,
"rewards/rejected": -0.3970545828342438,
"step": 2280
},
{
"epoch": 0.61,
"learning_rate": 1.9745315534350157e-06,
"logits/chosen": -1.5288734436035156,
"logits/rejected": -0.7481527328491211,
"logps/chosen": -627.6370849609375,
"logps/rejected": -1676.7398681640625,
"loss": 0.1255,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09700825065374374,
"rewards/margins": 0.31658482551574707,
"rewards/rejected": -0.4135931134223938,
"step": 2290
},
{
"epoch": 0.61,
"learning_rate": 1.9518034395302413e-06,
"logits/chosen": -1.2082513570785522,
"logits/rejected": -0.5326002836227417,
"logps/chosen": -606.5856323242188,
"logps/rejected": -1724.590087890625,
"loss": 0.1711,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08526596426963806,
"rewards/margins": 0.3748168349266052,
"rewards/rejected": -0.4600828289985657,
"step": 2300
},
{
"epoch": 0.62,
"learning_rate": 1.9291228247233607e-06,
"logits/chosen": -1.349005103111267,
"logits/rejected": -0.7115123867988586,
"logps/chosen": -850.7883911132812,
"logps/rejected": -1918.0609130859375,
"loss": 0.1049,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1696614772081375,
"rewards/margins": 0.3336087763309479,
"rewards/rejected": -0.5032702684402466,
"step": 2310
},
{
"epoch": 0.62,
"learning_rate": 1.9064916742013515e-06,
"logits/chosen": -1.4064449071884155,
"logits/rejected": -0.41218453645706177,
"logps/chosen": -922.0281372070312,
"logps/rejected": -1926.2939453125,
"loss": 0.1868,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.20819124579429626,
"rewards/margins": 0.350273996591568,
"rewards/rejected": -0.558465301990509,
"step": 2320
},
{
"epoch": 0.62,
"learning_rate": 1.883911948865306e-06,
"logits/chosen": -1.4412527084350586,
"logits/rejected": -0.3628607392311096,
"logps/chosen": -686.2496948242188,
"logps/rejected": -1755.6363525390625,
"loss": 0.1658,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1295745074748993,
"rewards/margins": 0.372491717338562,
"rewards/rejected": -0.5020662546157837,
"step": 2330
},
{
"epoch": 0.62,
"learning_rate": 1.8613856051605242e-06,
"logits/chosen": -1.6775572299957275,
"logits/rejected": -0.6549306511878967,
"logps/chosen": -597.9432373046875,
"logps/rejected": -1609.9539794921875,
"loss": 0.137,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10794492810964584,
"rewards/margins": 0.3688591420650482,
"rewards/rejected": -0.47680407762527466,
"step": 2340
},
{
"epoch": 0.63,
"learning_rate": 1.8389145949069953e-06,
"logits/chosen": -1.382621169090271,
"logits/rejected": -0.4452172815799713,
"logps/chosen": -683.6560668945312,
"logps/rejected": -1507.298095703125,
"loss": 0.2337,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.110257588326931,
"rewards/margins": 0.25430378317832947,
"rewards/rejected": -0.36456140875816345,
"step": 2350
},
{
"epoch": 0.63,
"learning_rate": 1.816500865130279e-06,
"logits/chosen": -1.467057228088379,
"logits/rejected": -0.31321266293525696,
"logps/chosen": -672.0931396484375,
"logps/rejected": -1508.775634765625,
"loss": 0.1942,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09393011033535004,
"rewards/margins": 0.24138036370277405,
"rewards/rejected": -0.33531051874160767,
"step": 2360
},
{
"epoch": 0.63,
"learning_rate": 1.7941463578928088e-06,
"logits/chosen": -1.6495239734649658,
"logits/rejected": -0.6112552285194397,
"logps/chosen": -657.7040405273438,
"logps/rejected": -1789.280029296875,
"loss": 0.149,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07512945681810379,
"rewards/margins": 0.32172054052352905,
"rewards/rejected": -0.39685001969337463,
"step": 2370
},
{
"epoch": 0.63,
"learning_rate": 1.7718530101256115e-06,
"logits/chosen": -1.4681954383850098,
"logits/rejected": -0.752936840057373,
"logps/chosen": -775.773193359375,
"logps/rejected": -1576.1943359375,
"loss": 0.2381,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12784716486930847,
"rewards/margins": 0.20212575793266296,
"rewards/rejected": -0.32997292280197144,
"step": 2380
},
{
"epoch": 0.64,
"learning_rate": 1.7496227534604859e-06,
"logits/chosen": -1.6714906692504883,
"logits/rejected": -1.0251556634902954,
"logps/chosen": -675.6273803710938,
"logps/rejected": -1630.176025390625,
"loss": 0.1692,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.0969114676117897,
"rewards/margins": 0.361731618642807,
"rewards/rejected": -0.4586430490016937,
"step": 2390
},
{
"epoch": 0.64,
"learning_rate": 1.7274575140626318e-06,
"logits/chosen": -1.5120388269424438,
"logits/rejected": -0.5438052415847778,
"logps/chosen": -692.0360717773438,
"logps/rejected": -1721.4326171875,
"loss": 0.1631,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11012852191925049,
"rewards/margins": 0.3708108365535736,
"rewards/rejected": -0.4809393882751465,
"step": 2400
},
{
"epoch": 0.64,
"learning_rate": 1.7053592124637557e-06,
"logits/chosen": -1.6181414127349854,
"logits/rejected": -0.6608393788337708,
"logps/chosen": -747.2515869140625,
"logps/rejected": -1643.6500244140625,
"loss": 0.2218,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.13988645374774933,
"rewards/margins": 0.2650088965892792,
"rewards/rejected": -0.4048953652381897,
"step": 2410
},
{
"epoch": 0.65,
"learning_rate": 1.6833297633956647e-06,
"logits/chosen": -1.3472005128860474,
"logits/rejected": 0.12191818654537201,
"logps/chosen": -693.1132202148438,
"logps/rejected": -1696.821533203125,
"loss": 0.1063,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07493006438016891,
"rewards/margins": 0.32017913460731506,
"rewards/rejected": -0.3951091766357422,
"step": 2420
},
{
"epoch": 0.65,
"learning_rate": 1.661371075624363e-06,
"logits/chosen": -1.4592828750610352,
"logits/rejected": -0.45236214995384216,
"logps/chosen": -696.2838745117188,
"logps/rejected": -1729.255126953125,
"loss": 0.1603,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.061267025768756866,
"rewards/margins": 0.35353952646255493,
"rewards/rejected": -0.4148065447807312,
"step": 2430
},
{
"epoch": 0.65,
"learning_rate": 1.6394850517846621e-06,
"logits/chosen": -1.6494789123535156,
"logits/rejected": -0.9741545915603638,
"logps/chosen": -759.4490966796875,
"logps/rejected": -1714.441162109375,
"loss": 0.1556,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10241154581308365,
"rewards/margins": 0.3450758159160614,
"rewards/rejected": -0.44748735427856445,
"step": 2440
},
{
"epoch": 0.65,
"learning_rate": 1.6176735882153284e-06,
"logits/chosen": -1.468384027481079,
"logits/rejected": -0.16317103803157806,
"logps/chosen": -605.1905517578125,
"logps/rejected": -1460.53955078125,
"loss": 0.1345,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0676204189658165,
"rewards/margins": 0.30662956833839417,
"rewards/rejected": -0.37424999475479126,
"step": 2450
},
{
"epoch": 0.66,
"learning_rate": 1.5959385747947697e-06,
"logits/chosen": -1.4594337940216064,
"logits/rejected": -0.8179009556770325,
"logps/chosen": -606.8065185546875,
"logps/rejected": -1657.3125,
"loss": 0.1104,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06961175799369812,
"rewards/margins": 0.3652622103691101,
"rewards/rejected": -0.4348739683628082,
"step": 2460
},
{
"epoch": 0.66,
"learning_rate": 1.5742818947772875e-06,
"logits/chosen": -1.5698840618133545,
"logits/rejected": 0.06299029290676117,
