| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.9705793145283591, |
| "eval_steps": 500, |
| "global_step": 200, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.004852896572641796, |
| "grad_norm": 3.7387311458587646, |
| "learning_rate": 6.451612903225806e-06, |
| "logits/chosen": -0.5510790944099426, |
| "logits/rejected": -0.5881398916244507, |
| "logps/chosen": -12.010782241821289, |
| "logps/rejected": -14.501535415649414, |
| "loss": 1.3914958238601685, |
| "memory(GiB)": 30.65, |
| "nll_loss": 0.6983487010002136, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.010451 |
| }, |
| { |
| "epoch": 0.009705793145283591, |
| "grad_norm": 5.074642658233643, |
| "learning_rate": 1.2903225806451613e-05, |
| "logits/chosen": -0.523028552532196, |
| "logits/rejected": -0.48799216747283936, |
| "logps/chosen": -15.598363876342773, |
| "logps/rejected": -12.491965293884277, |
| "loss": 1.3891178369522095, |
| "memory(GiB)": 33.32, |
| "nll_loss": 0.6959705948829651, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.010818 |
| }, |
| { |
| "epoch": 0.014558689717925387, |
| "grad_norm": 4.8900933265686035, |
| "learning_rate": 1.935483870967742e-05, |
| "logits/chosen": -0.6194146871566772, |
| "logits/rejected": -0.6501045227050781, |
| "logps/chosen": -4.106720924377441, |
| "logps/rejected": -9.803337097167969, |
| "loss": 1.0989456176757812, |
| "memory(GiB)": 33.32, |
| "nll_loss": 0.4074462950229645, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.0007492969743907452, |
| "rewards/margins": 0.003412690246477723, |
| "rewards/rejected": -0.0026633935049176216, |
| "step": 3, |
| "train_speed(iter/s)": 0.010979 |
| }, |
| { |
| "epoch": 0.019411586290567183, |
| "grad_norm": 4.551450729370117, |
| "learning_rate": 2.5806451612903226e-05, |
| "logits/chosen": -0.4892222285270691, |
| "logits/rejected": -0.6288261413574219, |
| "logps/chosen": -11.312190055847168, |
| "logps/rejected": -19.471235275268555, |
| "loss": 1.4801559448242188, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.788896381855011, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.009614704176783562, |
| "rewards/margins": 0.003918787930160761, |
| "rewards/rejected": 0.00569591810926795, |
| "step": 4, |
| "train_speed(iter/s)": 0.011017 |
| }, |
| { |
| "epoch": 0.024264482863208977, |
| "grad_norm": 5.593626499176025, |
| "learning_rate": 3.2258064516129034e-05, |
| "logits/chosen": -0.5494186282157898, |
| "logits/rejected": -0.6138424277305603, |
| "logps/chosen": -9.588997840881348, |
| "logps/rejected": -13.734201431274414, |
| "loss": 1.360138177871704, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.6719292998313904, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.006837800145149231, |
| "rewards/margins": 0.01003339048475027, |
| "rewards/rejected": -0.003195591736584902, |
| "step": 5, |
| "train_speed(iter/s)": 0.011063 |
| }, |
| { |
| "epoch": 0.029117379435850774, |
| "grad_norm": 4.7115631103515625, |
| "learning_rate": 3.870967741935484e-05, |
| "logits/chosen": -0.5902564525604248, |
| "logits/rejected": -0.6040255427360535, |
| "logps/chosen": -10.28292465209961, |
| "logps/rejected": -10.143241882324219, |
| "loss": 1.4910708665847778, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.7961816787719727, |
| "rewards/accuracies": 0.5, |
| "rewards/chosen": 0.010182744823396206, |
| "rewards/margins": -0.0029990212060511112, |
| "rewards/rejected": 0.013181766495108604, |
| "step": 6, |
| "train_speed(iter/s)": 0.011101 |
| }, |
| { |
| "epoch": 0.03397027600849257, |
| "grad_norm": 3.242079257965088, |
| "learning_rate": 4.516129032258064e-05, |
| "logits/chosen": -0.5178452730178833, |
| "logits/rejected": -0.5171934366226196, |
| "logps/chosen": -11.35531997680664, |
| "logps/rejected": -13.940824508666992, |
| "loss": 1.256138563156128, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.5596349239349365, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": 0.027008483186364174, |
| "rewards/margins": -0.005357709713280201, |
| "rewards/rejected": 0.03236619383096695, |
| "step": 7, |
| "train_speed(iter/s)": 0.01111 |
| }, |
| { |
| "epoch": 0.038823172581134366, |
| "grad_norm": 4.068852424621582, |
| "learning_rate": 5.161290322580645e-05, |
| "logits/chosen": -0.5522034168243408, |
| "logits/rejected": -0.5255574584007263, |
| "logps/chosen": -12.265619277954102, |
| "logps/rejected": -11.32555866241455, |
| "loss": 1.3358875513076782, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.6352814435958862, |
| "rewards/accuracies": 0.40625, |
| "rewards/chosen": 0.021265771239995956, |
| "rewards/margins": -0.011079448275268078, |
| "rewards/rejected": 0.03234522044658661, |
| "step": 8, |
| "train_speed(iter/s)": 0.011128 |
| }, |
| { |
| "epoch": 0.04367606915377616, |
| "grad_norm": 3.7364556789398193, |
| "learning_rate": 5.8064516129032266e-05, |
| "logits/chosen": -0.5157386064529419, |
| "logits/rejected": -0.5693979263305664, |
| "logps/chosen": -8.727921485900879, |
| "logps/rejected": -15.667279243469238, |
| "loss": 1.2342674732208252, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.565812349319458, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.06674117594957352, |
| "rewards/margins": 0.05803339183330536, |
| "rewards/rejected": 0.008707791566848755, |
| "step": 9, |
| "train_speed(iter/s)": 0.01114 |
| }, |
| { |
| "epoch": 0.048528965726417954, |
| "grad_norm": 5.376058578491211, |
| "learning_rate": 6.451612903225807e-05, |
| "logits/chosen": -0.5336275696754456, |
| "logits/rejected": -0.5664803385734558, |
| "logps/chosen": -10.200874328613281, |
| "logps/rejected": -14.752674102783203, |
| "loss": 1.3360637426376343, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.6595087647438049, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.05797234922647476, |
| "rewards/margins": 0.05362413078546524, |
| "rewards/rejected": 0.00434822216629982, |
| "step": 10, |
| "train_speed(iter/s)": 0.011149 |
| }, |
| { |
| "epoch": 0.05338186229905975, |
| "grad_norm": 3.41105318069458, |
| "learning_rate": 7.096774193548388e-05, |
| "logits/chosen": -0.5836127996444702, |
| "logits/rejected": -0.5781168341636658, |
| "logps/chosen": -16.744802474975586, |
| "logps/rejected": -10.854264259338379, |
| "loss": 1.4587984085083008, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.8293395042419434, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.10905362665653229, |
| "rewards/margins": 0.15107941627502441, |
| "rewards/rejected": -0.04202580079436302, |
| "step": 11, |
| "train_speed(iter/s)": 0.011156 |
| }, |
| { |
| "epoch": 0.05823475887170155, |
| "grad_norm": 3.8034191131591797, |
| "learning_rate": 7.741935483870968e-05, |
| "logits/chosen": -0.5040208101272583, |
| "logits/rejected": -0.5330135822296143, |
| "logps/chosen": -9.613079071044922, |
| "logps/rejected": -11.064685821533203, |
| "loss": 1.267149806022644, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.6177489161491394, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.02217767760157585, |
| "rewards/margins": 0.10778125375509262, |
| "rewards/rejected": -0.08560357987880707, |
| "step": 12, |
| "train_speed(iter/s)": 0.011164 |
| }, |
| { |
| "epoch": 0.06308765544434335, |
| "grad_norm": 3.453901767730713, |
| "learning_rate": 8.387096774193549e-05, |
| "logits/chosen": -0.5467038154602051, |
| "logits/rejected": -0.577677845954895, |
| "logps/chosen": -8.353002548217773, |
| "logps/rejected": -13.458589553833008, |
| "loss": 1.233524203300476, |
| "memory(GiB)": 36.76, |
| "nll_loss": 0.5735047459602356, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.0792345255613327, |
| "rewards/margins": 0.10825661569833755, |
| "rewards/rejected": -0.02902209758758545, |
| "step": 13, |
| "train_speed(iter/s)": 0.011171 |
| }, |
| { |
| "epoch": 0.06794055201698514, |
| "grad_norm": 3.1424827575683594, |
| "learning_rate": 9.032258064516129e-05, |
| "logits/chosen": -0.5190276503562927, |
| "logits/rejected": -0.5021806359291077, |
| "logps/chosen": -11.147318840026855, |
| "logps/rejected": -11.436860084533691, |
| "loss": 1.254595398902893, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5940842628479004, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": 0.12143825739622116, |
| "rewards/margins": 0.11505899578332901, |
| "rewards/rejected": 0.006379259284585714, |
| "step": 14, |
| "train_speed(iter/s)": 0.011173 |
| }, |
| { |
| "epoch": 0.07279344858962693, |
| "grad_norm": 3.430884599685669, |
| "learning_rate": 9.677419354838711e-05, |
| "logits/chosen": -0.4556944668292999, |
| "logits/rejected": -0.4863564968109131, |
| "logps/chosen": -13.696606636047363, |
| "logps/rejected": -16.93030548095703, |
| "loss": 1.4541280269622803, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7789333462715149, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.06093275174498558, |
| "rewards/margins": 0.06727998703718185, |
| "rewards/rejected": -0.006347230635583401, |
| "step": 15, |
| "train_speed(iter/s)": 0.011173 |
| }, |
| { |
| "epoch": 0.07764634516226873, |
| "grad_norm": 5.697832107543945, |
| "learning_rate": 0.0001032258064516129, |
| "logits/chosen": -0.5404947400093079, |
| "logits/rejected": -0.5623520016670227, |
| "logps/chosen": -8.202204704284668, |
| "logps/rejected": -12.407597541809082, |
| "loss": 1.244711995124817, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.601537823677063, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.14254164695739746, |
| "rewards/margins": 0.16008448600769043, |
| "rewards/rejected": -0.017542842775583267, |
| "step": 16, |
| "train_speed(iter/s)": 0.011181 |
| }, |
| { |
| "epoch": 0.08249924173491052, |
| "grad_norm": 3.0015194416046143, |
| "learning_rate": 0.00010967741935483871, |
| "logits/chosen": -0.5660006999969482, |
| "logits/rejected": -0.5559377074241638, |
| "logps/chosen": -10.180459022521973, |
| "logps/rejected": -10.599821090698242, |
| "loss": 1.2977104187011719, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6220536231994629, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.12847113609313965, |
| "rewards/margins": 0.0701126754283905, |
| "rewards/rejected": 0.058358460664749146, |
| "step": 17, |
| "train_speed(iter/s)": 0.011186 |
| }, |
| { |
| "epoch": 0.08735213830755233, |
| "grad_norm": 5.351679801940918, |
| "learning_rate": 0.00011612903225806453, |
| "logits/chosen": -0.4931953251361847, |
| "logits/rejected": -0.5151236653327942, |
| "logps/chosen": -9.91970157623291, |
| "logps/rejected": -17.587669372558594, |
| "loss": 1.2324532270431519, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6188042163848877, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.10037277638912201, |
| "rewards/margins": 0.1970798224210739, |
| "rewards/rejected": -0.0967070460319519, |
| "step": 18, |
| "train_speed(iter/s)": 0.011184 |
| }, |
| { |
| "epoch": 0.09220503488019412, |
| "grad_norm": 2.8047103881835938, |
| "learning_rate": 0.00012258064516129034, |
| "logits/chosen": -0.5366685390472412, |
| "logits/rejected": -0.5188369154930115, |
| "logps/chosen": -11.985616683959961, |
| "logps/rejected": -9.721867561340332, |
| "loss": 1.243293285369873, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5873799920082092, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.1405532956123352, |
| "rewards/margins": 0.13308030366897583, |
| "rewards/rejected": 0.007473001256585121, |
| "step": 19, |
| "train_speed(iter/s)": 0.011187 |
| }, |
| { |
| "epoch": 0.09705793145283591, |
| "grad_norm": 3.8547439575195312, |
| "learning_rate": 0.00012903225806451613, |
| "logits/chosen": -0.4702628254890442, |
| "logits/rejected": -0.45280686020851135, |
| "logps/chosen": -11.422282218933105, |
| "logps/rejected": -13.049072265625, |
| "loss": 1.1890830993652344, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.49578943848609924, |
| "rewards/accuracies": 0.5, |
| "rewards/chosen": 0.15254227817058563, |
| "rewards/margins": 0.06372587382793427, |
| "rewards/rejected": 0.08881638944149017, |
| "step": 20, |
| "train_speed(iter/s)": 0.011186 |
| }, |
| { |
| "epoch": 0.10191082802547771, |
| "grad_norm": 3.9698216915130615, |
| "learning_rate": 0.00013548387096774193, |
| "logits/chosen": -0.46789583563804626, |
| "logits/rejected": -0.511239230632782, |
| "logps/chosen": -6.999385833740234, |
| "logps/rejected": -15.381391525268555, |
| "loss": 1.1891487836837769, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4858560860157013, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.0561373308300972, |
| "rewards/margins": 0.021670352667570114, |
| "rewards/rejected": 0.034466978162527084, |
| "step": 21, |
| "train_speed(iter/s)": 0.01119 |
| }, |
| { |
| "epoch": 0.1067637245981195, |
| "grad_norm": 2.810508966445923, |
| "learning_rate": 0.00014193548387096775, |
| "logits/chosen": -0.43238648772239685, |
| "logits/rejected": -0.5333250164985657, |
| "logps/chosen": -10.029150009155273, |
| "logps/rejected": -19.058868408203125, |
| "loss": 1.1820727586746216, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5326224565505981, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.09418301284313202, |
| "rewards/margins": 0.12642550468444824, |
| "rewards/rejected": -0.03224247321486473, |
| "step": 22, |
| "train_speed(iter/s)": 0.011191 |
| }, |
| { |
| "epoch": 0.1116166211707613, |
| "grad_norm": 5.709966659545898, |
| "learning_rate": 0.00014838709677419355, |
| "logits/chosen": -0.42264610528945923, |
| "logits/rejected": -0.45910894870758057, |
| "logps/chosen": -16.804805755615234, |
| "logps/rejected": -18.372587203979492, |
| "loss": 1.3850518465042114, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8230935335159302, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.05581340938806534, |
| "rewards/margins": 0.3414239287376404, |
| "rewards/rejected": -0.285610556602478, |
| "step": 23, |
| "train_speed(iter/s)": 0.01119 |
| }, |
| { |
| "epoch": 0.1164695177434031, |
| "grad_norm": 8.213228225708008, |
| "learning_rate": 0.00015483870967741937, |
| "logits/chosen": -0.43471360206604004, |
| "logits/rejected": -0.4376092553138733, |
| "logps/chosen": -10.293994903564453, |
| "logps/rejected": -12.654733657836914, |
| "loss": 1.2845356464385986, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6288778781890869, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.019672194495797157, |
| "rewards/margins": 0.11052073538303375, |
| "rewards/rejected": -0.09084855020046234, |
| "step": 24, |
| "train_speed(iter/s)": 0.011191 |
| }, |
| { |
| "epoch": 0.12132241431604489, |
| "grad_norm": 2.5066726207733154, |
| "learning_rate": 0.00016129032258064516, |
| "logits/chosen": -0.45425349473953247, |
| "logits/rejected": -0.46823209524154663, |
| "logps/chosen": -5.9361162185668945, |
| "logps/rejected": -12.616898536682129, |
| "loss": 0.9625968933105469, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.3818866014480591, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.14780059456825256, |
| "rewards/margins": 0.3101717531681061, |
| "rewards/rejected": -0.16237111389636993, |