"logps/chosen": -768.6961669921875,
"logps/rejected": -1631.1295166015625,
"loss": 0.137,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09274804592132568,
"rewards/margins": 0.27431541681289673,
"rewards/rejected": -0.3670634627342224,
"step": 2470
},
{
"epoch": 0.66,
"learning_rate": 1.552705424629898e-06,
"logits/chosen": -1.4378259181976318,
"logits/rejected": -0.6328508257865906,
"logps/chosen": -757.0548095703125,
"logps/rejected": -1625.8509521484375,
"loss": 0.1496,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09750431030988693,
"rewards/margins": 0.25974932312965393,
"rewards/rejected": -0.35725364089012146,
"step": 2480
},
{
"epoch": 0.66,
"learning_rate": 1.5312110338697427e-06,
"logits/chosen": -1.429025411605835,
"logits/rejected": -1.0502439737319946,
"logps/chosen": -620.97998046875,
"logps/rejected": -1421.07470703125,
"loss": 0.1467,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.0735950618982315,
"rewards/margins": 0.2558991312980652,
"rewards/rejected": -0.3294941782951355,
"step": 2490
},
{
"epoch": 0.67,
"learning_rate": 1.509800584902108e-06,
"logits/chosen": -1.2865421772003174,
"logits/rejected": -0.21516099572181702,
"logps/chosen": -712.9671630859375,
"logps/rejected": -1577.719970703125,
"loss": 0.1254,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08029041439294815,
"rewards/margins": 0.2842417359352112,
"rewards/rejected": -0.36453211307525635,
"step": 2500
},
{
"epoch": 0.67,
"learning_rate": 1.4884759328590476e-06,
"logits/chosen": -1.7966272830963135,
"logits/rejected": -0.5356184244155884,
"logps/chosen": -774.8032836914062,
"logps/rejected": -1728.052490234375,
"loss": 0.1569,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13011132180690765,
"rewards/margins": 0.38214725255966187,
"rewards/rejected": -0.5122585892677307,
"step": 2510
},
{
"epoch": 0.67,
"learning_rate": 1.467238925438646e-06,
"logits/chosen": -1.6186519861221313,
"logits/rejected": 0.41130122542381287,
"logps/chosen": -872.7478637695312,
"logps/rejected": -1909.5640869140625,
"loss": 0.148,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1565207540988922,
"rewards/margins": 0.3737161159515381,
"rewards/rejected": -0.5302368402481079,
"step": 2520
},
{
"epoch": 0.67,
"learning_rate": 1.446091402744923e-06,
"logits/chosen": -1.2679738998413086,
"logits/rejected": -0.5303937196731567,
"logps/chosen": -702.9267578125,
"logps/rejected": -1642.1829833984375,
"loss": 0.1461,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1064765453338623,
"rewards/margins": 0.34039077162742615,
"rewards/rejected": -0.44686728715896606,
"step": 2530
},
{
"epoch": 0.68,
"learning_rate": 1.4250351971283937e-06,
"logits/chosen": -1.604107141494751,
"logits/rejected": 0.4407750070095062,
"logps/chosen": -770.2025146484375,
"logps/rejected": -1616.8951416015625,
"loss": 0.1425,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11423502117395401,
"rewards/margins": 0.29226523637771606,
"rewards/rejected": -0.4065002501010895,
"step": 2540
},
{
"epoch": 0.68,
"learning_rate": 1.4040721330273063e-06,
"logits/chosen": -1.4690310955047607,
"logits/rejected": 0.40654927492141724,
"logps/chosen": -752.0418090820312,
"logps/rejected": -1654.997802734375,
"loss": 0.1918,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11529115587472916,
"rewards/margins": 0.2776246964931488,
"rewards/rejected": -0.392915815114975,
"step": 2550
},
{
"epoch": 0.68,
"learning_rate": 1.3832040268095589e-06,
"logits/chosen": -1.2964470386505127,
"logits/rejected": -0.35982269048690796,
"logps/chosen": -707.73681640625,
"logps/rejected": -1798.7490234375,
"loss": 0.1272,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.09786094725131989,
"rewards/margins": 0.3779224455356598,
"rewards/rejected": -0.4757833480834961,
"step": 2560
},
{
"epoch": 0.69,
"learning_rate": 1.362432686615316e-06,
"logits/chosen": -1.4789488315582275,
"logits/rejected": -0.4336570203304291,
"logps/chosen": -560.9461669921875,
"logps/rejected": -1716.0234375,
"loss": 0.164,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.058935634791851044,
"rewards/margins": 0.3738183081150055,
"rewards/rejected": -0.43275389075279236,
"step": 2570
},
{
"epoch": 0.69,
"learning_rate": 1.3417599122003464e-06,
"logits/chosen": -1.5111268758773804,
"logits/rejected": -0.40306025743484497,
"logps/chosen": -659.0839233398438,
"logps/rejected": -1699.853515625,
"loss": 0.1041,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08692941069602966,
"rewards/margins": 0.35781151056289673,
"rewards/rejected": -0.4447408616542816,
"step": 2580
},
{
"epoch": 0.69,
"learning_rate": 1.3211874947800747e-06,
"logits/chosen": -1.552757978439331,
"logits/rejected": -0.733113169670105,
"logps/chosen": -668.4918212890625,
"logps/rejected": -1597.004150390625,
"loss": 0.1547,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.09527282416820526,
"rewards/margins": 0.2590247094631195,
"rewards/rejected": -0.35429757833480835,
"step": 2590
},
{
"epoch": 0.69,
"learning_rate": 1.3007172168743854e-06,
"logits/chosen": -1.4352308511734009,
"logits/rejected": -0.04908560588955879,
"logps/chosen": -665.5040893554688,
"logps/rejected": -1593.777587890625,
"loss": 0.1445,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08846768736839294,
"rewards/margins": 0.32587721943855286,
"rewards/rejected": -0.4143448770046234,
"step": 2600
},
{
"epoch": 0.7,
"learning_rate": 1.280350852153168e-06,
"logits/chosen": -1.3806612491607666,
"logits/rejected": -0.20799100399017334,
"logps/chosen": -762.9417724609375,
"logps/rejected": -1701.4697265625,
"loss": 0.148,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15475930273532867,
"rewards/margins": 0.29663509130477905,
"rewards/rejected": -0.45139437913894653,
"step": 2610
},
{
"epoch": 0.7,
"learning_rate": 1.260090165282645e-06,
"logits/chosen": -1.4364516735076904,
"logits/rejected": 0.2683382034301758,
"logps/chosen": -701.4534912109375,
"logps/rejected": -1679.4847412109375,
"loss": 0.1688,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.19048824906349182,
"rewards/margins": 0.27079007029533386,
"rewards/rejected": -0.46127834916114807,
"step": 2620
},
{
"epoch": 0.7,
"learning_rate": 1.2399369117724582e-06,
"logits/chosen": -1.4400079250335693,
"logits/rejected": -0.4409395158290863,
"logps/chosen": -797.1912841796875,
"logps/rejected": -1693.439453125,
"loss": 0.1558,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12387750297784805,
"rewards/margins": 0.3022800087928772,
"rewards/rejected": -0.42615753412246704,
"step": 2630
},
{
"epoch": 0.7,
"learning_rate": 1.2198928378235717e-06,
"logits/chosen": -1.5424561500549316,
"logits/rejected": 0.6223300695419312,
"logps/chosen": -694.9400024414062,
"logps/rejected": -1691.6669921875,
"loss": 0.1657,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0836716815829277,
"rewards/margins": 0.3352142870426178,
"rewards/rejected": -0.4188859462738037,
"step": 2640
},
{
"epoch": 0.71,
"learning_rate": 1.1999596801769617e-06,
"logits/chosen": -1.6569328308105469,
"logits/rejected": -0.4357103407382965,
"logps/chosen": -693.4281005859375,
"logps/rejected": -1608.7647705078125,
"loss": 0.1473,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08369255810976028,
"rewards/margins": 0.2921767234802246,
"rewards/rejected": -0.3758693337440491,
"step": 2650
},
{
"epoch": 0.71,
"learning_rate": 1.1801391659631423e-06,
"logits/chosen": -1.5653746128082275,
"logits/rejected": 0.7312607169151306,
"logps/chosen": -660.3027954101562,
"logps/rejected": -1588.2919921875,
"loss": 0.1214,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06633373349905014,
"rewards/margins": 0.30183035135269165,
"rewards/rejected": -0.3681640923023224,
"step": 2660
},
{
"epoch": 0.71,
"learning_rate": 1.160433012552508e-06,
"logits/chosen": -1.5090980529785156,
"logits/rejected": -0.8780859112739563,
"logps/chosen": -695.6319580078125,
"logps/rejected": -1500.2325439453125,
"loss": 0.1882,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.07338380068540573,
"rewards/margins": 0.2535490393638611,
"rewards/rejected": -0.326932817697525,
"step": 2670
},
{
"epoch": 0.71,
"learning_rate": 1.1408429274065418e-06,
"logits/chosen": -1.3488575220108032,
"logits/rejected": -0.6483387351036072,
"logps/chosen": -572.9019165039062,
"logps/rejected": -1599.5225830078125,
"loss": 0.1575,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07061342149972916,
"rewards/margins": 0.2944713532924652,
"rewards/rejected": -0.36508476734161377,
"step": 2680
},
{
"epoch": 0.72,
"learning_rate": 1.1213706079298566e-06,
"logits/chosen": -1.4909251928329468,
"logits/rejected": -0.27080297470092773,
"logps/chosen": -564.2769165039062,
"logps/rejected": -1579.8958740234375,
"loss": 0.1631,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.053073953837156296,
"rewards/margins": 0.32997792959213257,
"rewards/rejected": -0.38305193185806274,
"step": 2690
},
{
"epoch": 0.72,
"learning_rate": 1.1020177413231334e-06,
"logits/chosen": -1.5161458253860474,
"logits/rejected": -0.2539205551147461,
"logps/chosen": -698.4984741210938,
"logps/rejected": -1621.921630859375,
"loss": 0.1693,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09174446761608124,
"rewards/margins": 0.29953649640083313,
"rewards/rejected": -0.3912809491157532,
"step": 2700
},
{
"epoch": 0.72,
"learning_rate": 1.0827860044369226e-06,
"logits/chosen": -1.5986204147338867,
"logits/rejected": -0.2638501226902008,
"logps/chosen": -766.909423828125,
"logps/rejected": -1732.082275390625,
"loss": 0.1602,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12383983284235,
"rewards/margins": 0.32002753019332886,
"rewards/rejected": -0.44386744499206543,
"step": 2710
},
{
"epoch": 0.73,
"learning_rate": 1.06367706362636e-06,
"logits/chosen": -1.4342091083526611,
"logits/rejected": -0.8792727589607239,
"logps/chosen": -737.6398315429688,
"logps/rejected": -1701.187255859375,
"loss": 0.1727,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1248563751578331,
"rewards/margins": 0.2940226197242737,
"rewards/rejected": -0.41887903213500977,
"step": 2720
},
{
"epoch": 0.73,
"learning_rate": 1.0446925746067768e-06,
"logits/chosen": -1.3923488855361938,
"logits/rejected": -0.20048478245735168,
"logps/chosen": -803.1317138671875,
"logps/rejected": -1872.0126953125,
"loss": 0.1227,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1411004215478897,
"rewards/margins": 0.36530831456184387,
"rewards/rejected": -0.5064087510108948,
"step": 2730
},
{
"epoch": 0.73,
"learning_rate": 1.0258341823102418e-06,
"logits/chosen": -1.5749719142913818,
"logits/rejected": -0.40622204542160034,
"logps/chosen": -757.8853759765625,
"logps/rejected": -1855.6015625,
"loss": 0.1441,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1515471190214157,
"rewards/margins": 0.4066466689109802,
"rewards/rejected": -0.5581938624382019,
"step": 2740
},
{
"epoch": 0.73,
"learning_rate": 1.0071035207430352e-06,
"logits/chosen": -1.6754239797592163,
"logits/rejected": -0.716029167175293,
"logps/chosen": -787.7699584960938,
"logps/rejected": -1843.915283203125,
"loss": 0.1577,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1490515172481537,
"rewards/margins": 0.3542669713497162,
"rewards/rejected": -0.5033184885978699,
"step": 2750
},
{
"epoch": 0.74,
"learning_rate": 9.88502212844063e-07,
"logits/chosen": -1.3399386405944824,
"logits/rejected": -0.6510659456253052,
"logps/chosen": -645.7489013671875,
"logps/rejected": -1645.0814208984375,
"loss": 0.1528,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12475794553756714,
"rewards/margins": 0.3248399794101715,
"rewards/rejected": -0.44959789514541626,
"step": 2760
},
{
"epoch": 0.74,
"learning_rate": 9.700318703442437e-07,
"logits/chosen": -1.5086935758590698,
"logits/rejected": -0.44206541776657104,
"logps/chosen": -751.8470458984375,
"logps/rejected": -1835.744873046875,
"loss": 0.1461,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0970626249909401,
"rewards/margins": 0.4153032898902893,
"rewards/rejected": -0.5123659372329712,
"step": 2770
},
{
"epoch": 0.74,
"learning_rate": 9.516940936268504e-07,
"logits/chosen": -1.4146547317504883,
"logits/rejected": -0.39967912435531616,
"logps/chosen": -655.8258666992188,
"logps/rejected": -1597.619873046875,
"loss": 0.1199,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.10331164300441742,
"rewards/margins": 0.29826727509498596,
"rewards/rejected": -0.4015789031982422,
"step": 2780
},
{
"epoch": 0.74,
"learning_rate": 9.334904715888496e-07,
"logits/chosen": -1.5796334743499756,
"logits/rejected": -0.18904080986976624,
"logps/chosen": -649.2426147460938,
"logps/rejected": -1707.6396484375,
"loss": 0.1682,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08159293234348297,
"rewards/margins": 0.3529675602912903,
"rewards/rejected": -0.43456047773361206,
"step": 2790
},
{
"epoch": 0.75,
"learning_rate": 9.154225815032242e-07,
"logits/chosen": -1.5274991989135742,
"logits/rejected": -0.7836966514587402,
"logps/chosen": -610.8500366210938,
"logps/rejected": -1656.6763916015625,
"loss": 0.1472,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06673729419708252,
"rewards/margins": 0.36455440521240234,
"rewards/rejected": -0.4312916696071625,
"step": 2800
},
{
"epoch": 0.75,
"learning_rate": 8.974919888823164e-07,
"logits/chosen": -1.4849748611450195,
"logits/rejected": -0.4282095432281494,
"logps/chosen": -743.9561767578125,
"logps/rejected": -1814.4990234375,
"loss": 0.1699,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09201905876398087,
"rewards/margins": 0.36516329646110535,
"rewards/rejected": -0.4571823179721832,
"step": 2810
},
{
"epoch": 0.75,
"learning_rate": 8.797002473421729e-07,
"logits/chosen": -1.4566679000854492,
"logits/rejected": -0.5152000188827515,
"logps/chosen": -523.7813720703125,
"logps/rejected": -1502.1715087890625,
"loss": 0.1961,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.058007530868053436,
"rewards/margins": 0.3089088201522827,
"rewards/rejected": -0.36691635847091675,
"step": 2820
},
{
"epoch": 0.75,
"learning_rate": 8.620488984679378e-07,
"logits/chosen": -1.7398220300674438,
"logits/rejected": -0.2614799737930298,
"logps/chosen": -641.33447265625,
"logps/rejected": -1628.4310302734375,
"loss": 0.1416,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.05575539544224739,
"rewards/margins": 0.34235674142837524,
"rewards/rejected": -0.39811214804649353,
"step": 2830
},
{
"epoch": 0.76,
"learning_rate": 8.445394716802754e-07,
"logits/chosen": -1.5343586206436157,
"logits/rejected": -0.6991978883743286,