| "step": 25, |
| "train_speed(iter/s)": 0.011195 |
| }, |
| { |
| "epoch": 0.1261753108886867, |
| "grad_norm": 3.1320712566375732, |
| "learning_rate": 0.00016774193548387098, |
| "logits/chosen": -0.47818127274513245, |
| "logits/rejected": -0.43204230070114136, |
| "logps/chosen": -13.965362548828125, |
| "logps/rejected": -15.635419845581055, |
| "loss": 1.1598286628723145, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6072496175765991, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.06804848462343216, |
| "rewards/margins": 0.44113460183143616, |
| "rewards/rejected": -0.3730861246585846, |
| "step": 26, |
| "train_speed(iter/s)": 0.011195 |
| }, |
| { |
| "epoch": 0.13102820746132848, |
| "grad_norm": 3.8042213916778564, |
| "learning_rate": 0.00017419354838709678, |
| "logits/chosen": -0.449899286031723, |
| "logits/rejected": -0.4721740484237671, |
| "logps/chosen": -9.528651237487793, |
| "logps/rejected": -12.257569313049316, |
| "loss": 1.2201637029647827, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6225947737693787, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.07078885287046432, |
| "rewards/margins": 0.28294628858566284, |
| "rewards/rejected": -0.21215742826461792, |
| "step": 27, |
| "train_speed(iter/s)": 0.011198 |
| }, |
| { |
| "epoch": 0.13588110403397027, |
| "grad_norm": 2.5645625591278076, |
| "learning_rate": 0.00018064516129032257, |
| "logits/chosen": -0.3968259394168854, |
| "logits/rejected": -0.41532400250434875, |
| "logps/chosen": -7.651938438415527, |
| "logps/rejected": -15.575130462646484, |
| "loss": 0.9698889851570129, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4171852469444275, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.1875855028629303, |
| "rewards/margins": 0.46900543570518494, |
| "rewards/rejected": -0.28141990303993225, |
| "step": 28, |
| "train_speed(iter/s)": 0.011199 |
| }, |
| { |
| "epoch": 0.14073400060661206, |
| "grad_norm": 3.2087931632995605, |
| "learning_rate": 0.0001870967741935484, |
| "logits/chosen": -0.38016778230667114, |
| "logits/rejected": -0.45883065462112427, |
| "logps/chosen": -9.820419311523438, |
| "logps/rejected": -17.21317481994629, |
| "loss": 1.156844973564148, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5829570293426514, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.21587063372135162, |
| "rewards/margins": 0.3926744759082794, |
| "rewards/rejected": -0.1768038272857666, |
| "step": 29, |
| "train_speed(iter/s)": 0.0112 |
| }, |
| { |
| "epoch": 0.14558689717925385, |
| "grad_norm": 3.5056474208831787, |
| "learning_rate": 0.00019354838709677422, |
| "logits/chosen": -0.43780362606048584, |
| "logits/rejected": -0.46428602933883667, |
| "logps/chosen": -12.7692289352417, |
| "logps/rejected": -16.947650909423828, |
| "loss": 1.2092680931091309, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7365323305130005, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.19128330051898956, |
| "rewards/margins": 0.7160227298736572, |
| "rewards/rejected": -0.5247394442558289, |
| "step": 30, |
| "train_speed(iter/s)": 0.011202 |
| }, |
| { |
| "epoch": 0.15043979375189567, |
| "grad_norm": 7.983421802520752, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.3911149799823761, |
| "logits/rejected": -0.4533422887325287, |
| "logps/chosen": -11.242105484008789, |
| "logps/rejected": -16.97354507446289, |
| "loss": 1.3208513259887695, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7320083975791931, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.08527207374572754, |
| "rewards/margins": 0.4059477746486664, |
| "rewards/rejected": -0.4912198483943939, |
| "step": 31, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.15529269032453746, |
| "grad_norm": 7.263556957244873, |
| "learning_rate": 0.0001999985678367986, |
| "logits/chosen": -0.4142841398715973, |
| "logits/rejected": -0.49599555134773254, |
| "logps/chosen": -8.028633117675781, |
| "logps/rejected": -19.247087478637695, |
| "loss": 1.1268211603164673, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5463874936103821, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.16298356652259827, |
| "rewards/margins": 0.4510626792907715, |
| "rewards/rejected": -0.6140461564064026, |
| "step": 32, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.16014558689717925, |
| "grad_norm": 7.569846153259277, |
| "learning_rate": 0.0001999942713882162, |
| "logits/chosen": -0.4265003502368927, |
| "logits/rejected": -0.42608755826950073, |
| "logps/chosen": -9.83013916015625, |
| "logps/rejected": -17.054443359375, |
| "loss": 1.2255939245224, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6033804416656494, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": -0.24369043111801147, |
| "rewards/margins": 0.24309512972831726, |
| "rewards/rejected": -0.48678556084632874, |
| "step": 33, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.16499848346982104, |
| "grad_norm": 3.240605592727661, |
| "learning_rate": 0.00019998711077731716, |
| "logits/chosen": -0.4495273232460022, |
| "logits/rejected": -0.3768068850040436, |
| "logps/chosen": -12.2802095413208, |
| "logps/rejected": -12.816421508789062, |
| "loss": 1.085668683052063, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5540680289268494, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0720956027507782, |
| "rewards/margins": 0.47421759366989136, |
| "rewards/rejected": -0.40212199091911316, |
| "step": 34, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.16985138004246284, |
| "grad_norm": 2.972562551498413, |
| "learning_rate": 0.00019997708620920465, |
| "logits/chosen": -0.4633820354938507, |
| "logits/rejected": -0.5037291646003723, |
| "logps/chosen": -8.281206130981445, |
| "logps/rejected": -17.908266067504883, |
| "loss": 1.0561590194702148, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6168872714042664, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13331066071987152, |
| "rewards/margins": 0.9570158123970032, |
| "rewards/rejected": -0.8237051963806152, |
| "step": 35, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.17470427661510465, |
| "grad_norm": 5.980532646179199, |
| "learning_rate": 0.00019996419797101514, |
| "logits/chosen": -0.39558717608451843, |
| "logits/rejected": -0.431841105222702, |
| "logps/chosen": -10.584002494812012, |
| "logps/rejected": -19.50037956237793, |
| "loss": 1.2352135181427002, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6387596130371094, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.029237767681479454, |
| "rewards/margins": 0.46423807740211487, |
| "rewards/rejected": -0.43500036001205444, |
| "step": 36, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.17955717318774644, |
| "grad_norm": 5.786473751068115, |
| "learning_rate": 0.00019994844643190975, |
| "logits/chosen": -0.3947299122810364, |
| "logits/rejected": -0.5335777997970581, |
| "logps/chosen": -10.800368309020996, |
| "logps/rejected": -19.217914581298828, |
| "loss": 1.4149001836776733, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8904857635498047, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.04618726670742035, |
| "rewards/margins": 0.49434420466423035, |
| "rewards/rejected": -0.5405314564704895, |
| "step": 37, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.18441006976038823, |
| "grad_norm": 2.6556568145751953, |
| "learning_rate": 0.00019992983204306403, |
| "logits/chosen": -0.39078089594841003, |
| "logits/rejected": -0.39872410893440247, |
| "logps/chosen": -12.065397262573242, |
| "logps/rejected": -19.61814308166504, |
| "loss": 1.1651582717895508, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5659552812576294, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.06024760380387306, |
| "rewards/margins": 0.4251976013183594, |
| "rewards/rejected": -0.36494994163513184, |
| "step": 38, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.18926296633303002, |
| "grad_norm": 4.3966779708862305, |
| "learning_rate": 0.0001999083553376548, |
| "logits/chosen": -0.4178510308265686, |
| "logits/rejected": -0.42042505741119385, |
| "logps/chosen": -10.572690963745117, |
| "logps/rejected": -17.888080596923828, |
| "loss": 1.1389288902282715, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6220788955688477, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.14774595201015472, |
| "rewards/margins": 0.5430118441581726, |
| "rewards/rejected": -0.3952659070491791, |
| "step": 39, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.19411586290567182, |
| "grad_norm": 2.347978115081787, |
| "learning_rate": 0.00019988401693084502, |
| "logits/chosen": -0.4415282905101776, |
| "logits/rejected": -0.4420144259929657, |
| "logps/chosen": -10.284622192382812, |
| "logps/rejected": -15.651193618774414, |
| "loss": 1.131446123123169, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5388365387916565, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.08026871085166931, |
| "rewards/margins": 0.271782785654068, |
| "rewards/rejected": -0.19151407480239868, |
| "step": 40, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.1989687594783136, |
| "grad_norm": 3.0811686515808105, |
| "learning_rate": 0.0001998568175197661, |
| "logits/chosen": -0.48944562673568726, |
| "logits/rejected": -0.48836228251457214, |
| "logps/chosen": -10.908992767333984, |
| "logps/rejected": -16.806554794311523, |
| "loss": 1.1378778219223022, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5336601734161377, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.026157543063163757, |
| "rewards/margins": 0.3137216567993164, |
| "rewards/rejected": -0.33987918496131897, |
| "step": 41, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.20382165605095542, |
| "grad_norm": 3.496178150177002, |
| "learning_rate": 0.00019982675788349794, |
| "logits/chosen": -0.4175988435745239, |
| "logits/rejected": -0.49307048320770264, |
| "logps/chosen": -8.874513626098633, |
| "logps/rejected": -16.90668296813965, |
| "loss": 1.237853765487671, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6065552830696106, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.028882348909974098, |
| "rewards/margins": 0.22182509303092957, |
| "rewards/rejected": -0.19294273853302002, |
| "step": 42, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.20867455262359721, |
| "grad_norm": 2.585383892059326, |
| "learning_rate": 0.00019979383888304666, |
| "logits/chosen": -0.40949565172195435, |
| "logits/rejected": -0.3888520896434784, |
| "logps/chosen": -12.942455291748047, |
| "logps/rejected": -16.93043327331543, |
| "loss": 1.1456714868545532, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5834133625030518, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.029632963240146637, |
| "rewards/margins": 0.47202715277671814, |
| "rewards/rejected": -0.4423941969871521, |
| "step": 43, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.213527449196239, |
| "grad_norm": 2.6477532386779785, |
| "learning_rate": 0.0001997580614613199, |
| "logits/chosen": -0.4283292591571808, |
| "logits/rejected": -0.417165070772171, |
| "logps/chosen": -11.886260986328125, |
| "logps/rejected": -22.782238006591797, |
| "loss": 1.126604676246643, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6129148602485657, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.020920906215906143, |
| "rewards/margins": 0.8850383162498474, |
| "rewards/rejected": -0.9059593677520752, |
| "step": 44, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.2183803457688808, |
| "grad_norm": 2.563398599624634, |
| "learning_rate": 0.00019971942664309968, |
| "logits/chosen": -0.4108792245388031, |
| "logits/rejected": -0.4292621612548828, |
| "logps/chosen": -10.143468856811523, |
| "logps/rejected": -21.13189125061035, |
| "loss": 0.957160472869873, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4789010286331177, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.04898708686232567, |
| "rewards/margins": 0.9526273012161255, |
| "rewards/rejected": -0.9036401510238647, |
| "step": 45, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.2232332423415226, |
| "grad_norm": 5.1497344970703125, |
| "learning_rate": 0.00019967793553501346, |
| "logits/chosen": -0.3103107213973999, |
| "logits/rejected": -0.3952590227127075, |
| "logps/chosen": -12.430742263793945, |
| "logps/rejected": -20.878339767456055, |
| "loss": 1.2279822826385498, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6726821660995483, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.020153822377324104, |
| "rewards/margins": 0.7637170553207397, |
| "rewards/rejected": -0.7435632348060608, |
| "step": 46, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.2280861389141644, |
| "grad_norm": 3.151014566421509, |
| "learning_rate": 0.00019963358932550192, |
| "logits/chosen": -0.3890960216522217, |
| "logits/rejected": -0.4021611511707306, |
| "logps/chosen": -10.522011756896973, |
| "logps/rejected": -18.351852416992188, |
| "loss": 1.1763105392456055, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5507612228393555, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.15107618272304535, |
| "rewards/margins": 0.30810290575027466, |
| "rewards/rejected": -0.4591791033744812, |
| "step": 47, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.2329390354868062, |
| "grad_norm": 4.961374759674072, |
| "learning_rate": 0.00019958638928478523, |
| "logits/chosen": -0.31074726581573486, |
| "logits/rejected": -0.3801327645778656, |
| "logps/chosen": -9.19954776763916, |
| "logps/rejected": -21.655811309814453, |
| "loss": 0.9774930477142334, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5149185657501221, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.030308127403259277, |
| "rewards/margins": 0.8868625164031982, |
| "rewards/rejected": -0.8565543293952942, |
| "step": 48, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.23779193205944799, |
| "grad_norm": 4.1918511390686035, |
| "learning_rate": 0.00019953633676482669, |
| "logits/chosen": -0.3845188617706299, |
| "logits/rejected": -0.4037459194660187, |
| "logps/chosen": -14.745367050170898, |
| "logps/rejected": -18.788145065307617, |
| "loss": 1.2587707042694092, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7012693285942078, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.11992412805557251, |
| "rewards/margins": 0.648144543170929, |
| "rewards/rejected": -0.5282202959060669, |
| "step": 49, |
| "train_speed(iter/s)": 0.011207 |
| }, |
| { |
| "epoch": 0.24264482863208978, |
| "grad_norm": 3.7718863487243652, |
| "learning_rate": 0.00019948343319929374, |
| "logits/chosen": -0.3694141209125519, |
| "logits/rejected": -0.474610835313797, |
| "logps/chosen": -10.025066375732422, |
| "logps/rejected": -20.226930618286133, |
| "loss": 1.0972646474838257, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5580352544784546, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.015783965587615967, |
| "rewards/margins": 0.48214325308799744, |
| "rewards/rejected": -0.4979272484779358, |
| "step": 50, |