"logps/chosen": -695.736328125,
"logps/rejected": -1697.3258056640625,
"loss": 0.17,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0863880068063736,
"rewards/margins": 0.3482551872730255,
"rewards/rejected": -0.4346431791782379,
"step": 2840
},
{
"epoch": 0.76,
"learning_rate": 8.271734841028553e-07,
"logits/chosen": -1.3757431507110596,
"logits/rejected": -0.4847453236579895,
"logps/chosen": -624.9798583984375,
"logps/rejected": -1600.3311767578125,
"loss": 0.1704,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07361427694559097,
"rewards/margins": 0.3176981210708618,
"rewards/rejected": -0.3913124203681946,
"step": 2850
},
{
"epoch": 0.76,
"learning_rate": 8.099524404308948e-07,
"logits/chosen": -1.463273525238037,
"logits/rejected": -0.44319063425064087,
"logps/chosen": -733.7213745117188,
"logps/rejected": -1909.9351806640625,
"loss": 0.1014,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1007709726691246,
"rewards/margins": 0.38669538497924805,
"rewards/rejected": -0.487466424703598,
"step": 2860
},
{
"epoch": 0.77,
"learning_rate": 7.928778328007918e-07,
"logits/chosen": -1.7150824069976807,
"logits/rejected": -0.10497765243053436,
"logps/chosen": -676.3504028320312,
"logps/rejected": -1732.474609375,
"loss": 0.1293,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07949225604534149,
"rewards/margins": 0.3401327431201935,
"rewards/rejected": -0.4196249842643738,
"step": 2870
},
{
"epoch": 0.77,
"learning_rate": 7.759511406608255e-07,
"logits/chosen": -1.4279053211212158,
"logits/rejected": -0.10710684955120087,
"logps/chosen": -716.6738891601562,
"logps/rejected": -1643.4683837890625,
"loss": 0.1605,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09047757089138031,
"rewards/margins": 0.2613658308982849,
"rewards/rejected": -0.35184338688850403,
"step": 2880
},
{
"epoch": 0.77,
"learning_rate": 7.591738306429769e-07,
"logits/chosen": -1.4157854318618774,
"logits/rejected": -0.362983763217926,
"logps/chosen": -813.336181640625,
"logps/rejected": -1738.5064697265625,
"loss": 0.1146,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08977197110652924,
"rewards/margins": 0.333584189414978,
"rewards/rejected": -0.42335623502731323,
"step": 2890
},
{
"epoch": 0.77,
"learning_rate": 7.425473564358457e-07,
"logits/chosen": -1.2446939945220947,
"logits/rejected": -0.24595150351524353,
"logps/chosen": -870.3673706054688,
"logps/rejected": -1808.233642578125,
"loss": 0.2744,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14074157178401947,
"rewards/margins": 0.33063870668411255,
"rewards/rejected": -0.4713803231716156,
"step": 2900
},
{
"epoch": 0.78,
"learning_rate": 7.260731586586983e-07,
"logits/chosen": -1.6176745891571045,
"logits/rejected": -0.5913767218589783,
"logps/chosen": -506.3935546875,
"logps/rejected": -1553.77099609375,
"loss": 0.1245,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04700837656855583,
"rewards/margins": 0.33918124437332153,
"rewards/rejected": -0.38618963956832886,
"step": 2910
},
{
"epoch": 0.78,
"learning_rate": 7.097526647366379e-07,
"logits/chosen": -1.7674744129180908,
"logits/rejected": -0.6992005705833435,
"logps/chosen": -651.4013061523438,
"logps/rejected": -1570.8125,
"loss": 0.1434,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.058572918176651,
"rewards/margins": 0.3027827739715576,
"rewards/rejected": -0.3613556921482086,
"step": 2920
},
{
"epoch": 0.78,
"learning_rate": 6.935872887769299e-07,
"logits/chosen": -1.6404222249984741,
"logits/rejected": -0.85943204164505,
"logps/chosen": -782.1417846679688,
"logps/rejected": -1594.085205078125,
"loss": 0.2041,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.0960140973329544,
"rewards/margins": 0.2894643545150757,
"rewards/rejected": -0.3854784369468689,
"step": 2930
},
{
"epoch": 0.78,
"learning_rate": 6.775784314464717e-07,
"logits/chosen": -1.4033766984939575,
"logits/rejected": -0.9920269846916199,
"logps/chosen": -629.6160888671875,
"logps/rejected": -1653.112548828125,
"loss": 0.1683,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.07426479458808899,
"rewards/margins": 0.29479536414146423,
"rewards/rejected": -0.36906009912490845,
"step": 2940
},
{
"epoch": 0.79,
"learning_rate": 6.617274798504286e-07,
"logits/chosen": -1.6468874216079712,
"logits/rejected": -0.37529969215393066,
"logps/chosen": -776.8147583007812,
"logps/rejected": -1830.4251708984375,
"loss": 0.1382,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07489114999771118,
"rewards/margins": 0.3607185184955597,
"rewards/rejected": -0.4356096684932709,
"step": 2950
},
{
"epoch": 0.79,
"learning_rate": 6.460358074120518e-07,
"logits/chosen": -1.5551801919937134,
"logits/rejected": -0.9461766481399536,
"logps/chosen": -714.2229614257812,
"logps/rejected": -1528.100341796875,
"loss": 0.1998,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06713708490133286,
"rewards/margins": 0.25901907682418823,
"rewards/rejected": -0.3261561691761017,
"step": 2960
},
{
"epoch": 0.79,
"learning_rate": 6.305047737536707e-07,
"logits/chosen": -1.5757367610931396,
"logits/rejected": -1.0922707319259644,
"logps/chosen": -647.8770751953125,
"logps/rejected": -1738.0703125,
"loss": 0.1411,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.06151123717427254,
"rewards/margins": 0.37969422340393066,
"rewards/rejected": -0.4412055015563965,
"step": 2970
},
{
"epoch": 0.79,
"learning_rate": 6.151357245788917e-07,
"logits/chosen": -1.52052903175354,
"logits/rejected": -0.5594236850738525,
"logps/chosen": -708.1317138671875,
"logps/rejected": -1907.6968994140625,
"loss": 0.1327,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08830620348453522,
"rewards/margins": 0.3933480679988861,
"rewards/rejected": -0.4816543161869049,
"step": 2980
},
{
"epoch": 0.8,
"learning_rate": 5.999299915559956e-07,
"logits/chosen": -1.422783374786377,
"logits/rejected": -0.7664706110954285,
"logps/chosen": -648.51171875,
"logps/rejected": -1761.1148681640625,
"loss": 0.1385,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07798387855291367,
"rewards/margins": 0.35578176379203796,
"rewards/rejected": -0.4337656497955322,
"step": 2990
},
{
"epoch": 0.8,
"learning_rate": 5.848888922025553e-07,
"logits/chosen": -1.4832850694656372,
"logits/rejected": -0.704194962978363,
"logps/chosen": -635.0929565429688,
"logps/rejected": -1439.96630859375,
"loss": 0.2039,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.09270008653402328,
"rewards/margins": 0.2593896985054016,
"rewards/rejected": -0.3520897328853607,
"step": 3000
},
{
"epoch": 0.8,
"learning_rate": 5.700137297712749e-07,
"logits/chosen": -1.560736894607544,
"logits/rejected": -0.25968214869499207,
"logps/chosen": -725.5360107421875,
"logps/rejected": -1755.325927734375,
"loss": 0.1363,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08439795672893524,
"rewards/margins": 0.36337295174598694,
"rewards/rejected": -0.4477709233760834,
"step": 3010
},
{
"epoch": 0.81,
"learning_rate": 5.553057931370729e-07,
"logits/chosen": -1.666855812072754,
"logits/rejected": -0.44792652130126953,
"logps/chosen": -635.80419921875,
"logps/rejected": -1666.586669921875,
"loss": 0.1712,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07980332523584366,
"rewards/margins": 0.3341625928878784,
"rewards/rejected": -0.41396594047546387,
"step": 3020
},
{
"epoch": 0.81,
"learning_rate": 5.407663566854008e-07,
"logits/chosen": -1.5879957675933838,
"logits/rejected": -0.9095015525817871,
"logps/chosen": -787.1989135742188,
"logps/rejected": -1740.4361572265625,
"loss": 0.1449,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08927465230226517,
"rewards/margins": 0.3183870315551758,
"rewards/rejected": -0.4076617360115051,
"step": 3030
},
{
"epoch": 0.81,
"learning_rate": 5.263966802018275e-07,
"logits/chosen": -1.6660913228988647,
"logits/rejected": -0.7353159785270691,
"logps/chosen": -805.4395751953125,
"logps/rejected": -1653.5638427734375,
"loss": 0.1664,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08214981853961945,
"rewards/margins": 0.31050822138786316,
"rewards/rejected": -0.3926580250263214,
"step": 3040
},
{
"epoch": 0.81,
"learning_rate": 5.121980087628802e-07,
"logits/chosen": -1.63559091091156,
"logits/rejected": -0.620925784111023,
"logps/chosen": -642.1688232421875,
"logps/rejected": -1510.009521484375,
"loss": 0.186,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.04546310380101204,
"rewards/margins": 0.30619844794273376,
"rewards/rejected": -0.3516615927219391,
"step": 3050
},
{
"epoch": 0.82,
"learning_rate": 4.981715726281666e-07,
"logits/chosen": -1.423666000366211,
"logits/rejected": -0.12835507094860077,
"logps/chosen": -631.965576171875,
"logps/rejected": -1562.4158935546875,
"loss": 0.2021,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06688202172517776,
"rewards/margins": 0.2849853038787842,
"rewards/rejected": -0.35186734795570374,
"step": 3060
},
{
"epoch": 0.82,
"learning_rate": 4.843185871337722e-07,
"logits/chosen": -1.3471550941467285,
"logits/rejected": -0.5594549179077148,
"logps/chosen": -620.7034301757812,
"logps/rejected": -1736.4898681640625,
"loss": 0.1361,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.05083579942584038,
"rewards/margins": 0.33074623346328735,
"rewards/rejected": -0.3815820515155792,
"step": 3070
},
{
"epoch": 0.82,
"learning_rate": 4.706402525869633e-07,
"logits/chosen": -1.2812891006469727,
"logits/rejected": -0.4547126293182373,
"logps/chosen": -562.1158447265625,
"logps/rejected": -1769.2896728515625,
"loss": 0.1329,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04002733901143074,
"rewards/margins": 0.383873850107193,
"rewards/rejected": -0.423901230096817,
"step": 3080
},
{
"epoch": 0.82,
"learning_rate": 4.5713775416217884e-07,
"logits/chosen": -1.571263313293457,
"logits/rejected": -0.40705451369285583,
"logps/chosen": -605.5236206054688,
"logps/rejected": -1484.4984130859375,
"loss": 0.1846,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.05240412801504135,
"rewards/margins": 0.2970747649669647,
"rewards/rejected": -0.34947890043258667,
"step": 3090
},
{
"epoch": 0.83,
"learning_rate": 4.438122617983442e-07,
"logits/chosen": -1.3903917074203491,
"logits/rejected": -0.20577654242515564,
"logps/chosen": -603.0784301757812,
"logps/rejected": -1667.8463134765625,
"loss": 0.1127,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.0214361734688282,
"rewards/margins": 0.37630894780158997,
"rewards/rejected": -0.3977451026439667,
"step": 3100
},
{
"epoch": 0.83,
"learning_rate": 4.3066493009749853e-07,
"logits/chosen": -1.6927878856658936,
"logits/rejected": -0.6562541723251343,
"logps/chosen": -784.9993896484375,
"logps/rejected": -1762.9515380859375,
"loss": 0.19,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07724924385547638,
"rewards/margins": 0.29403752088546753,
"rewards/rejected": -0.3712867796421051,
"step": 3110
},
{
"epoch": 0.83,
"learning_rate": 4.1769689822475147e-07,
"logits/chosen": -1.457176923751831,
"logits/rejected": -0.6649435758590698,
"logps/chosen": -731.3934326171875,
"logps/rejected": -1780.1842041015625,
"loss": 0.1694,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06437239050865173,
"rewards/margins": 0.33346226811408997,
"rewards/rejected": -0.3978345990180969,
"step": 3120
},
{
"epoch": 0.83,
"learning_rate": 4.049092898095816e-07,
"logits/chosen": -1.2814868688583374,
"logits/rejected": -0.6186938285827637,
"logps/chosen": -650.4574584960938,
"logps/rejected": -1637.374755859375,
"loss": 0.1541,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06206426024436951,
"rewards/margins": 0.26436498761177063,
"rewards/rejected": -0.32642924785614014,
"step": 3130
},
{
"epoch": 0.84,
"learning_rate": 3.9230321284847856e-07,
"logits/chosen": -1.5192755460739136,
"logits/rejected": -0.237101748585701,
"logps/chosen": -517.3258666992188,
"logps/rejected": -1354.155517578125,
"loss": 0.1789,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.014683857560157776,
"rewards/margins": 0.2717076539993286,
"rewards/rejected": -0.2863914966583252,
"step": 3140
},
{
"epoch": 0.84,
"learning_rate": 3.798797596089351e-07,
"logits/chosen": -1.514344573020935,
"logits/rejected": -1.0580737590789795,
"logps/chosen": -651.7825927734375,
"logps/rejected": -1709.284423828125,
"loss": 0.1427,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06703333556652069,
"rewards/margins": 0.3800427317619324,
"rewards/rejected": -0.4470759928226471,
"step": 3150
},
{
"epoch": 0.84,
"learning_rate": 3.6764000653481263e-07,
"logits/chosen": -1.5731451511383057,
"logits/rejected": -0.5786491632461548,
"logps/chosen": -680.1246337890625,
"logps/rejected": -1564.331787109375,
"loss": 0.1837,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08038660138845444,
"rewards/margins": 0.2984033226966858,
"rewards/rejected": -0.3787899315357208,
"step": 3160
},
{
"epoch": 0.85,
"learning_rate": 3.555850141530659e-07,
"logits/chosen": -1.620054006576538,
"logits/rejected": -0.6548138856887817,
"logps/chosen": -725.5187377929688,
"logps/rejected": -1559.7091064453125,
"loss": 0.1401,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.058062393218278885,
"rewards/margins": 0.3103691041469574,
"rewards/rejected": -0.3684315085411072,
"step": 3170
},
{
"epoch": 0.85,
"learning_rate": 3.4371582698185636e-07,
"logits/chosen": -1.4922726154327393,
"logits/rejected": -0.5721961259841919,
"logps/chosen": -574.6033935546875,
"logps/rejected": -1700.106689453125,
"loss": 0.1561,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.055240534245967865,
"rewards/margins": 0.36047983169555664,
"rewards/rejected": -0.4157203733921051,
"step": 3180
},
{
"epoch": 0.85,
"learning_rate": 3.3203347344004737e-07,
"logits/chosen": -1.260347604751587,
"logits/rejected": -0.1747826784849167,
"logps/chosen": -695.7683715820312,
"logps/rejected": -1630.91064453125,
"loss": 0.1731,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.13098278641700745,
"rewards/margins": 0.2832750380039215,
"rewards/rejected": -0.41425782442092896,
"step": 3190
},
{
"epoch": 0.85,
"learning_rate": 3.2053896575809426e-07,
"logits/chosen": -1.39348566532135,
"logits/rejected": -0.686271607875824,
"logps/chosen": -608.1319580078125,
"logps/rejected": -1700.7486572265625,
"loss": 0.0992,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.0731215551495552,
"rewards/margins": 0.36609500646591187,
"rewards/rejected": -0.4392165243625641,
"step": 3200
},
{
"epoch": 0.86,
"learning_rate": 3.092332998903416e-07,
"logits/chosen": -1.3719862699508667,
"logits/rejected": -0.6818580031394958,