| "train_speed(iter/s)": 0.011207 |
| }, |
| { |
| "epoch": 0.24749772520473157, |
| "grad_norm": 3.204578161239624, |
| "learning_rate": 0.00019942768010351727, |
| "logits/chosen": -0.5031870007514954, |
| "logits/rejected": -0.46329694986343384, |
| "logps/chosen": -13.154851913452148, |
| "logps/rejected": -16.299560546875, |
| "loss": 1.1670925617218018, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.648939311504364, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.07203729450702667, |
| "rewards/margins": 0.5248441696166992, |
| "rewards/rejected": -0.5968814492225647, |
| "step": 51, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.2523506217773734, |
| "grad_norm": 3.282288074493408, |
| "learning_rate": 0.00019936907907444788, |
| "logits/chosen": -0.40597957372665405, |
| "logits/rejected": -0.4711161255836487, |
| "logps/chosen": -10.530755996704102, |
| "logps/rejected": -18.246557235717773, |
| "loss": 1.1418707370758057, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6192983984947205, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.012871884740889072, |
| "rewards/margins": 0.4392779469490051, |
| "rewards/rejected": -0.4264060854911804, |
| "step": 52, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.25720351835001515, |
| "grad_norm": 3.151282787322998, |
| "learning_rate": 0.00019930763179061036, |
| "logits/chosen": -0.4330732226371765, |
| "logits/rejected": -0.456554651260376, |
| "logps/chosen": -8.236556053161621, |
| "logps/rejected": -15.27733039855957, |
| "loss": 1.168614149093628, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6144590973854065, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.12388759851455688, |
| "rewards/margins": 0.6339080333709717, |
| "rewards/rejected": -0.5100203156471252, |
| "step": 53, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.26205641492265697, |
| "grad_norm": 3.4259731769561768, |
| "learning_rate": 0.00019924334001205542, |
| "logits/chosen": -0.4073447287082672, |
| "logits/rejected": -0.42736995220184326, |
| "logps/chosen": -11.766214370727539, |
| "logps/rejected": -20.65784454345703, |
| "loss": 1.234757423400879, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6889488697052002, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08055731654167175, |
| "rewards/margins": 0.7608071565628052, |
| "rewards/rejected": -0.680249810218811, |
| "step": 54, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.26690931149529873, |
| "grad_norm": 2.5751075744628906, |
| "learning_rate": 0.00019917620558030947, |
| "logits/chosen": -0.45161640644073486, |
| "logits/rejected": -0.40232208371162415, |
| "logps/chosen": -10.377983093261719, |
| "logps/rejected": -15.469682693481445, |
| "loss": 1.1288338899612427, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6078510284423828, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.13011425733566284, |
| "rewards/margins": 0.7133916020393372, |
| "rewards/rejected": -0.5832774043083191, |
| "step": 55, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.27176220806794055, |
| "grad_norm": 1.9270464181900024, |
| "learning_rate": 0.0001991062304183218, |
| "logits/chosen": -0.3679361641407013, |
| "logits/rejected": -0.4820949137210846, |
| "logps/chosen": -6.5226335525512695, |
| "logps/rejected": -21.38983917236328, |
| "loss": 0.9311513900756836, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.48036620020866394, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1770131140947342, |
| "rewards/margins": 0.9639515280723572, |
| "rewards/rejected": -0.7869384288787842, |
| "step": 56, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.27661510464058237, |
| "grad_norm": 2.343013048171997, |
| "learning_rate": 0.0001990334165304094, |
| "logits/chosen": -0.3758355975151062, |
| "logits/rejected": -0.446993350982666, |
| "logps/chosen": -10.01812744140625, |
| "logps/rejected": -22.936779022216797, |
| "loss": 1.017271637916565, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5227488875389099, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.20892874896526337, |
| "rewards/margins": 0.9375370144844055, |
| "rewards/rejected": -0.728608250617981, |
| "step": 57, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.28146800121322413, |
| "grad_norm": 3.4284698963165283, |
| "learning_rate": 0.00019895776600219973, |
| "logits/chosen": -0.3820294141769409, |
| "logits/rejected": -0.4261220097541809, |
| "logps/chosen": -12.147420883178711, |
| "logps/rejected": -21.496156692504883, |
| "loss": 1.173353910446167, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.579372763633728, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.036380209028720856, |
| "rewards/margins": 0.430204302072525, |
| "rewards/rejected": -0.46658453345298767, |
| "step": 58, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.28632089778586595, |
| "grad_norm": 3.7487685680389404, |
| "learning_rate": 0.0001988792810005708, |
| "logits/chosen": -0.3050299882888794, |
| "logits/rejected": -0.40809452533721924, |
| "logps/chosen": -10.987959861755371, |
| "logps/rejected": -24.92184066772461, |
| "loss": 1.2469043731689453, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6960954070091248, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.009784307330846786, |
| "rewards/margins": 0.4835450351238251, |
| "rewards/rejected": -0.473760724067688, |
| "step": 59, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.2911737943585077, |
| "grad_norm": 3.093308448791504, |
| "learning_rate": 0.00019879796377358927, |
| "logits/chosen": -0.37827837467193604, |
| "logits/rejected": -0.4184655547142029, |
| "logps/chosen": -7.49070930480957, |
| "logps/rejected": -17.80307388305664, |
| "loss": 1.0403252840042114, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.49073174595832825, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.00874985009431839, |
| "rewards/margins": 0.46609920263290405, |
| "rewards/rejected": -0.47484904527664185, |
| "step": 60, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.2960266909311495, |
| "grad_norm": 3.3970422744750977, |
| "learning_rate": 0.00019871381665044594, |
| "logits/chosen": -0.3776484727859497, |
| "logits/rejected": -0.3863822817802429, |
| "logps/chosen": -9.921972274780273, |
| "logps/rejected": -15.067032814025879, |
| "loss": 1.250232219696045, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7360854744911194, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.09412748366594315, |
| "rewards/margins": 0.5320085287094116, |
| "rewards/rejected": -0.4378809928894043, |
| "step": 61, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.30087958750379135, |
| "grad_norm": 2.651646137237549, |
| "learning_rate": 0.00019862684204138907, |
| "logits/chosen": -0.3813178539276123, |
| "logits/rejected": -0.3218821585178375, |
| "logps/chosen": -12.840286254882812, |
| "logps/rejected": -17.035186767578125, |
| "loss": 1.1197326183319092, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6345531940460205, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.10404972732067108, |
| "rewards/margins": 0.6996921896934509, |
| "rewards/rejected": -0.5956425070762634, |
| "step": 62, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.3057324840764331, |
| "grad_norm": 2.7409791946411133, |
| "learning_rate": 0.00019853704243765532, |
| "logits/chosen": -0.21578770875930786, |
| "logits/rejected": -0.40845662355422974, |
| "logps/chosen": -9.395906448364258, |
| "logps/rejected": -28.66230583190918, |
| "loss": 1.0102654695510864, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6361392736434937, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.23129011690616608, |
| "rewards/margins": 1.06308913230896, |
| "rewards/rejected": -0.8317990303039551, |
| "step": 63, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.3105853806490749, |
| "grad_norm": 2.692429304122925, |
| "learning_rate": 0.00019844442041139847, |
| "logits/chosen": -0.35631364583969116, |
| "logits/rejected": -0.3108435571193695, |
| "logps/chosen": -11.395570755004883, |
| "logps/rejected": -16.973485946655273, |
| "loss": 0.9985603094100952, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5419827699661255, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.07737573981285095, |
| "rewards/margins": 0.7498005628585815, |
| "rewards/rejected": -0.672424852848053, |
| "step": 64, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.3154382772217167, |
| "grad_norm": 5.938281536102295, |
| "learning_rate": 0.00019834897861561573, |
| "logits/chosen": -0.3479470908641815, |
| "logits/rejected": -0.3646903932094574, |
| "logps/chosen": -8.458881378173828, |
| "logps/rejected": -18.25616455078125, |
| "loss": 1.0791547298431396, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5545474290847778, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.05356256291270256, |
| "rewards/margins": 0.5557125210762024, |
| "rewards/rejected": -0.6092751026153564, |
| "step": 65, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.3202911737943585, |
| "grad_norm": 2.738814353942871, |
| "learning_rate": 0.00019825071978407156, |
| "logits/chosen": -0.29352137446403503, |
| "logits/rejected": -0.2870010733604431, |
| "logps/chosen": -12.466264724731445, |
| "logps/rejected": -22.961450576782227, |
| "loss": 0.9836492538452148, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5442343354225159, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.123663991689682, |
| "rewards/margins": 0.9405882954597473, |
| "rewards/rejected": -0.8169243335723877, |
| "step": 66, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.3251440703670003, |
| "grad_norm": 3.51104998588562, |
| "learning_rate": 0.00019814964673121966, |
| "logits/chosen": -0.3337523937225342, |
| "logits/rejected": -0.3147628903388977, |
| "logps/chosen": -11.947603225708008, |
| "logps/rejected": -17.78832244873047, |
| "loss": 1.2882845401763916, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8149017095565796, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.04635877162218094, |
| "rewards/margins": 0.7259438037872314, |
| "rewards/rejected": -0.7723026275634766, |
| "step": 67, |
| "train_speed(iter/s)": 0.011213 |
| }, |
| { |
| "epoch": 0.3299969669396421, |
| "grad_norm": 3.068816661834717, |
| "learning_rate": 0.00019804576235212218, |
| "logits/chosen": -0.2215232253074646, |
| "logits/rejected": -0.2657161355018616, |
| "logps/chosen": -11.033540725708008, |
| "logps/rejected": -25.50714874267578, |
| "loss": 0.8847059607505798, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5216297507286072, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09357042610645294, |
| "rewards/margins": 1.1900687217712402, |
| "rewards/rejected": -1.0964982509613037, |
| "step": 68, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.3348498635122839, |
| "grad_norm": 2.5830941200256348, |
| "learning_rate": 0.00019793906962236682, |
| "logits/chosen": -0.24091127514839172, |
| "logits/rejected": -0.29128578305244446, |
| "logps/chosen": -10.63074016571045, |
| "logps/rejected": -26.628215789794922, |
| "loss": 1.0002039670944214, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5858277082443237, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06674355268478394, |
| "rewards/margins": 1.1017783880233765, |
| "rewards/rejected": -1.0350347757339478, |
| "step": 69, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.33970276008492567, |
| "grad_norm": 2.7311224937438965, |
| "learning_rate": 0.00019782957159798157, |
| "logits/chosen": -0.30429190397262573, |
| "logits/rejected": -0.2560291290283203, |
| "logps/chosen": -12.162567138671875, |
| "logps/rejected": -17.434350967407227, |
| "loss": 1.063048005104065, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6400512456893921, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.18220071494579315, |
| "rewards/margins": 1.0044963359832764, |
| "rewards/rejected": -0.8222954869270325, |
| "step": 70, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.3445556566575675, |
| "grad_norm": 3.414921760559082, |
| "learning_rate": 0.0001977172714153473, |
| "logits/chosen": -0.28515180945396423, |
| "logits/rejected": -0.3108617961406708, |
| "logps/chosen": -10.81340503692627, |
| "logps/rejected": -26.520463943481445, |
| "loss": 1.1172614097595215, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.685325026512146, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1235567182302475, |
| "rewards/margins": 1.0347576141357422, |
| "rewards/rejected": -0.9112008810043335, |
| "step": 71, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.3494085532302093, |
| "grad_norm": 4.576763153076172, |
| "learning_rate": 0.00019760217229110777, |
| "logits/chosen": -0.36723142862319946, |
| "logits/rejected": -0.3557523488998413, |
| "logps/chosen": -12.69983196258545, |
| "logps/rejected": -20.64535140991211, |
| "loss": 1.1201751232147217, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.692030668258667, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.049573928117752075, |
| "rewards/margins": 1.0138797760009766, |
| "rewards/rejected": -0.9643057584762573, |
| "step": 72, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.35426144980285107, |
| "grad_norm": 4.538747310638428, |
| "learning_rate": 0.00019748427752207755, |
| "logits/chosen": -0.34310680627822876, |
| "logits/rejected": -0.366294801235199, |
| "logps/chosen": -12.221488952636719, |
| "logps/rejected": -21.165037155151367, |
| "loss": 1.2563344240188599, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7162244915962219, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.12930546700954437, |
| "rewards/margins": 0.6588571667671204, |
| "rewards/rejected": -0.7881627082824707, |
| "step": 73, |
| "train_speed(iter/s)": 0.011213 |
| }, |
| { |
| "epoch": 0.3591143463754929, |
| "grad_norm": 2.0262386798858643, |
| "learning_rate": 0.00019736359048514766, |
| "logits/chosen": -0.334412157535553, |
| "logits/rejected": -0.37951967120170593, |
| "logps/chosen": -14.834428787231445, |
| "logps/rejected": -21.54924964904785, |
| "loss": 1.0132083892822266, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6331067085266113, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1725282073020935, |
| "rewards/margins": 1.028924822807312, |
| "rewards/rejected": -0.8563966751098633, |
| "step": 74, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.36396724294813465, |
| "grad_norm": 3.2811484336853027, |
| "learning_rate": 0.00019724011463718885, |
| "logits/chosen": -0.360372394323349, |
| "logits/rejected": -0.35564228892326355, |
| "logps/chosen": -11.420612335205078, |
| "logps/rejected": -19.772428512573242, |
| "loss": 1.2705371379852295, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6594747304916382, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.09604446589946747, |
| "rewards/margins": 0.5641387701034546, |
| "rewards/rejected": -0.660183310508728, |
| "step": 75, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.36882013952077647, |
| "grad_norm": 3.251281976699829, |