"logps/chosen": -606.0661010742188,
"logps/rejected": -1553.185791015625,
"loss": 0.1497,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06806117296218872,
"rewards/margins": 0.30099108815193176,
"rewards/rejected": -0.3690522313117981,
"step": 3210
},
{
"epoch": 0.86,
"learning_rate": 2.981174554287239e-07,
"logits/chosen": -1.3921059370040894,
"logits/rejected": -0.7532171010971069,
"logps/chosen": -703.4324951171875,
"logps/rejected": -1927.8841552734375,
"loss": 0.1379,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07026232779026031,
"rewards/margins": 0.38420066237449646,
"rewards/rejected": -0.4544629156589508,
"step": 3220
},
{
"epoch": 0.86,
"learning_rate": 2.871923955178918e-07,
"logits/chosen": -1.2359908819198608,
"logits/rejected": -0.5525861978530884,
"logps/chosen": -692.30419921875,
"logps/rejected": -1788.237060546875,
"loss": 0.1208,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08661970496177673,
"rewards/margins": 0.37772631645202637,
"rewards/rejected": -0.4643460810184479,
"step": 3230
},
{
"epoch": 0.86,
"learning_rate": 2.764590667717562e-07,
"logits/chosen": -1.4797961711883545,
"logits/rejected": -0.3416746258735657,
"logps/chosen": -617.2152099609375,
"logps/rejected": -1625.399658203125,
"loss": 0.1826,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0840892642736435,
"rewards/margins": 0.2994067072868347,
"rewards/rejected": -0.3834959864616394,
"step": 3240
},
{
"epoch": 0.87,
"learning_rate": 2.6591839919146963e-07,
"logits/chosen": -1.503493070602417,
"logits/rejected": 0.3321295380592346,
"logps/chosen": -688.2984619140625,
"logps/rejected": -1707.145263671875,
"loss": 0.1299,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08260496705770493,
"rewards/margins": 0.3498184084892273,
"rewards/rejected": -0.43242329359054565,
"step": 3250
},
{
"epoch": 0.87,
"learning_rate": 2.555713060848433e-07,
"logits/chosen": -1.3645676374435425,
"logits/rejected": -0.04500112682580948,
"logps/chosen": -643.5718994140625,
"logps/rejected": -1641.6656494140625,
"loss": 0.1557,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0737595409154892,
"rewards/margins": 0.3242550790309906,
"rewards/rejected": -0.3980146050453186,
"step": 3260
},
{
"epoch": 0.87,
"learning_rate": 2.454186839872158e-07,
"logits/chosen": -1.6310592889785767,
"logits/rejected": -0.6514387130737305,
"logps/chosen": -804.0472412109375,
"logps/rejected": -1859.6383056640625,
"loss": 0.1372,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12043702602386475,
"rewards/margins": 0.3875194489955902,
"rewards/rejected": -0.5079564452171326,
"step": 3270
},
{
"epoch": 0.87,
"learning_rate": 2.3546141258376786e-07,
"logits/chosen": -1.3590515851974487,
"logits/rejected": -0.43198928236961365,
"logps/chosen": -733.1363525390625,
"logps/rejected": -1653.435791015625,
"loss": 0.1397,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11448194831609726,
"rewards/margins": 0.33477213978767395,
"rewards/rejected": -0.4492540955543518,
"step": 3280
},
{
"epoch": 0.88,
"learning_rate": 2.257003546333042e-07,
"logits/chosen": -1.4484094381332397,
"logits/rejected": -0.5377854108810425,
"logps/chosen": -742.8893432617188,
"logps/rejected": -1920.5609130859375,
"loss": 0.1697,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11689500510692596,
"rewards/margins": 0.35355696082115173,
"rewards/rejected": -0.4704520106315613,
"step": 3290
},
{
"epoch": 0.88,
"learning_rate": 2.1613635589349756e-07,
"logits/chosen": -1.6482737064361572,
"logits/rejected": -0.8282219767570496,
"logps/chosen": -565.1773681640625,
"logps/rejected": -1599.753662109375,
"loss": 0.1429,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.056347571313381195,
"rewards/margins": 0.3239589333534241,
"rewards/rejected": -0.38030651211738586,
"step": 3300
},
{
"epoch": 0.88,
"learning_rate": 2.0677024504760752e-07,
"logits/chosen": -1.502986192703247,
"logits/rejected": -0.47991952300071716,
"logps/chosen": -559.9749755859375,
"logps/rejected": -1734.3677978515625,
"loss": 0.1338,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.041349343955516815,
"rewards/margins": 0.3722584545612335,
"rewards/rejected": -0.41360777616500854,
"step": 3310
},
{
"epoch": 0.89,
"learning_rate": 1.9760283363267684e-07,
"logits/chosen": -1.6863027811050415,
"logits/rejected": -0.2434253990650177,
"logps/chosen": -727.5802612304688,
"logps/rejected": -1704.158203125,
"loss": 0.1468,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07457654923200607,
"rewards/margins": 0.3438524007797241,
"rewards/rejected": -0.4184289872646332,
"step": 3320
},
{
"epoch": 0.89,
"learning_rate": 1.8863491596921745e-07,
"logits/chosen": -1.3177156448364258,
"logits/rejected": -0.1511184722185135,
"logps/chosen": -603.4699096679688,
"logps/rejected": -1553.7645263671875,
"loss": 0.114,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.07171906530857086,
"rewards/margins": 0.3328208923339844,
"rewards/rejected": -0.40453997254371643,
"step": 3330
},
{
"epoch": 0.89,
"learning_rate": 1.798672690923828e-07,
"logits/chosen": -1.5630302429199219,
"logits/rejected": -0.8224193453788757,
"logps/chosen": -620.0535888671875,
"logps/rejected": -1780.6702880859375,
"loss": 0.0976,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06969192624092102,
"rewards/margins": 0.3709218502044678,
"rewards/rejected": -0.4406138062477112,
"step": 3340
},
{
"epoch": 0.89,
"learning_rate": 1.713006526846439e-07,
"logits/chosen": -1.5444920063018799,
"logits/rejected": -0.24943670630455017,
"logps/chosen": -683.787841796875,
"logps/rejected": -1977.3343505859375,
"loss": 0.0802,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.06767871230840683,
"rewards/margins": 0.4605562686920166,
"rewards/rejected": -0.5282350778579712,
"step": 3350
},
{
"epoch": 0.9,
"learning_rate": 1.629358090099639e-07,
"logits/chosen": -1.5846331119537354,
"logits/rejected": -0.536185085773468,
"logps/chosen": -703.1005859375,
"logps/rejected": -1691.5517578125,
"loss": 0.1945,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07973551005125046,
"rewards/margins": 0.3327817916870117,
"rewards/rejected": -0.4125173091888428,
"step": 3360
},
{
"epoch": 0.9,
"learning_rate": 1.5477346284948292e-07,
"logits/chosen": -1.572539210319519,
"logits/rejected": -0.8882861137390137,
"logps/chosen": -748.1613159179688,
"logps/rejected": -1634.09521484375,
"loss": 0.1691,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09525839984416962,
"rewards/margins": 0.28481072187423706,
"rewards/rejected": -0.38006913661956787,
"step": 3370
},
{
"epoch": 0.9,
"learning_rate": 1.4681432143872133e-07,
"logits/chosen": -1.2806552648544312,
"logits/rejected": -0.47933030128479004,
"logps/chosen": -767.8941650390625,
"logps/rejected": -1739.5386962890625,
"loss": 0.1768,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09690088033676147,
"rewards/margins": 0.33204126358032227,
"rewards/rejected": -0.42894211411476135,
"step": 3380
},
{
"epoch": 0.9,
"learning_rate": 1.3905907440629752e-07,
"logits/chosen": -1.619302749633789,
"logits/rejected": -0.4543309211730957,
"logps/chosen": -705.6532592773438,
"logps/rejected": -1661.1187744140625,
"loss": 0.2429,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.09244809299707413,
"rewards/margins": 0.27722662687301636,