| "learning_rate": 0.0001971138535149523, |
| "logits/chosen": -0.3563613295555115, |
| "logits/rejected": -0.38878530263900757, |
| "logps/chosen": -12.576432228088379, |
| "logps/rejected": -21.920602798461914, |
| "loss": 0.9680290222167969, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5066214799880981, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.009179912507534027, |
| "rewards/margins": 0.7688724994659424, |
| "rewards/rejected": -0.7780523896217346, |
| "step": 76, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.37367303609341823, |
| "grad_norm": 2.962810516357422, |
| "learning_rate": 0.00019698481073496878, |
| "logits/chosen": -0.3119925856590271, |
| "logits/rejected": -0.4056414067745209, |
| "logps/chosen": -9.659420013427734, |
| "logps/rejected": -23.653026580810547, |
| "loss": 0.9925091862678528, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5779968500137329, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.030417252331972122, |
| "rewards/margins": 0.987002968788147, |
| "rewards/rejected": -0.956585705280304, |
| "step": 77, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.37852593266606005, |
| "grad_norm": 10.599248886108398, |
| "learning_rate": 0.00019685298999344458, |
| "logits/chosen": -0.35163983702659607, |
| "logits/rejected": -0.3840740919113159, |
| "logps/chosen": -8.082690238952637, |
| "logps/rejected": -25.49774169921875, |
| "loss": 1.0417295694351196, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7148680686950684, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.03415816277265549, |
| "rewards/margins": 1.2999666929244995, |
| "rewards/rejected": -1.3341249227523804, |
| "step": 78, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.38337882923870187, |
| "grad_norm": 3.775139331817627, |
| "learning_rate": 0.00019671839506615614, |
| "logits/chosen": -0.3263033926486969, |
| "logits/rejected": -0.336015522480011, |
| "logps/chosen": -12.873096466064453, |
| "logps/rejected": -25.73821258544922, |
| "loss": 1.011587142944336, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5964220762252808, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15943416953086853, |
| "rewards/margins": 1.001997947692871, |
| "rewards/rejected": -0.8425638675689697, |
| "step": 79, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.38823172581134363, |
| "grad_norm": 2.6962356567382812, |
| "learning_rate": 0.00019658102980834146, |
| "logits/chosen": -0.42636168003082275, |
| "logits/rejected": -0.40870201587677, |
| "logps/chosen": -9.675471305847168, |
| "logps/rejected": -17.18560791015625, |
| "loss": 1.0823123455047607, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6633837819099426, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.09595775604248047, |
| "rewards/margins": 0.8637285232543945, |
| "rewards/rejected": -0.7677708268165588, |
| "step": 80, |
| "train_speed(iter/s)": 0.011213 |
| }, |
| { |
| "epoch": 0.39308462238398545, |
| "grad_norm": 3.252925395965576, |
| "learning_rate": 0.0001964408981545898, |
| "logits/chosen": -0.3987385034561157, |
| "logits/rejected": -0.3940485715866089, |
| "logps/chosen": -11.167372703552246, |
| "logps/rejected": -19.442848205566406, |
| "loss": 1.2748379707336426, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8224515914916992, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.004448680207133293, |
| "rewards/margins": 0.9587394595146179, |
| "rewards/rejected": -0.9631881713867188, |
| "step": 81, |
| "train_speed(iter/s)": 0.011213 |
| }, |
| { |
| "epoch": 0.3979375189566272, |
| "grad_norm": 2.8878297805786133, |
| "learning_rate": 0.00019629800411872928, |
| "logits/chosen": -0.35343819856643677, |
| "logits/rejected": -0.4096568822860718, |
| "logps/chosen": -8.287420272827148, |
| "logps/rejected": -20.879899978637695, |
| "loss": 0.9840810894966125, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5656735897064209, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.13655401766300201, |
| "rewards/margins": 0.9052618145942688, |
| "rewards/rejected": -0.768707811832428, |
| "step": 82, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.40279041552926903, |
| "grad_norm": 2.34645676612854, |
| "learning_rate": 0.00019615235179371135, |
| "logits/chosen": -0.31612086296081543, |
| "logits/rejected": -0.39879709482192993, |
| "logps/chosen": -9.401403427124023, |
| "logps/rejected": -27.325342178344727, |
| "loss": 0.888444721698761, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.569918692111969, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.23369260132312775, |
| "rewards/margins": 1.3009941577911377, |
| "rewards/rejected": -1.0673015117645264, |
| "step": 83, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.40764331210191085, |
| "grad_norm": 2.7571706771850586, |
| "learning_rate": 0.00019600394535149404, |
| "logits/chosen": -0.34237152338027954, |
| "logits/rejected": -0.38029745221138, |
| "logps/chosen": -9.792423248291016, |
| "logps/rejected": -21.61672592163086, |
| "loss": 1.0087002515792847, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5312238335609436, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.07024291902780533, |
| "rewards/margins": 0.8068092465400696, |
| "rewards/rejected": -0.7365662455558777, |
| "step": 84, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.4124962086745526, |
| "grad_norm": 2.0822556018829346, |
| "learning_rate": 0.0001958527890429223, |
| "logits/chosen": -0.3475739657878876, |
| "logits/rejected": -0.31929460167884827, |
| "logps/chosen": -13.526400566101074, |
| "logps/rejected": -25.060049057006836, |
| "loss": 0.8273504376411438, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5433682203292847, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2506234347820282, |
| "rewards/margins": 1.4649745225906372, |
| "rewards/rejected": -1.2143512964248657, |
| "step": 85, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.41734910524719443, |
| "grad_norm": 2.4087798595428467, |
| "learning_rate": 0.0001956988871976062, |
| "logits/chosen": -0.27175548672676086, |
| "logits/rejected": -0.3286592662334442, |
| "logps/chosen": -7.984870910644531, |
| "logps/rejected": -24.256772994995117, |
| "loss": 0.7937061786651611, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.42936819791793823, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.059691205620765686, |
| "rewards/margins": 1.0793851613998413, |
| "rewards/rejected": -1.0196938514709473, |
| "step": 86, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.4222020018198362, |
| "grad_norm": 2.174376964569092, |
| "learning_rate": 0.00019554224422379686, |
| "logits/chosen": -0.23324547708034515, |
| "logits/rejected": -0.30267754197120667, |
| "logps/chosen": -11.11595344543457, |
| "logps/rejected": -27.55086326599121, |
| "loss": 0.799884021282196, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4499448537826538, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.12257401645183563, |
| "rewards/margins": 1.2369678020477295, |
| "rewards/rejected": -1.114393711090088, |
| "step": 87, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.427054898392478, |
| "grad_norm": 4.111970901489258, |
| "learning_rate": 0.00019538286460826041, |
| "logits/chosen": -0.2658248543739319, |
| "logits/rejected": -0.2973652482032776, |
| "logps/chosen": -8.75698471069336, |
| "logps/rejected": -22.72409439086914, |
| "loss": 0.8912484645843506, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.48959213495254517, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09087920188903809, |
| "rewards/margins": 1.290701150894165, |
| "rewards/rejected": -1.199821949005127, |
| "step": 88, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.43190779496511983, |
| "grad_norm": 3.519275665283203, |
| "learning_rate": 0.0001952207529161492, |
| "logits/chosen": -0.3236900866031647, |
| "logits/rejected": -0.29652339220046997, |
| "logps/chosen": -15.083149909973145, |
| "logps/rejected": -29.6583251953125, |
| "loss": 1.028883695602417, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7164204120635986, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.056427597999572754, |
| "rewards/margins": 1.5590038299560547, |
| "rewards/rejected": -1.5025763511657715, |
| "step": 89, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.4367606915377616, |
| "grad_norm": 5.104607105255127, |
| "learning_rate": 0.00019505591379087128, |
| "logits/chosen": -0.30673032999038696, |
| "logits/rejected": -0.284626841545105, |
| "logps/chosen": -18.149417877197266, |
| "logps/rejected": -27.66114616394043, |
| "loss": 1.1724226474761963, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8443087935447693, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12549515068531036, |
| "rewards/margins": 1.6692070960998535, |
| "rewards/rejected": -1.5437119007110596, |
| "step": 90, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.4416135881104034, |
| "grad_norm": 5.404626846313477, |
| "learning_rate": 0.00019488835195395718, |
| "logits/chosen": -0.25754019618034363, |
| "logits/rejected": -0.25054478645324707, |
| "logps/chosen": -11.13974380493164, |
| "logps/rejected": -29.229190826416016, |
| "loss": 0.9712298512458801, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.614799976348877, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.06202991306781769, |
| "rewards/margins": 1.4658480882644653, |
| "rewards/rejected": -1.4038183689117432, |
| "step": 91, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.4464664846830452, |
| "grad_norm": 4.149505138397217, |
| "learning_rate": 0.00019471807220492487, |
| "logits/chosen": -0.3054165542125702, |
| "logits/rejected": -0.3217582106590271, |
| "logps/chosen": -7.596218585968018, |
| "logps/rejected": -24.510986328125, |
| "loss": 0.8925672173500061, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5096707344055176, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.05373791605234146, |
| "rewards/margins": 1.6141862869262695, |
| "rewards/rejected": -1.560448408126831, |
| "step": 92, |
| "train_speed(iter/s)": 0.011214 |
| }, |
| { |
| "epoch": 0.451319381255687, |
| "grad_norm": 2.699763536453247, |
| "learning_rate": 0.00019454507942114218, |
| "logits/chosen": -0.4031309485435486, |
| "logits/rejected": -0.3797576129436493, |
| "logps/chosen": -11.298748970031738, |
| "logps/rejected": -21.17618751525879, |
| "loss": 0.9288226366043091, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5129250884056091, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.09262087196111679, |
| "rewards/margins": 1.186156988143921, |
| "rewards/rejected": -1.0935360193252563, |
| "step": 93, |
| "train_speed(iter/s)": 0.011215 |
| }, |
| { |
| "epoch": 0.4561722778283288, |
| "grad_norm": 2.987886905670166, |
| "learning_rate": 0.00019436937855768705, |
| "logits/chosen": -0.3467521667480469, |
| "logits/rejected": -0.38047531247138977, |
| "logps/chosen": -8.127204895019531, |
| "logps/rejected": -27.24859619140625, |
| "loss": 0.9090497493743896, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5626466274261475, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.01449224166572094, |
| "rewards/margins": 1.3198893070220947, |
| "rewards/rejected": -1.3053967952728271, |
| "step": 94, |
| "train_speed(iter/s)": 0.011216 |
| }, |
| { |
| "epoch": 0.46102517440097057, |
| "grad_norm": 6.682155609130859, |
| "learning_rate": 0.0001941909746472057, |
| "logits/chosen": -0.36795341968536377, |
| "logits/rejected": -0.33038249611854553, |
| "logps/chosen": -14.30661392211914, |
| "logps/rejected": -27.33529281616211, |
| "loss": 1.2022923231124878, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6685870289802551, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.14697454869747162, |
| "rewards/margins": 0.7069932818412781, |
| "rewards/rejected": -0.8539677858352661, |
| "step": 95, |
| "train_speed(iter/s)": 0.011215 |
| }, |
| { |
| "epoch": 0.4658780709736124, |
| "grad_norm": 6.086058616638184, |
| "learning_rate": 0.00019400987279976847, |
| "logits/chosen": -0.4552175998687744, |
| "logits/rejected": -0.3753393590450287, |
| "logps/chosen": -12.3525390625, |
| "logps/rejected": -24.96709632873535, |
| "loss": 1.1781623363494873, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8305119276046753, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.04669938236474991, |
| "rewards/margins": 1.6278446912765503, |
| "rewards/rejected": -1.674544095993042, |
| "step": 96, |
| "train_speed(iter/s)": 0.011216 |
| }, |
| { |
| "epoch": 0.47073096754625415, |
| "grad_norm": 2.856499671936035, |
| "learning_rate": 0.00019382607820272343, |
| "logits/chosen": -0.34398210048675537, |
| "logits/rejected": -0.3632335662841797, |
| "logps/chosen": -13.498963356018066, |
| "logps/rejected": -23.799148559570312, |
| "loss": 0.9125381708145142, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6083617806434631, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16915112733840942, |
| "rewards/margins": 1.5896618366241455, |
| "rewards/rejected": -1.4205105304718018, |
| "step": 97, |
| "train_speed(iter/s)": 0.011216 |
| }, |
| { |
| "epoch": 0.47558386411889597, |
| "grad_norm": 2.2536661624908447, |
| "learning_rate": 0.00019363959612054768, |
| "logits/chosen": -0.3522893190383911, |
| "logits/rejected": -0.3821050822734833, |
| "logps/chosen": -7.958831310272217, |
| "logps/rejected": -22.634958267211914, |
| "loss": 0.9238535165786743, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5074257254600525, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05879054591059685, |
| "rewards/margins": 1.1695164442062378, |
| "rewards/rejected": -1.1107258796691895, |
| "step": 98, |
| "train_speed(iter/s)": 0.011217 |
| }, |
| { |
| "epoch": 0.48043676069153773, |
| "grad_norm": 4.499397277832031, |
| "learning_rate": 0.00019345043189469675, |
| "logits/chosen": -0.3616553843021393, |
| "logits/rejected": -0.3694867789745331, |
| "logps/chosen": -9.15666389465332, |
| "logps/rejected": -22.706497192382812, |
| "loss": 0.8626678586006165, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4881688058376312, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.22088897228240967, |
| "rewards/margins": 1.2141640186309814, |
| "rewards/rejected": -0.993274986743927, |
| "step": 99, |
| "train_speed(iter/s)": 0.011217 |
| }, |
| { |
| "epoch": 0.48528965726417955, |
| "grad_norm": 2.5381503105163574, |
| "learning_rate": 0.00019325859094345148, |
| "logits/chosen": -0.3770003914833069, |
| "logits/rejected": -0.4174918234348297, |
| "logps/chosen": -10.83204460144043, |
| "logps/rejected": -20.01091194152832, |
| "loss": 1.1128623485565186, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6556410193443298, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08420430123806, |
| "rewards/margins": 0.7714729905128479, |
| "rewards/rejected": -0.6872686743736267, |
| "step": 100, |
| "train_speed(iter/s)": 0.011217 |
| }, |
| { |
| "epoch": 0.49014255383682137, |
| "grad_norm": 3.571274995803833, |