"rewards/rejected": -0.3696747124195099,
"step": 3390
},
{
"epoch": 0.91,
"learning_rate": 1.31508393714177e-07,
"logits/chosen": -1.576765537261963,
"logits/rejected": -0.9673765301704407,
"logps/chosen": -700.4979248046875,
"logps/rejected": -1635.6195068359375,
"loss": 0.1433,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08088205009698868,
"rewards/margins": 0.33026668429374695,
"rewards/rejected": -0.41114872694015503,
"step": 3400
},
{
"epoch": 0.91,
"learning_rate": 1.241629335994471e-07,
"logits/chosen": -1.3158491849899292,
"logits/rejected": -0.2821682393550873,
"logps/chosen": -567.6187133789062,
"logps/rejected": -1448.58837890625,
"loss": 0.2002,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.06683605909347534,
"rewards/margins": 0.2942168712615967,
"rewards/rejected": -0.361052930355072,
"step": 3410
},
{
"epoch": 0.91,
"learning_rate": 1.1702333051763271e-07,
"logits/chosen": -1.630567193031311,
"logits/rejected": -0.7469101548194885,
"logps/chosen": -718.3992919921875,
"logps/rejected": -1641.1156005859375,
"loss": 0.1399,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10309512913227081,
"rewards/margins": 0.3157398998737335,
"rewards/rejected": -0.4188350737094879,
"step": 3420
},
{
"epoch": 0.91,
"learning_rate": 1.1009020308754587e-07,
"logits/chosen": -1.4223135709762573,
"logits/rejected": -0.4289831519126892,
"logps/chosen": -924.49609375,
"logps/rejected": -1811.7252197265625,
"loss": 0.2089,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13550476729869843,
"rewards/margins": 0.3179660737514496,
"rewards/rejected": -0.4534708857536316,
"step": 3430
},
{
"epoch": 0.92,
"learning_rate": 1.0336415203768962e-07,
"logits/chosen": -1.4638688564300537,
"logits/rejected": 0.3213498592376709,
"logps/chosen": -778.1163330078125,
"logps/rejected": -1671.373779296875,
"loss": 0.1553,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07180814445018768,
"rewards/margins": 0.3324059545993805,
"rewards/rejected": -0.404214084148407,
"step": 3440
},
{
"epoch": 0.92,
"learning_rate": 9.684576015420277e-08,
"logits/chosen": -1.3937320709228516,
"logits/rejected": -0.29455748200416565,
"logps/chosen": -558.4961547851562,
"logps/rejected": -1537.4754638671875,
"loss": 0.1932,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07989932596683502,
"rewards/margins": 0.2945837080478668,
"rewards/rejected": -0.37448304891586304,
"step": 3450
},
{
"epoch": 0.92,
"learning_rate": 9.053559223036746e-08,
"logits/chosen": -1.3823366165161133,
"logits/rejected": -0.34489625692367554,
"logps/chosen": -750.284423828125,
"logps/rejected": -1547.081298828125,
"loss": 0.1481,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07611394673585892,
"rewards/margins": 0.2931041121482849,
"rewards/rejected": -0.36921799182891846,
"step": 3460
},
{
"epoch": 0.93,
"learning_rate": 8.44341950176683e-08,
"logits/chosen": -1.58815598487854,
"logits/rejected": -0.0034358978737145662,
"logps/chosen": -567.5117797851562,
"logps/rejected": -1468.9776611328125,
"loss": 0.1411,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.03928015008568764,
"rewards/margins": 0.3018725514411926,
"rewards/rejected": -0.34115269780158997,
"step": 3470
},
{
"epoch": 0.93,
"learning_rate": 7.854209717842231e-08,
"logits/chosen": -1.5472205877304077,
"logits/rejected": -0.9048103094100952,
"logps/chosen": -627.4727783203125,
"logps/rejected": -1713.844970703125,
"loss": 0.0771,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.04558110982179642,
"rewards/margins": 0.3975328207015991,
"rewards/rejected": -0.44311389327049255,
"step": 3480
},
{
"epoch": 0.93,
"learning_rate": 7.285980923996989e-08,
"logits/chosen": -1.4914085865020752,
"logits/rejected": -0.47641056776046753,
"logps/chosen": -650.9934692382812,
"logps/rejected": -1637.280029296875,
"loss": 0.1166,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.054985035210847855,
"rewards/margins": 0.3437821865081787,
"rewards/rejected": -0.3987672030925751,
"step": 3490
},
{
"epoch": 0.93,
"learning_rate": 6.738782355044048e-08,
"logits/chosen": -1.564391016960144,
"logits/rejected": -0.30673253536224365,
"logps/chosen": -597.55322265625,
"logps/rejected": -1578.419677734375,
"loss": 0.1353,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04587271437048912,
"rewards/margins": 0.32063284516334534,
"rewards/rejected": -0.36650553345680237,
"step": 3500
},
{
"epoch": 0.94,
"learning_rate": 6.212661423609184e-08,
"logits/chosen": -1.298776626586914,
"logits/rejected": -0.1061747819185257,
"logps/chosen": -640.8598022460938,
"logps/rejected": -1533.592529296875,
"loss": 0.1815,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.05050724744796753,
"rewards/margins": 0.27905362844467163,
"rewards/rejected": -0.32956087589263916,
"step": 3510
},
{
"epoch": 0.94,
"learning_rate": 5.707663716023021e-08,
"logits/chosen": -1.5522761344909668,
"logits/rejected": -0.26008373498916626,
"logps/chosen": -703.5093994140625,
"logps/rejected": -1482.5135498046875,
"loss": 0.1582,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.06401093304157257,
"rewards/margins": 0.2837941348552704,
"rewards/rejected": -0.34780508279800415,
"step": 3520
},
{
"epoch": 0.94,
"learning_rate": 5.22383298837098e-08,
"logits/chosen": -1.5810019969940186,
"logits/rejected": 0.17494335770606995,
"logps/chosen": -811.2449340820312,
"logps/rejected": -1590.089599609375,
"loss": 0.0992,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09410594403743744,
"rewards/margins": 0.29233235120773315,
"rewards/rejected": -0.3864383101463318,
"step": 3530
},
{
"epoch": 0.94,
"learning_rate": 4.761211162702117e-08,
"logits/chosen": -1.484390139579773,
"logits/rejected": -0.5827163457870483,
"logps/chosen": -721.3407592773438,
"logps/rejected": -1430.334716796875,
"loss": 0.1909,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10846199095249176,
"rewards/margins": 0.22209401428699493,
"rewards/rejected": -0.3305560052394867,
"step": 3540
},
{
"epoch": 0.95,
"learning_rate": 4.319838323396691e-08,
"logits/chosen": -1.5092474222183228,
"logits/rejected": -0.8641079068183899,
"logps/chosen": -654.3946533203125,
"logps/rejected": -1624.559326171875,
"loss": 0.1749,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07206736505031586,
"rewards/margins": 0.3177274167537689,
"rewards/rejected": -0.3897947669029236,
"step": 3550
},
{
"epoch": 0.95,
"learning_rate": 3.8997527136930004e-08,
"logits/chosen": -1.3939855098724365,
"logits/rejected": -0.9460384249687195,
"logps/chosen": -721.7215576171875,
"logps/rejected": -1758.743408203125,
"loss": 0.1378,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1156727522611618,
"rewards/margins": 0.3109985291957855,
"rewards/rejected": -0.4266713261604309,
"step": 3560
},
{
"epoch": 0.95,
"learning_rate": 3.5009907323737826e-08,
"logits/chosen": -1.390904188156128,
"logits/rejected": -0.4405292570590973,
"logps/chosen": -571.7150268554688,
"logps/rejected": -1526.440673828125,
"loss": 0.172,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.04342179745435715,
"rewards/margins": 0.2945147156715393,
"rewards/rejected": -0.33793652057647705,
"step": 3570
},
{
"epoch": 0.95,
"learning_rate": 3.1235869306123766e-08,
"logits/chosen": -1.743300199508667,
"logits/rejected": -0.7482790350914001,