| "learning_rate": 0.000193064078761763, |
| "logits/chosen": -0.3813248872756958, |
| "logits/rejected": -0.4329451024532318, |
| "logps/chosen": -11.104312896728516, |
| "logps/rejected": -24.20327377319336, |
| "loss": 1.1288418769836426, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7659333944320679, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13299986720085144, |
| "rewards/margins": 1.1876386404037476, |
| "rewards/rejected": -1.0546387434005737, |
| "step": 101, |
| "train_speed(iter/s)": 0.011197 |
| }, |
| { |
| "epoch": 0.49499545040946313, |
| "grad_norm": 3.4420199394226074, |
| "learning_rate": 0.00019286690092109495, |
| "logits/chosen": -0.3884170353412628, |
| "logits/rejected": -0.4291474521160126, |
| "logps/chosen": -8.859447479248047, |
| "logps/rejected": -19.800003051757812, |
| "loss": 1.0438815355300903, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6051428318023682, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.10340974479913712, |
| "rewards/margins": 0.9933844804763794, |
| "rewards/rejected": -0.8899747729301453, |
| "step": 102, |
| "train_speed(iter/s)": 0.011197 |
| }, |
| { |
| "epoch": 0.49984834698210495, |
| "grad_norm": 3.0272207260131836, |
| "learning_rate": 0.00019266706306926437, |
| "logits/chosen": -0.28712770342826843, |
| "logits/rejected": -0.482435941696167, |
| "logps/chosen": -8.935839653015137, |
| "logps/rejected": -26.6412353515625, |
| "loss": 1.1441514492034912, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6857534646987915, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1457400619983673, |
| "rewards/margins": 1.0326664447784424, |
| "rewards/rejected": -0.8869264125823975, |
| "step": 103, |
| "train_speed(iter/s)": 0.011197 |
| }, |
| { |
| "epoch": 0.5047012435547468, |
| "grad_norm": 2.2835068702697754, |
| "learning_rate": 0.00019246457093027955, |
| "logits/chosen": -0.3885180354118347, |
| "logits/rejected": -0.4434734582901001, |
| "logps/chosen": -7.748335838317871, |
| "logps/rejected": -27.667823791503906, |
| "loss": 0.7652894258499146, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.3754329979419708, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.06499351561069489, |
| "rewards/margins": 1.0574610233306885, |
| "rewards/rejected": -0.9924674034118652, |
| "step": 104, |
| "train_speed(iter/s)": 0.011197 |
| }, |
| { |
| "epoch": 0.5095541401273885, |
| "grad_norm": 3.3575081825256348, |
| "learning_rate": 0.0001922594303041764, |
| "logits/chosen": -0.314767062664032, |
| "logits/rejected": -0.418552041053772, |
| "logps/chosen": -9.141714096069336, |
| "logps/rejected": -26.07691192626953, |
| "loss": 0.8440835475921631, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.53468257188797, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18071942031383514, |
| "rewards/margins": 1.4176182746887207, |
| "rewards/rejected": -1.2368988990783691, |
| "step": 105, |
| "train_speed(iter/s)": 0.011197 |
| }, |
| { |
| "epoch": 0.5144070367000303, |
| "grad_norm": 3.938190460205078, |
| "learning_rate": 0.0001920516470668519, |
| "logits/chosen": -0.38045060634613037, |
| "logits/rejected": -0.46996402740478516, |
| "logps/chosen": -9.252923965454102, |
| "logps/rejected": -25.073986053466797, |
| "loss": 0.8858380913734436, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5138120651245117, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.24657589197158813, |
| "rewards/margins": 1.3090505599975586, |
| "rewards/rejected": -1.0624747276306152, |
| "step": 106, |
| "train_speed(iter/s)": 0.011198 |
| }, |
| { |
| "epoch": 0.5192599332726722, |
| "grad_norm": 3.807465076446533, |
| "learning_rate": 0.00019184122716989625, |
| "logits/chosen": -0.38855308294296265, |
| "logits/rejected": -0.384641170501709, |
| "logps/chosen": -10.419635772705078, |
| "logps/rejected": -23.18486213684082, |
| "loss": 1.1566160917282104, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6819642782211304, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.015541527420282364, |
| "rewards/margins": 1.0274978876113892, |
| "rewards/rejected": -1.0430394411087036, |
| "step": 107, |
| "train_speed(iter/s)": 0.011198 |
| }, |
| { |
| "epoch": 0.5241128298453139, |
| "grad_norm": 3.046457529067993, |
| "learning_rate": 0.0001916281766404221, |
| "logits/chosen": -0.41123640537261963, |
| "logits/rejected": -0.4135390818119049, |
| "logps/chosen": -10.575555801391602, |
| "logps/rejected": -25.36039924621582, |
| "loss": 0.9802708625793457, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5873898863792419, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.016565335914492607, |
| "rewards/margins": 1.2314006090164185, |
| "rewards/rejected": -1.2479660511016846, |
| "step": 108, |
| "train_speed(iter/s)": 0.011199 |
| }, |
| { |
| "epoch": 0.5289657264179557, |
| "grad_norm": 5.713778495788574, |
| "learning_rate": 0.00019141250158089202, |
| "logits/chosen": -0.36335325241088867, |
| "logits/rejected": -0.42550188302993774, |
| "logps/chosen": -11.970105171203613, |
| "logps/rejected": -30.854656219482422, |
| "loss": 1.0406579971313477, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7013589143753052, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.020902972668409348, |
| "rewards/margins": 1.5762500762939453, |
| "rewards/rejected": -1.5553470849990845, |
| "step": 109, |
| "train_speed(iter/s)": 0.011199 |
| }, |
| { |
| "epoch": 0.5338186229905975, |
| "grad_norm": 3.487213373184204, |
| "learning_rate": 0.0001911942081689437, |
| "logits/chosen": -0.33955633640289307, |
| "logits/rejected": -0.4554653465747833, |
| "logps/chosen": -13.618782997131348, |
| "logps/rejected": -27.65629005432129, |
| "loss": 1.2663241624832153, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.8004875183105469, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.034927308559417725, |
| "rewards/margins": 1.3416428565979004, |
| "rewards/rejected": -1.306715488433838, |
| "step": 110, |
| "train_speed(iter/s)": 0.011199 |
| }, |
| { |
| "epoch": 0.5386715195632393, |
| "grad_norm": 5.0910491943359375, |
| "learning_rate": 0.00019097330265721295, |
| "logits/chosen": -0.2648548185825348, |
| "logits/rejected": -0.3673378527164459, |
| "logps/chosen": -12.02265453338623, |
| "logps/rejected": -30.836153030395508, |
| "loss": 1.0310111045837402, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5808767676353455, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.09649889171123505, |
| "rewards/margins": 1.15187406539917, |
| "rewards/rejected": -1.248373031616211, |
| "step": 111, |
| "train_speed(iter/s)": 0.011199 |
| }, |
| { |
| "epoch": 0.5435244161358811, |
| "grad_norm": 2.8877944946289062, |
| "learning_rate": 0.00019074979137315468, |
| "logits/chosen": -0.4370982348918915, |
| "logits/rejected": -0.41886797547340393, |
| "logps/chosen": -8.66671371459961, |
| "logps/rejected": -21.121248245239258, |
| "loss": 0.915783703327179, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5682031512260437, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.0029869861900806427, |
| "rewards/margins": 1.161545753479004, |
| "rewards/rejected": -1.1645326614379883, |
| "step": 112, |
| "train_speed(iter/s)": 0.0112 |
| }, |
| { |
| "epoch": 0.5483773127085229, |
| "grad_norm": 6.939223289489746, |
| "learning_rate": 0.0001905236807188616, |
| "logits/chosen": -0.32761722803115845, |
| "logits/rejected": -0.3690508306026459, |
| "logps/chosen": -9.920032501220703, |
| "logps/rejected": -33.124969482421875, |
| "loss": 0.9988139867782593, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6303930282592773, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.05219712853431702, |
| "rewards/margins": 1.4993199110031128, |
| "rewards/rejected": -1.5515170097351074, |
| "step": 113, |
| "train_speed(iter/s)": 0.0112 |
| }, |
| { |
| "epoch": 0.5532302092811647, |
| "grad_norm": 3.1117734909057617, |
| "learning_rate": 0.00019029497717088086, |
| "logits/chosen": -0.3898760974407196, |
| "logits/rejected": -0.42370229959487915, |
| "logps/chosen": -8.5205078125, |
| "logps/rejected": -28.223445892333984, |
| "loss": 0.9178063869476318, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5913013219833374, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.11722102761268616, |
| "rewards/margins": 1.5129557847976685, |
| "rewards/rejected": -1.3957347869873047, |
| "step": 114, |
| "train_speed(iter/s)": 0.0112 |
| }, |
| { |
| "epoch": 0.5580831058538065, |
| "grad_norm": 9.934938430786133, |
| "learning_rate": 0.00019006368728002862, |
| "logits/chosen": -0.3294893503189087, |
| "logits/rejected": -0.4310123026371002, |
| "logps/chosen": -7.87459135055542, |
| "logps/rejected": -26.407917022705078, |
| "loss": 1.0052833557128906, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6052643656730652, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.08163556456565857, |
| "rewards/margins": 1.377348780632019, |
| "rewards/rejected": -1.295713186264038, |
| "step": 115, |
| "train_speed(iter/s)": 0.011201 |
| }, |
| { |
| "epoch": 0.5629360024264483, |
| "grad_norm": 2.110900640487671, |
| "learning_rate": 0.00018982981767120231, |
| "logits/chosen": -0.2874082624912262, |
| "logits/rejected": -0.27834200859069824, |
| "logps/chosen": -11.620476722717285, |
| "logps/rejected": -28.311389923095703, |
| "loss": 0.8064985871315002, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4605460464954376, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.03009045496582985, |
| "rewards/margins": 1.43373441696167, |
| "rewards/rejected": -1.4036438465118408, |
| "step": 116, |
| "train_speed(iter/s)": 0.0112 |
| }, |
| { |
| "epoch": 0.5677888989990901, |
| "grad_norm": 4.351490497589111, |
| "learning_rate": 0.0001895933750431908, |
| "logits/chosen": -0.31533291935920715, |
| "logits/rejected": -0.33318501710891724, |
| "logps/chosen": -9.928375244140625, |
| "logps/rejected": -26.295669555664062, |
| "loss": 1.0582681894302368, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6660478711128235, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.0869743824005127, |
| "rewards/margins": 1.2630525827407837, |
| "rewards/rejected": -1.3500269651412964, |
| "step": 117, |
| "train_speed(iter/s)": 0.011201 |
| }, |
| { |
| "epoch": 0.5726417955717319, |
| "grad_norm": 2.8951752185821533, |
| "learning_rate": 0.0001893543661684828, |
| "logits/chosen": -0.33839067816734314, |
| "logits/rejected": -0.3445620536804199, |
| "logps/chosen": -8.102829933166504, |
| "logps/rejected": -22.460590362548828, |
| "loss": 0.7739503979682922, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.43036913871765137, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1979444921016693, |
| "rewards/margins": 1.243354082107544, |
| "rewards/rejected": -1.0454095602035522, |
| "step": 118, |
| "train_speed(iter/s)": 0.011201 |
| }, |
| { |
| "epoch": 0.5774946921443737, |
| "grad_norm": 7.867568492889404, |
| "learning_rate": 0.00018911279789307254, |
| "logits/chosen": -0.2516986131668091, |
| "logits/rejected": -0.30931952595710754, |
| "logps/chosen": -8.088445663452148, |
| "logps/rejected": -28.628847122192383, |
| "loss": 0.8760090470314026, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6033605933189392, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18770426511764526, |
| "rewards/margins": 1.7780780792236328, |
| "rewards/rejected": -1.5903738737106323, |
| "step": 119, |
| "train_speed(iter/s)": 0.011201 |
| }, |
| { |
| "epoch": 0.5823475887170154, |
| "grad_norm": 9.06499195098877, |
| "learning_rate": 0.00018886867713626394, |
| "logits/chosen": -0.31181472539901733, |
| "logits/rejected": -0.35031330585479736, |
| "logps/chosen": -6.005623817443848, |
| "logps/rejected": -24.879501342773438, |
| "loss": 0.7998319864273071, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5190199613571167, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.15923379361629486, |
| "rewards/margins": 1.547129511833191, |
| "rewards/rejected": -1.3878957033157349, |
| "step": 120, |
| "train_speed(iter/s)": 0.011202 |
| }, |
| { |
| "epoch": 0.5872004852896573, |
| "grad_norm": 2.1969258785247803, |
| "learning_rate": 0.00018862201089047228, |
| "logits/chosen": -0.36267367005348206, |
| "logits/rejected": -0.35898861289024353, |
| "logps/chosen": -9.9768648147583, |
| "logps/rejected": -31.001258850097656, |
| "loss": 0.8342351913452148, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6177542805671692, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3305099904537201, |
| "rewards/margins": 2.1078670024871826, |
| "rewards/rejected": -1.7773568630218506, |
| "step": 121, |
| "train_speed(iter/s)": 0.011202 |
| }, |
| { |
| "epoch": 0.592053381862299, |
| "grad_norm": 4.638126850128174, |
| "learning_rate": 0.000188372806221024, |
| "logits/chosen": -0.2871503531932831, |
| "logits/rejected": -0.3778407573699951, |
| "logps/chosen": -11.12118911743164, |
| "logps/rejected": -28.530237197875977, |
| "loss": 1.1293538808822632, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7672280669212341, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.024127217009663582, |
| "rewards/margins": 1.5403187274932861, |
| "rewards/rejected": -1.5644460916519165, |
| "step": 122, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.5969062784349408, |
| "grad_norm": 2.8399314880371094, |
| "learning_rate": 0.0001881210702659542, |
| "logits/chosen": -0.2696555554866791, |
| "logits/rejected": -0.2814023196697235, |
| "logps/chosen": -12.141545295715332, |
| "logps/rejected": -28.03343391418457, |
| "loss": 0.8627264499664307, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5336970090866089, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.07849591970443726, |
| "rewards/margins": 1.3175827264785767, |
| "rewards/rejected": -1.2390867471694946, |
| "step": 123, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.6017591750075827, |
| "grad_norm": 2.7301621437072754, |
| "learning_rate": 0.00018786681023580235, |
| "logits/chosen": -0.26133084297180176, |
| "logits/rejected": -0.2869059145450592, |
| "logps/chosen": -12.609733581542969, |
| "logps/rejected": -32.32643508911133, |
| "loss": 0.7636784315109253, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.4288833439350128, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08298669010400772, |
| "rewards/margins": 1.697174072265625, |
| "rewards/rejected": -1.6141873598098755, |
| "step": 124, |
| "train_speed(iter/s)": 0.011202 |
| }, |
| { |
| "epoch": 0.6066120715802245, |
| "grad_norm": 4.0161309242248535, |
| "learning_rate": 0.0001876100334134056, |
| "logits/chosen": -0.28447672724723816, |
| "logits/rejected": -0.3584819436073303, |
| "logps/chosen": -17.320146560668945, |
| "logps/rejected": -27.78953742980957, |
| "loss": 1.2440454959869385, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.867017924785614, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12772110104560852, |
| "rewards/margins": 1.4953205585479736, |
| "rewards/rejected": -1.3675994873046875, |
| "step": 125, |
| "train_speed(iter/s)": 0.011202 |
| }, |
| { |
| "epoch": 0.6114649681528662, |
| "grad_norm": 4.7182159423828125, |
| "learning_rate": 0.00018735074715369026, |
| "logits/chosen": -0.3351546823978424, |
| "logits/rejected": -0.34775683283805847, |
| "logps/chosen": -12.783681869506836, |
| "logps/rejected": -28.113178253173828, |
| "loss": 1.1283553838729858, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7738282084465027, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.028632041066884995, |
| "rewards/margins": 1.5667368173599243, |
| "rewards/rejected": -1.5381048917770386, |
| "step": 126, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.6163178647255081, |
| "grad_norm": 3.1524689197540283, |
| "learning_rate": 0.00018708895888346118, |
| "logits/chosen": -0.33259081840515137, |
| "logits/rejected": -0.38818880915641785, |
| "logps/chosen": -8.918370246887207, |
| "logps/rejected": -27.71927261352539, |
| "loss": 0.9142798185348511, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6048630475997925, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.16594702005386353, |
| "rewards/margins": 1.7087321281433105, |
| "rewards/rejected": -1.5427850484848022, |
| "step": 127, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.6211707612981499, |
| "grad_norm": 2.636528491973877, |
| "learning_rate": 0.00018682467610118883, |
| "logits/chosen": -0.2616899013519287, |
| "logits/rejected": -0.31121164560317993, |
| "logps/chosen": -7.396364212036133, |
| "logps/rejected": -26.648653030395508, |
| "loss": 0.685588538646698, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.40424948930740356, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18888305127620697, |
| "rewards/margins": 1.6681468486785889, |
| "rewards/rejected": -1.4792636632919312, |
| "step": 128, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.6260236578707916, |
| "grad_norm": 5.758305072784424, |
| "learning_rate": 0.00018655790637679478, |
| "logits/chosen": -0.3434281349182129, |
| "logits/rejected": -0.36044564843177795, |
| "logps/chosen": -8.97478199005127, |
| "logps/rejected": -31.279216766357422, |
| "loss": 0.9071744680404663, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5397835373878479, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.0950365662574768, |
| "rewards/margins": 1.9973032474517822, |
| "rewards/rejected": -1.9022667407989502, |
| "step": 129, |
| "train_speed(iter/s)": 0.011203 |
| }, |
| { |
| "epoch": 0.6308765544434334, |
| "grad_norm": 5.446237087249756, |
| "learning_rate": 0.00018628865735143463, |
| "logits/chosen": -0.40321969985961914, |
| "logits/rejected": -0.44060012698173523, |
| "logps/chosen": -7.757773399353027, |
| "logps/rejected": -26.798311233520508, |
| "loss": 0.9266724586486816, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5693855881690979, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.0821087509393692, |
| "rewards/margins": 1.3789818286895752, |
| "rewards/rejected": -1.4610905647277832, |
| "step": 130, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.6357294510160753, |
| "grad_norm": 3.8205740451812744, |
| "learning_rate": 0.00018601693673727933, |
| "logits/chosen": -0.40578073263168335, |
| "logits/rejected": -0.45195069909095764, |
| "logps/chosen": -12.803248405456543, |
| "logps/rejected": -31.195695877075195, |
| "loss": 0.9055405855178833, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6313671469688416, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.27764347195625305, |
| "rewards/margins": 1.7698627710342407, |
| "rewards/rejected": -1.492219090461731, |
| "step": 131, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.640582347588717, |
| "grad_norm": 4.257453918457031, |
| "learning_rate": 0.00018574275231729418, |
| "logits/chosen": -0.3832516670227051, |
| "logits/rejected": -0.40236127376556396, |
| "logps/chosen": -10.488710403442383, |
| "logps/rejected": -29.920001983642578, |
| "loss": 0.9821866154670715, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6411001682281494, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.07582154870033264, |
| "rewards/margins": 1.663382649421692, |
| "rewards/rejected": -1.7392042875289917, |
| "step": 132, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.6454352441613588, |
| "grad_norm": 5.575984954833984, |
| "learning_rate": 0.00018546611194501594, |
| "logits/chosen": -0.3834778666496277, |
| "logits/rejected": -0.4196530878543854, |
| "logps/chosen": -8.930424690246582, |
| "logps/rejected": -34.139522552490234, |
| "loss": 0.9412792921066284, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5637305974960327, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.2074286788702011, |
| "rewards/margins": 1.6988741159439087, |
| "rewards/rejected": -1.9063026905059814, |
| "step": 133, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.6502881407340007, |
| "grad_norm": 2.6553735733032227, |
| "learning_rate": 0.00018518702354432774, |
| "logits/chosen": -0.302835613489151, |
| "logits/rejected": -0.4271845519542694, |
| "logps/chosen": -6.828452110290527, |
| "logps/rejected": -32.71015167236328, |
| "loss": 0.8524077534675598, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5101207494735718, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.007050292566418648, |
| "rewards/margins": 1.6347914934158325, |
| "rewards/rejected": -1.6277412176132202, |
| "step": 134, |
| "train_speed(iter/s)": 0.011204 |
| }, |
| { |
| "epoch": 0.6551410373066424, |
| "grad_norm": 2.3347177505493164, |
| "learning_rate": 0.0001849054951092324, |
| "logits/chosen": -0.4007658362388611, |
| "logits/rejected": -0.4008050560951233, |
| "logps/chosen": -10.33335018157959, |
| "logps/rejected": -23.038896560668945, |
| "loss": 0.8887284398078918, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5501910448074341, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2812037765979767, |
| "rewards/margins": 1.3908374309539795, |
| "rewards/rejected": -1.1096336841583252, |
| "step": 135, |
| "train_speed(iter/s)": 0.011205 |
| }, |
| { |
| "epoch": 0.6599939338792842, |
| "grad_norm": 2.9783730506896973, |
| "learning_rate": 0.00018462153470362322, |
| "logits/chosen": -0.3893599808216095, |
| "logits/rejected": -0.44760093092918396, |
| "logps/chosen": -7.99445915222168, |
| "logps/rejected": -26.292232513427734, |
| "loss": 0.9434449672698975, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5910047292709351, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.11856328696012497, |
| "rewards/margins": 1.402287244796753, |
| "rewards/rejected": -1.2837239503860474, |
| "step": 136, |
| "train_speed(iter/s)": 0.011205 |
| }, |
| { |
| "epoch": 0.6648468304519259, |
| "grad_norm": 2.7139804363250732, |
| "learning_rate": 0.00018433515046105306, |
| "logits/chosen": -0.4038199782371521, |
| "logits/rejected": -0.3786419630050659, |
| "logps/chosen": -11.153128623962402, |
| "logps/rejected": -27.12870216369629, |
| "loss": 0.8474470376968384, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5868622660636902, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18890687823295593, |
| "rewards/margins": 1.7894312143325806, |
| "rewards/rejected": -1.6005244255065918, |
| "step": 137, |
| "train_speed(iter/s)": 0.011205 |
| }, |
| { |
| "epoch": 0.6696997270245678, |
| "grad_norm": 2.877267599105835, |
| "learning_rate": 0.0001840463505845014, |
| "logits/chosen": -0.30394551157951355, |
| "logits/rejected": -0.42835158109664917, |
| "logps/chosen": -7.929500102996826, |
| "logps/rejected": -30.011865615844727, |
| "loss": 0.806150496006012, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.49172037839889526, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15592113137245178, |
| "rewards/margins": 1.6279278993606567, |
| "rewards/rejected": -1.4720066785812378, |
| "step": 138, |
| "train_speed(iter/s)": 0.011205 |
| }, |
| { |
| "epoch": 0.6745526235972096, |
| "grad_norm": 2.5812747478485107, |
| "learning_rate": 0.0001837551433461393, |
| "logits/chosen": -0.3570711016654968, |
| "logits/rejected": -0.40324440598487854, |
| "logps/chosen": -9.332073211669922, |
| "logps/rejected": -23.627119064331055, |
| "loss": 0.7940301299095154, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.480025053024292, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.24301281571388245, |
| "rewards/margins": 1.4770238399505615, |
| "rewards/rejected": -1.234011173248291, |
| "step": 139, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.6794055201698513, |
| "grad_norm": 9.010900497436523, |
| "learning_rate": 0.00018346153708709268, |
| "logits/chosen": -0.36091697216033936, |
| "logits/rejected": -0.36146706342697144, |
| "logps/chosen": -11.872391700744629, |
| "logps/rejected": -22.261938095092773, |
| "loss": 1.231181263923645, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7421593070030212, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.055155493319034576, |
| "rewards/margins": 0.9281628131866455, |
| "rewards/rejected": -0.8730072975158691, |
| "step": 140, |
| "train_speed(iter/s)": 0.011205 |
| }, |
| { |
| "epoch": 0.6842584167424932, |
| "grad_norm": 2.032696485519409, |
| "learning_rate": 0.00018316554021720306, |
| "logits/chosen": -0.32833001017570496, |
| "logits/rejected": -0.44503307342529297, |
| "logps/chosen": -7.086223602294922, |
| "logps/rejected": -27.835424423217773, |
| "loss": 0.8454415202140808, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.49735450744628906, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.16136682033538818, |
| "rewards/margins": 1.4422115087509155, |
| "rewards/rejected": -1.2808446884155273, |
| "step": 141, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.689111313315135, |
| "grad_norm": 8.792712211608887, |
| "learning_rate": 0.00018286716121478693, |
| "logits/chosen": -0.32804882526397705, |
| "logits/rejected": -0.4235531687736511, |
| "logps/chosen": -7.5324883460998535, |
| "logps/rejected": -31.770389556884766, |
| "loss": 0.6662007570266724, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.43374398350715637, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18283605575561523, |
| "rewards/margins": 2.050058364868164, |
| "rewards/rejected": -1.8672223091125488, |
| "step": 142, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.6939642098877767, |
| "grad_norm": 3.7807235717773438, |
| "learning_rate": 0.00018256640862639287, |
| "logits/chosen": -0.31429523229599, |
| "logits/rejected": -0.42787081003189087, |
| "logps/chosen": -7.7822160720825195, |
| "logps/rejected": -30.884796142578125, |
| "loss": 0.823533296585083, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.556084394454956, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.06609532237052917, |
| "rewards/margins": 1.6945431232452393, |
| "rewards/rejected": -1.6284478902816772, |
| "step": 143, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.6988171064604186, |
| "grad_norm": 2.3116719722747803, |
| "learning_rate": 0.00018226329106655668, |
| "logits/chosen": -0.4266669750213623, |
| "logits/rejected": -0.4430805444717407, |
| "logps/chosen": -9.213752746582031, |
| "logps/rejected": -29.852783203125, |
| "loss": 0.671653687953949, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.434116929769516, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.08867526799440384, |
| "rewards/margins": 1.9470399618148804, |
| "rewards/rejected": -1.8583647012710571, |
| "step": 144, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.7036700030330604, |
| "grad_norm": 3.421821355819702, |
| "learning_rate": 0.00018195781721755465, |
| "logits/chosen": -0.44033709168434143, |
| "logits/rejected": -0.4357948899269104, |
| "logps/chosen": -9.914063453674316, |
| "logps/rejected": -33.468997955322266, |
| "loss": 0.911195695400238, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5890682935714722, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.049831733107566833, |
| "rewards/margins": 2.241945505142212, |
| "rewards/rejected": -2.1921138763427734, |
| "step": 145, |
| "train_speed(iter/s)": 0.011207 |
| }, |
| { |
| "epoch": 0.7085228996057021, |
| "grad_norm": 2.5243051052093506, |
| "learning_rate": 0.00018164999582915488, |
| "logits/chosen": -0.3629850447177887, |
| "logits/rejected": -0.4418475925922394, |
| "logps/chosen": -12.760583877563477, |
| "logps/rejected": -46.22138595581055, |
| "loss": 0.8992067575454712, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.685957133769989, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3213002681732178, |
| "rewards/margins": 3.270735502243042, |
| "rewards/rejected": -2.9494354724884033, |
| "step": 146, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.7133757961783439, |
| "grad_norm": 4.871396064758301, |
| "learning_rate": 0.00018133983571836667, |
| "logits/chosen": -0.33027127385139465, |
| "logits/rejected": -0.4520616829395294, |
| "logps/chosen": -10.541389465332031, |
| "logps/rejected": -41.49027633666992, |
| "loss": 1.0133702754974365, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6680970788002014, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.13710376620292664, |
| "rewards/margins": 2.4736220836639404, |
| "rewards/rejected": -2.3365182876586914, |
| "step": 147, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.7182286927509858, |
| "grad_norm": 3.6936795711517334, |
| "learning_rate": 0.000181027345769188, |
| "logits/chosen": -0.41163426637649536, |
| "logits/rejected": -0.40933167934417725, |
| "logps/chosen": -9.165144920349121, |
| "logps/rejected": -30.310758590698242, |
| "loss": 0.7363777160644531, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.42432036995887756, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.02400527521967888, |
| "rewards/margins": 1.7538728713989258, |
| "rewards/rejected": -1.7778780460357666, |
| "step": 148, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.7230815893236275, |
| "grad_norm": 6.49341344833374, |
| "learning_rate": 0.00018071253493235094, |
| "logits/chosen": -0.4786967933177948, |
| "logits/rejected": -0.4735773801803589, |
| "logps/chosen": -12.567697525024414, |
| "logps/rejected": -31.06718635559082, |
| "loss": 1.0480177402496338, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7017223834991455, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.04890735074877739, |
| "rewards/margins": 1.8125377893447876, |
| "rewards/rejected": -1.8614450693130493, |
| "step": 149, |
| "train_speed(iter/s)": 0.011206 |
| }, |
| { |
| "epoch": 0.7279344858962693, |
| "grad_norm": 11.014755249023438, |
| "learning_rate": 0.0001803954122250654, |
| "logits/chosen": -0.493363618850708, |
| "logits/rejected": -0.560202419757843, |
| "logps/chosen": -11.512059211730957, |
| "logps/rejected": -37.07139587402344, |
| "loss": 1.237532138824463, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7943273782730103, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.1521718055009842, |
| "rewards/margins": 2.0435235500335693, |
| "rewards/rejected": -2.195695161819458, |
| "step": 150, |