"logps/chosen": -623.4041748046875,
"logps/rejected": -1765.6488037109375,
"loss": 0.1102,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.057156067341566086,
"rewards/margins": 0.3713191747665405,
"rewards/rejected": -0.4284752309322357,
"step": 3580
},
{
"epoch": 0.96,
"learning_rate": 2.767574008979007e-08,
"logits/chosen": -1.62771475315094,
"logits/rejected": -0.4970121383666992,
"logps/chosen": -712.7774047851562,
"logps/rejected": -1704.322021484375,
"loss": 0.1109,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07325717061758041,
"rewards/margins": 0.3189954161643982,
"rewards/rejected": -0.3922525942325592,
"step": 3590
},
{
"epoch": 0.96,
"learning_rate": 2.4329828146074096e-08,
"logits/chosen": -1.5333154201507568,
"logits/rejected": -0.363052636384964,
"logps/chosen": -755.8592529296875,
"logps/rejected": -1792.1357421875,
"loss": 0.1142,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06991975009441376,
"rewards/margins": 0.3443168103694916,
"rewards/rejected": -0.41423654556274414,
"step": 3600
},
{
"epoch": 0.96,
"learning_rate": 2.1198423385220822e-08,
"logits/chosen": -1.5201735496520996,
"logits/rejected": -0.7737599611282349,
"logps/chosen": -648.6341552734375,
"logps/rejected": -1637.8931884765625,
"loss": 0.1757,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.046808890998363495,
"rewards/margins": 0.346151202917099,
"rewards/rejected": -0.3929601311683655,
"step": 3610
},
{
"epoch": 0.97,
"learning_rate": 1.82817971312621e-08,
"logits/chosen": -1.6378930807113647,
"logits/rejected": -0.5780726671218872,
"logps/chosen": -671.5693969726562,
"logps/rejected": -1713.7747802734375,
"loss": 0.1688,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07191529124975204,
"rewards/margins": 0.33269262313842773,
"rewards/rejected": -0.4046078622341156,
"step": 3620
},
{
"epoch": 0.97,
"learning_rate": 1.5580202098509078e-08,
"logits/chosen": -1.6032174825668335,
"logits/rejected": -0.24984005093574524,
"logps/chosen": -638.01123046875,
"logps/rejected": -1515.741943359375,
"loss": 0.1446,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06576866656541824,
"rewards/margins": 0.31248709559440613,
"rewards/rejected": -0.37825578451156616,
"step": 3630
},
{
"epoch": 0.97,
"learning_rate": 1.3093872369654148e-08,
"logits/chosen": -1.462215781211853,
"logits/rejected": 0.025119613856077194,
"logps/chosen": -680.1953125,
"logps/rejected": -1622.0228271484375,
"loss": 0.2123,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08275710791349411,
"rewards/margins": 0.2952966094017029,
"rewards/rejected": -0.3780537247657776,
"step": 3640
},
{
"epoch": 0.97,
"learning_rate": 1.0823023375489128e-08,
"logits/chosen": -1.6868362426757812,
"logits/rejected": -0.5641859173774719,
"logps/chosen": -618.9349975585938,
"logps/rejected": -1718.207275390625,
"loss": 0.107,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.05680794641375542,
"rewards/margins": 0.3729218542575836,
"rewards/rejected": -0.42972975969314575,
"step": 3650
},
{
"epoch": 0.98,
"learning_rate": 8.767851876239075e-09,
"logits/chosen": -1.2520034313201904,
"logits/rejected": -0.46264973282814026,
"logps/chosen": -756.2525024414062,
"logps/rejected": -1568.090576171875,
"loss": 0.1772,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10075948387384415,
"rewards/margins": 0.28606051206588745,
"rewards/rejected": -0.3868200182914734,
"step": 3660
},
{
"epoch": 0.98,
"learning_rate": 6.9285359445145366e-09,
"logits/chosen": -1.6095876693725586,
"logits/rejected": -0.49465712904930115,
"logps/chosen": -636.206298828125,
"logps/rejected": -1638.2972412109375,
"loss": 0.1288,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.03626566380262375,
"rewards/margins": 0.34111684560775757,
"rewards/rejected": -0.3773825466632843,
"step": 3670
},
{
"epoch": 0.98,
"learning_rate": 5.305234949880001e-09,
"logits/chosen": -1.4576680660247803,
"logits/rejected": -0.8192172050476074,
"logps/chosen": -602.4050903320312,
"logps/rejected": -1541.21142578125,
"loss": 0.2052,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.056397415697574615,
"rewards/margins": 0.27618226408958435,
"rewards/rejected": -0.33257967233657837,
"step": 3680
},
{
"epoch": 0.98,
"learning_rate": 3.8980895450474455e-09,
"logits/chosen": -1.4624910354614258,
"logits/rejected": -0.06623925268650055,
"logps/chosen": -615.46142578125,
"logps/rejected": -1512.7928466796875,
"loss": 0.1562,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.050837211310863495,
"rewards/margins": 0.3221223056316376,
"rewards/rejected": -0.37295952439308167,
"step": 3690
},
{
"epoch": 0.99,
"learning_rate": 2.7072216536885855e-09,
"logits/chosen": -1.576015591621399,
"logits/rejected": -0.6331445574760437,
"logps/chosen": -652.0988159179688,
"logps/rejected": -1652.5101318359375,
"loss": 0.0999,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0720575600862503,
"rewards/margins": 0.32736578583717346,
"rewards/rejected": -0.3994233012199402,
"step": 3700
},
{
"epoch": 0.99,
"learning_rate": 1.7327344598702667e-09,
"logits/chosen": -1.4119715690612793,
"logits/rejected": -0.8666298985481262,
"logps/chosen": -634.987548828125,
"logps/rejected": -1896.078125,
"loss": 0.0889,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.053931184113025665,
"rewards/margins": 0.42121997475624084,
"rewards/rejected": -0.4751511216163635,
"step": 3710
},
{
"epoch": 0.99,
"learning_rate": 9.747123991141193e-10,
"logits/chosen": -1.3923088312149048,
"logits/rejected": -0.4560534358024597,
"logps/chosen": -703.5958862304688,
"logps/rejected": -1692.0693359375,
"loss": 0.1027,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07977847754955292,
"rewards/margins": 0.34622496366500854,
"rewards/rejected": -0.42600345611572266,
"step": 3720
},
{
"epoch": 0.99,
"learning_rate": 4.332211510807427e-10,
"logits/chosen": -1.3781083822250366,
"logits/rejected": -0.20015454292297363,
"logps/chosen": -547.2339477539062,
"logps/rejected": -1573.963623046875,
"loss": 0.1506,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0455051027238369,
"rewards/margins": 0.3225332200527191,
"rewards/rejected": -0.3680383563041687,
"step": 3730
},
{
"epoch": 1.0,
"learning_rate": 1.0830763387897902e-10,
"logits/chosen": -1.4161818027496338,
"logits/rejected": 0.6542869806289673,
"logps/chosen": -718.9110107421875,
"logps/rejected": -1792.678955078125,
"loss": 0.0937,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.0775211825966835,
"rewards/margins": 0.3633776307106018,
"rewards/rejected": -0.4408988058567047,
"step": 3740
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"logits/chosen": -1.728247046470642,
"logits/rejected": -1.0278023481369019,
"logps/chosen": -706.55908203125,
"logps/rejected": -1668.5732421875,
"loss": 0.1603,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07834981381893158,
"rewards/margins": 0.3416889011859894,
"rewards/rejected": -0.42003870010375977,
"step": 3750
},
{
"epoch": 1.0,
"step": 3750,
"total_flos": 0.0,
"train_loss": 0.1829929338614146,
"train_runtime": 17952.2095,
"train_samples_per_second": 0.836,
"train_steps_per_second": 0.209
}
],
"logging_steps": 10,
"max_steps": 3750,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 0.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}