| "train_speed(iter/s)": 0.011207 |
| }, |
| { |
| "epoch": 0.7327873824689112, |
| "grad_norm": 4.69052791595459, |
| "learning_rate": 0.00018007598673076093, |
| "logits/chosen": -0.4235774278640747, |
| "logits/rejected": -0.4463633596897125, |
| "logps/chosen": -11.031900405883789, |
| "logps/rejected": -27.883403778076172, |
| "loss": 1.107750415802002, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.7605136632919312, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.05042722821235657, |
| "rewards/margins": 1.6558500528335571, |
| "rewards/rejected": -1.7062770128250122, |
| "step": 151, |
| "train_speed(iter/s)": 0.011207 |
| }, |
| { |
| "epoch": 0.7376402790415529, |
| "grad_norm": 2.4655253887176514, |
| "learning_rate": 0.00017975426759882614, |
| "logits/chosen": -0.44894903898239136, |
| "logits/rejected": -0.4738643765449524, |
| "logps/chosen": -9.968032836914062, |
| "logps/rejected": -29.113271713256836, |
| "loss": 0.7568652629852295, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5070943236351013, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.34214267134666443, |
| "rewards/margins": 1.9045073986053467, |
| "rewards/rejected": -1.5623646974563599, |
| "step": 152, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7424931756141947, |
| "grad_norm": 3.0639827251434326, |
| "learning_rate": 0.00017943026404434719, |
| "logits/chosen": -0.45172956585884094, |
| "logits/rejected": -0.5009413957595825, |
| "logps/chosen": -9.160325050354004, |
| "logps/rejected": -31.873594284057617, |
| "loss": 0.9305005669593811, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.6987206935882568, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.11259810626506805, |
| "rewards/margins": 1.9045584201812744, |
| "rewards/rejected": -1.7919602394104004, |
| "step": 153, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7473460721868365, |
| "grad_norm": 3.0101685523986816, |
| "learning_rate": 0.00017910398534784337, |
| "logits/chosen": -0.3823431730270386, |
| "logits/rejected": -0.47442248463630676, |
| "logps/chosen": -10.423698425292969, |
| "logps/rejected": -23.59076690673828, |
| "loss": 1.046834945678711, |
| "memory(GiB)": 40.31, |
| "nll_loss": 0.5919281244277954, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.2845015525817871, |
| "rewards/margins": 1.1300426721572876, |
| "rewards/rejected": -0.8455410599708557, |
| "step": 154, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7521989687594783, |
| "grad_norm": 6.684966087341309, |
| "learning_rate": 0.00017877544085500156, |
| "logits/chosen": -0.32389581203460693, |
| "logits/rejected": -0.4660918116569519, |
| "logps/chosen": -5.926966190338135, |
| "logps/rejected": -32.869140625, |
| "loss": 0.8766897916793823, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5428420305252075, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.006232857704162598, |
| "rewards/margins": 1.7234890460968018, |
| "rewards/rejected": -1.7172563076019287, |
| "step": 155, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7570518653321201, |
| "grad_norm": 3.565051317214966, |
| "learning_rate": 0.00017844463997640842, |
| "logits/chosen": -0.40359729528427124, |
| "logits/rejected": -0.44952863454818726, |
| "logps/chosen": -11.961484909057617, |
| "logps/rejected": -28.9517879486084, |
| "loss": 1.0656830072402954, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6759463548660278, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06430176645517349, |
| "rewards/margins": 1.5096461772918701, |
| "rewards/rejected": -1.44534432888031, |
| "step": 156, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7619047619047619, |
| "grad_norm": 2.5383567810058594, |
| "learning_rate": 0.00017811159218728082, |
| "logits/chosen": -0.39084097743034363, |
| "logits/rejected": -0.4242747128009796, |
| "logps/chosen": -8.804425239562988, |
| "logps/rejected": -22.249990463256836, |
| "loss": 0.9147839546203613, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.590728759765625, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2870722711086273, |
| "rewards/margins": 1.4324426651000977, |
| "rewards/rejected": -1.1453704833984375, |
| "step": 157, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7667576584774037, |
| "grad_norm": 2.662156343460083, |
| "learning_rate": 0.00017777630702719453, |
| "logits/chosen": -0.32536131143569946, |
| "logits/rejected": -0.397830992937088, |
| "logps/chosen": -9.738006591796875, |
| "logps/rejected": -30.069774627685547, |
| "loss": 0.8968651294708252, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.527851402759552, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08923595398664474, |
| "rewards/margins": 1.6620792150497437, |
| "rewards/rejected": -1.5728431940078735, |
| "step": 158, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7716105550500455, |
| "grad_norm": 3.2361485958099365, |
| "learning_rate": 0.00017743879409981095, |
| "logits/chosen": -0.3466249704360962, |
| "logits/rejected": -0.41640138626098633, |
| "logps/chosen": -11.293127059936523, |
| "logps/rejected": -30.880006790161133, |
| "loss": 0.9047757387161255, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5920196175575256, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.030911901965737343, |
| "rewards/margins": 1.6093332767486572, |
| "rewards/rejected": -1.5784213542938232, |
| "step": 159, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7764634516226873, |
| "grad_norm": 2.2754669189453125, |
| "learning_rate": 0.00017709906307260194, |
| "logits/chosen": -0.3480105400085449, |
| "logits/rejected": -0.4323396682739258, |
| "logps/chosen": -5.929109573364258, |
| "logps/rejected": -34.05678939819336, |
| "loss": 0.7185795903205872, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4936397075653076, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.06835521012544632, |
| "rewards/margins": 2.1114068031311035, |
| "rewards/rejected": -2.0430517196655273, |
| "step": 160, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.7813163481953291, |
| "grad_norm": 2.226016044616699, |
| "learning_rate": 0.000176757123676573, |
| "logits/chosen": -0.39186131954193115, |
| "logits/rejected": -0.3231448531150818, |
| "logps/chosen": -12.739521026611328, |
| "logps/rejected": -28.13015365600586, |
| "loss": 0.8966748714447021, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5328811407089233, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.17805510759353638, |
| "rewards/margins": 1.8448935747146606, |
| "rewards/rejected": -1.6668384075164795, |
| "step": 161, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7861692447679709, |
| "grad_norm": 2.2700612545013428, |
| "learning_rate": 0.00017641298570598459, |
| "logits/chosen": -0.36923766136169434, |
| "logits/rejected": -0.41111743450164795, |
| "logps/chosen": -6.442410469055176, |
| "logps/rejected": -29.95729637145996, |
| "loss": 0.7110913991928101, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4247634708881378, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08174540847539902, |
| "rewards/margins": 1.7869547605514526, |
| "rewards/rejected": -1.7052093744277954, |
| "step": 162, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7910221413406127, |
| "grad_norm": 5.662917613983154, |
| "learning_rate": 0.0001760666590180714, |
| "logits/chosen": -0.2990395426750183, |
| "logits/rejected": -0.4443129301071167, |
| "logps/chosen": -11.827997207641602, |
| "logps/rejected": -39.40394592285156, |
| "loss": 0.9818216562271118, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6332793831825256, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.07979752123355865, |
| "rewards/margins": 2.349473237991333, |
| "rewards/rejected": -2.4292705059051514, |
| "step": 163, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.7958750379132544, |
| "grad_norm": 2.659315586090088, |
| "learning_rate": 0.0001757181535327602, |
| "logits/chosen": -0.41728484630584717, |
| "logits/rejected": -0.44338852167129517, |
| "logps/chosen": -7.938527584075928, |
| "logps/rejected": -30.74227523803711, |
| "loss": 0.8022794723510742, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4814991056919098, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.056870464235544205, |
| "rewards/margins": 1.9418084621429443, |
| "rewards/rejected": -1.8849378824234009, |
| "step": 164, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.8007279344858963, |
| "grad_norm": 2.9071202278137207, |
| "learning_rate": 0.00017536747923238566, |
| "logits/chosen": -0.3409942090511322, |
| "logits/rejected": -0.457685649394989, |
| "logps/chosen": -7.500731468200684, |
| "logps/rejected": -44.705909729003906, |
| "loss": 0.6390520930290222, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.38755032420158386, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.19773569703102112, |
| "rewards/margins": 3.0309081077575684, |
| "rewards/rejected": -2.83317232131958, |
| "step": 165, |
| "train_speed(iter/s)": 0.011208 |
| }, |
| { |
| "epoch": 0.8055808310585381, |
| "grad_norm": 3.0054497718811035, |
| "learning_rate": 0.00017501464616140436, |
| "logits/chosen": -0.41681885719299316, |
| "logits/rejected": -0.45619145035743713, |
| "logps/chosen": -8.748014450073242, |
| "logps/rejected": -35.70191955566406, |
| "loss": 0.7890005111694336, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5446341633796692, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16440145671367645, |
| "rewards/margins": 2.74184250831604, |
| "rewards/rejected": -2.5774412155151367, |
| "step": 166, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8104337276311798, |
| "grad_norm": 3.300097942352295, |
| "learning_rate": 0.000174659664426107, |
| "logits/chosen": -0.4365924298763275, |
| "logits/rejected": -0.45622488856315613, |
| "logps/chosen": -11.21609115600586, |
| "logps/rejected": -34.401222229003906, |
| "loss": 0.9579750299453735, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.7187469601631165, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.14886964857578278, |
| "rewards/margins": 2.472686529159546, |
| "rewards/rejected": -2.323817253112793, |
| "step": 167, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8152866242038217, |
| "grad_norm": 2.5296413898468018, |
| "learning_rate": 0.00017430254419432931, |
| "logits/chosen": -0.41913431882858276, |
| "logits/rejected": -0.41736355423927307, |
| "logps/chosen": -13.593805313110352, |
| "logps/rejected": -33.63285827636719, |
| "loss": 0.6976242065429688, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.46450674533843994, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2216261476278305, |
| "rewards/margins": 2.4022936820983887, |
| "rewards/rejected": -2.1806676387786865, |
| "step": 168, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8201395207764635, |
| "grad_norm": 21.119770050048828, |
| "learning_rate": 0.00017394329569516028, |
| "logits/chosen": -0.3196732997894287, |
| "logits/rejected": -0.47898930311203003, |
| "logps/chosen": -11.645589828491211, |
| "logps/rejected": -45.33073425292969, |
| "loss": 1.067216157913208, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.8613059520721436, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.024522848427295685, |
| "rewards/margins": 2.9725117683410645, |
| "rewards/rejected": -2.947988748550415, |
| "step": 169, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8249924173491052, |
| "grad_norm": 2.541898012161255, |
| "learning_rate": 0.00017358192921864953, |
| "logits/chosen": -0.40079769492149353, |
| "logits/rejected": -0.4508993923664093, |
| "logps/chosen": -10.424609184265137, |
| "logps/rejected": -34.24469757080078, |
| "loss": 0.7411442399024963, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4615729749202728, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.05271195247769356, |
| "rewards/margins": 2.12577223777771, |
| "rewards/rejected": -2.0730605125427246, |
| "step": 170, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8298453139217471, |
| "grad_norm": 4.101256370544434, |
| "learning_rate": 0.00017321845511551244, |
| "logits/chosen": -0.38883864879608154, |
| "logits/rejected": -0.46420595049858093, |
| "logps/chosen": -9.728652954101562, |
| "logps/rejected": -49.25832748413086, |
| "loss": 0.7956501245498657, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6269941926002502, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.07354284077882767, |
| "rewards/margins": 3.7279632091522217, |
| "rewards/rejected": -3.6544198989868164, |
| "step": 171, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8346982104943889, |
| "grad_norm": 6.031051158905029, |
| "learning_rate": 0.00017285288379683376, |
| "logits/chosen": -0.4501774311065674, |
| "logits/rejected": -0.49218156933784485, |
| "logps/chosen": -12.754474639892578, |
| "logps/rejected": -35.773963928222656, |
| "loss": 0.9267921447753906, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6455193758010864, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.23461803793907166, |
| "rewards/margins": 2.5618717670440674, |
| "rewards/rejected": -2.327253580093384, |
| "step": 172, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8395511070670306, |
| "grad_norm": 2.7078492641448975, |
| "learning_rate": 0.00017248522573376932, |
| "logits/chosen": -0.5002922415733337, |
| "logits/rejected": -0.4927571713924408, |
| "logps/chosen": -12.274182319641113, |
| "logps/rejected": -38.476226806640625, |
| "loss": 0.8795382976531982, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.679415225982666, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.20902375876903534, |
| "rewards/margins": 2.886625289916992, |
| "rewards/rejected": -2.6776018142700195, |
| "step": 173, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8444040036396724, |
| "grad_norm": 3.0864455699920654, |
| "learning_rate": 0.00017211549145724602, |
| "logits/chosen": -0.44441068172454834, |
| "logits/rejected": -0.5031411647796631, |
| "logps/chosen": -6.964417934417725, |
| "logps/rejected": -47.59633255004883, |
| "loss": 0.5743454694747925, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.39823874831199646, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.1409493237733841, |
| "rewards/margins": 3.664174795150757, |
| "rewards/rejected": -3.5232253074645996, |
| "step": 174, |
| "train_speed(iter/s)": 0.011209 |
| }, |
| { |
| "epoch": 0.8492569002123143, |
| "grad_norm": 13.482826232910156, |
| "learning_rate": 0.00017174369155766037, |
| "logits/chosen": -0.44830185174942017, |
| "logits/rejected": -0.4931170344352722, |
| "logps/chosen": -11.00506591796875, |
| "logps/rejected": -38.13721466064453, |
| "loss": 1.102919578552246, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.8681702017784119, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.020436841994524002, |
| "rewards/margins": 2.716992139816284, |
| "rewards/rejected": -2.6965548992156982, |
| "step": 175, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.854109796784956, |
| "grad_norm": 2.788137435913086, |
| "learning_rate": 0.0001713698366845751, |
| "logits/chosen": -0.41955456137657166, |
| "logits/rejected": -0.4849671423435211, |
| "logps/chosen": -10.713014602661133, |
| "logps/rejected": -38.87740707397461, |
| "loss": 0.8198617696762085, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5618011951446533, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.052750274538993835, |
| "rewards/margins": 2.4263105392456055, |
| "rewards/rejected": -2.3735604286193848, |
| "step": 176, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.8589626933575978, |
| "grad_norm": 2.8578059673309326, |
| "learning_rate": 0.00017099393754641402, |
| "logits/chosen": -0.4903329014778137, |
| "logits/rejected": -0.44835591316223145, |
| "logps/chosen": -12.698493957519531, |
| "logps/rejected": -31.339452743530273, |
| "loss": 0.8736453056335449, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5660265684127808, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12503781914710999, |
| "rewards/margins": 2.115137815475464, |
| "rewards/rejected": -1.9901000261306763, |
| "step": 177, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.8638155899302397, |
| "grad_norm": 14.433143615722656, |
| "learning_rate": 0.0001706160049101554, |
| "logits/chosen": -0.39836981892585754, |
| "logits/rejected": -0.4502709209918976, |
| "logps/chosen": -12.003418922424316, |
| "logps/rejected": -33.39105224609375, |
| "loss": 0.9455697536468506, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6791843771934509, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2257295846939087, |
| "rewards/margins": 2.307054042816162, |
| "rewards/rejected": -2.081324577331543, |
| "step": 178, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.8686684865028814, |
| "grad_norm": 4.972922325134277, |
| "learning_rate": 0.00017023604960102355, |
| "logits/chosen": -0.48707541823387146, |
| "logits/rejected": -0.43483617901802063, |
| "logps/chosen": -15.320083618164062, |
| "logps/rejected": -26.741153717041016, |
| "loss": 1.0946437120437622, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.7049025893211365, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.14932136237621307, |
| "rewards/margins": 1.6685651540756226, |
| "rewards/rejected": -1.519243836402893, |
| "step": 179, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.8735213830755232, |
| "grad_norm": 7.616319179534912, |
| "learning_rate": 0.00016985408250217864, |
| "logits/chosen": -0.48446008563041687, |
| "logits/rejected": -0.5007644891738892, |
| "logps/chosen": -9.878619194030762, |
| "logps/rejected": -33.84723663330078, |
| "loss": 0.91806560754776, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.588477373123169, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13273106515407562, |
| "rewards/margins": 2.1440441608428955, |
| "rewards/rejected": -2.0113131999969482, |
| "step": 180, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.878374279648165, |
| "grad_norm": 5.360206604003906, |
| "learning_rate": 0.00016947011455440523, |
| "logits/chosen": -0.45152074098587036, |
| "logits/rejected": -0.4405006766319275, |
| "logps/chosen": -12.04659652709961, |
| "logps/rejected": -29.337995529174805, |
| "loss": 0.9476042985916138, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5956300497055054, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.007146604359149933, |
| "rewards/margins": 1.6998629570007324, |
| "rewards/rejected": -1.6927162408828735, |
| "step": 181, |
| "train_speed(iter/s)": 0.01121 |
| }, |
| { |
| "epoch": 0.8832271762208068, |
| "grad_norm": 2.752962589263916, |
| "learning_rate": 0.00016908415675579854, |
| "logits/chosen": -0.400238573551178, |
| "logits/rejected": -0.5106572508811951, |
| "logps/chosen": -5.792488098144531, |
| "logps/rejected": -31.905841827392578, |
| "loss": 0.7257025241851807, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4335445463657379, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.0758802518248558, |
| "rewards/margins": 1.8862504959106445, |
| "rewards/rejected": -1.8103703260421753, |
| "step": 182, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.8880800727934486, |
| "grad_norm": 2.3120791912078857, |
| "learning_rate": 0.00016869622016144973, |
| "logits/chosen": -0.5098201632499695, |
| "logits/rejected": -0.4519084095954895, |
| "logps/chosen": -8.241873741149902, |
| "logps/rejected": -25.190824508666992, |
| "loss": 0.7128751277923584, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.44536417722702026, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2868940234184265, |
| "rewards/margins": 1.7276583909988403, |
| "rewards/rejected": -1.4407645463943481, |
| "step": 183, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.8929329693660903, |
| "grad_norm": 2.231217861175537, |
| "learning_rate": 0.0001683063158831291, |
| "logits/chosen": -0.4472496509552002, |
| "logits/rejected": -0.4929869771003723, |
| "logps/chosen": -7.548064231872559, |
| "logps/rejected": -31.63290023803711, |
| "loss": 0.7525639533996582, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4533500671386719, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2907182276248932, |
| "rewards/margins": 2.194798231124878, |
| "rewards/rejected": -1.9040801525115967, |
| "step": 184, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.8977858659387322, |
| "grad_norm": 2.4883766174316406, |
| "learning_rate": 0.00016791445508896783, |
| "logits/chosen": -0.36559921503067017, |
| "logits/rejected": -0.42163488268852234, |
| "logps/chosen": -11.388641357421875, |
| "logps/rejected": -32.0931510925293, |
| "loss": 0.7775099873542786, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5461112856864929, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22250865399837494, |
| "rewards/margins": 1.9723868370056152, |
| "rewards/rejected": -1.749878168106079, |
| "step": 185, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.902638762511374, |
| "grad_norm": 3.259838819503784, |
| "learning_rate": 0.00016752064900313815, |
| "logits/chosen": -0.5012748837471008, |
| "logits/rejected": -0.4815691113471985, |
| "logps/chosen": -8.668580055236816, |
| "logps/rejected": -27.685821533203125, |
| "loss": 0.7447406649589539, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.46015244722366333, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.07047421485185623, |
| "rewards/margins": 1.7800168991088867, |
| "rewards/rejected": -1.7095428705215454, |
| "step": 186, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9074916590840157, |
| "grad_norm": 4.5533576011657715, |
| "learning_rate": 0.0001671249089055317, |
| "logits/chosen": -0.5065349340438843, |
| "logits/rejected": -0.4818389117717743, |
| "logps/chosen": -9.181870460510254, |
| "logps/rejected": -33.568580627441406, |
| "loss": 0.7996796369552612, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5168719291687012, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.10808134078979492, |
| "rewards/margins": 2.115391731262207, |
| "rewards/rejected": -2.007310628890991, |
| "step": 187, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9123445556566576, |
| "grad_norm": 4.431750297546387, |
| "learning_rate": 0.0001667272461314366, |
| "logits/chosen": -0.48659980297088623, |
| "logits/rejected": -0.487091988325119, |
| "logps/chosen": -10.220385551452637, |
| "logps/rejected": -35.60252380371094, |
| "loss": 0.9328407049179077, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6412612199783325, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.01909228041768074, |
| "rewards/margins": 2.4949698448181152, |
| "rewards/rejected": -2.475877285003662, |
| "step": 188, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.9171974522292994, |
| "grad_norm": 8.757882118225098, |
| "learning_rate": 0.00016632767207121267, |
| "logits/chosen": -0.33902058005332947, |
| "logits/rejected": -0.4704265594482422, |
| "logps/chosen": -10.807303428649902, |
| "logps/rejected": -48.82100296020508, |
| "loss": 1.2881890535354614, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.8564696907997131, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.06860647350549698, |
| "rewards/margins": 2.6622214317321777, |
| "rewards/rejected": -2.7308273315429688, |
| "step": 189, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9220503488019411, |
| "grad_norm": 9.251166343688965, |
| "learning_rate": 0.00016592619816996527, |
| "logits/chosen": -0.47989535331726074, |
| "logits/rejected": -0.5115556716918945, |
| "logps/chosen": -15.828009605407715, |
| "logps/rejected": -35.08710861206055, |
| "loss": 0.9385941624641418, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.632335901260376, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.0631108283996582, |
| "rewards/margins": 1.8624895811080933, |
| "rewards/rejected": -1.799378752708435, |
| "step": 190, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9269032453745829, |
| "grad_norm": 4.995007038116455, |
| "learning_rate": 0.0001655228359272173, |
| "logits/chosen": -0.45928898453712463, |
| "logits/rejected": -0.42803335189819336, |
| "logps/chosen": -11.98594856262207, |
| "logps/rejected": -27.537172317504883, |
| "loss": 1.0162464380264282, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6742801070213318, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.0864061713218689, |
| "rewards/margins": 1.6110825538635254, |
| "rewards/rejected": -1.5246765613555908, |
| "step": 191, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9317561419472248, |
| "grad_norm": 8.054399490356445, |
| "learning_rate": 0.00016511759689657999, |
| "logits/chosen": -0.3773999512195587, |
| "logits/rejected": -0.42789533734321594, |
| "logps/chosen": -12.406736373901367, |
| "logps/rejected": -39.963958740234375, |
| "loss": 0.7866171598434448, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5565248131752014, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.005689222365617752, |
| "rewards/margins": 2.343252658843994, |
| "rewards/rejected": -2.3489420413970947, |
| "step": 192, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9366090385198665, |
| "grad_norm": 3.3172736167907715, |
| "learning_rate": 0.00016471049268542188, |
| "logits/chosen": -0.46051445603370667, |
| "logits/rejected": -0.41965189576148987, |
| "logps/chosen": -10.034582138061523, |
| "logps/rejected": -36.16367721557617, |
| "loss": 0.7486264109611511, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4820393919944763, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.0357435941696167, |
| "rewards/margins": 2.641000747680664, |
| "rewards/rejected": -2.605257034301758, |
| "step": 193, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9414619350925083, |
| "grad_norm": 2.3992457389831543, |
| "learning_rate": 0.00016430153495453642, |
| "logits/chosen": -0.5328064560890198, |
| "logits/rejected": -0.47638142108917236, |
| "logps/chosen": -11.91832160949707, |
| "logps/rejected": -36.306785583496094, |
| "loss": 0.8761617541313171, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.6264892220497131, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1922227442264557, |
| "rewards/margins": 2.7257025241851807, |
| "rewards/rejected": -2.533479690551758, |
| "step": 194, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9463148316651502, |
| "grad_norm": 4.527428150177002, |
| "learning_rate": 0.00016389073541780782, |
| "logits/chosen": -0.46726107597351074, |
| "logits/rejected": -0.47500157356262207, |
| "logps/chosen": -8.38770580291748, |
| "logps/rejected": -29.87854766845703, |
| "loss": 0.7214972376823425, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4556006193161011, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08811609447002411, |
| "rewards/margins": 1.9756295680999756, |
| "rewards/rejected": -1.8875133991241455, |
| "step": 195, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9511677282377919, |
| "grad_norm": 3.743474006652832, |
| "learning_rate": 0.00016347810584187568, |
| "logits/chosen": -0.5369599461555481, |
| "logits/rejected": -0.5066350102424622, |
| "logps/chosen": -8.895236015319824, |
| "logps/rejected": -26.281404495239258, |
| "loss": 0.8112626075744629, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5115976333618164, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.04646964743733406, |
| "rewards/margins": 1.599694848060608, |
| "rewards/rejected": -1.5532252788543701, |
| "step": 196, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.9560206248104337, |
| "grad_norm": 2.772148370742798, |
| "learning_rate": 0.00016306365804579794, |
| "logits/chosen": -0.36604201793670654, |
| "logits/rejected": -0.4730686545372009, |
| "logps/chosen": -9.536896705627441, |
| "logps/rejected": -40.390995025634766, |
| "loss": 0.6381856799125671, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.4115173816680908, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2083117514848709, |
| "rewards/margins": 2.512366771697998, |
| "rewards/rejected": -2.3040552139282227, |
| "step": 197, |
| "train_speed(iter/s)": 0.011211 |
| }, |
| { |
| "epoch": 0.9608735213830755, |
| "grad_norm": 3.5839767456054688, |
| "learning_rate": 0.0001626474039007122, |
| "logits/chosen": -0.4816967844963074, |
| "logits/rejected": -0.47818446159362793, |
| "logps/chosen": -7.637480735778809, |
| "logps/rejected": -29.050764083862305, |
| "loss": 0.6914876103401184, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.3912826478481293, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2960473895072937, |
| "rewards/margins": 2.151468515396118, |
| "rewards/rejected": -1.8554210662841797, |
| "step": 198, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.9657264179557173, |
| "grad_norm": 2.633289098739624, |
| "learning_rate": 0.00016222935532949587, |
| "logits/chosen": -0.4851294755935669, |
| "logits/rejected": -0.4861046373844147, |
| "logps/chosen": -10.165040969848633, |
| "logps/rejected": -34.33792495727539, |
| "loss": 0.7900617718696594, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.5045729279518127, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.33929818868637085, |
| "rewards/margins": 2.4269657135009766, |
| "rewards/rejected": -2.08766770362854, |
| "step": 199, |
| "train_speed(iter/s)": 0.011212 |
| }, |
| { |
| "epoch": 0.9705793145283591, |
| "grad_norm": 4.324284553527832, |
| "learning_rate": 0.0001618095243064245, |
| "logits/chosen": -0.486560583114624, |
| "logits/rejected": -0.46493133902549744, |
| "logps/chosen": -7.55566930770874, |
| "logps/rejected": -32.536399841308594, |
| "loss": 0.7669270038604736, |
| "memory(GiB)": 43.93, |
| "nll_loss": 0.45861828327178955, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0628332868218422, |
| "rewards/margins": 2.340352773666382, |
| "rewards/rejected": -2.277519464492798, |
| "step": 200, |
| "train_speed(iter/s)": 0.011212 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 618, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.080336546969682e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
