| { |
| "best_global_step": 300, |
| "best_metric": 0.79432702, |
| "best_model_checkpoint": "/data2/kdd/crag/cjz/output/task1+3_dpo_absgt10_noidk__pari0.8_0608/v0-20250611-010914/checkpoint-300", |
| "epoch": 1.563758389261745, |
| "eval_steps": 300, |
| "global_step": 350, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0044742729306487695, |
| "grad_norm": 4.521431922912598, |
| "learning_rate": 8.695652173913044e-06, |
| "logits/chosen": -0.37893733382225037, |
| "logits/rejected": -0.4597730040550232, |
| "logps/chosen": -11.001541137695312, |
| "logps/rejected": -17.05784034729004, |
| "loss": 1.3586950302124023, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6655479669570923, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.005374 |
| }, |
| { |
| "epoch": 0.008948545861297539, |
| "grad_norm": 3.470855951309204, |
| "learning_rate": 1.739130434782609e-05, |
| "logits/chosen": -0.3633067309856415, |
| "logits/rejected": -0.4392029941082001, |
| "logps/chosen": -7.973292350769043, |
| "logps/rejected": -13.90955924987793, |
| "loss": 1.2437297105789185, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5505824685096741, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.013422818791946308, |
| "grad_norm": 5.51271390914917, |
| "learning_rate": 2.608695652173913e-05, |
| "logits/chosen": -0.26578277349472046, |
| "logits/rejected": -0.45589232444763184, |
| "logps/chosen": -9.600592613220215, |
| "logps/rejected": -22.355058670043945, |
| "loss": 1.392338514328003, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6997992396354675, |
| "rewards/accuracies": 0.453125, |
| "rewards/chosen": 4.354381235316396e-05, |
| "rewards/margins": 0.0013819998130202293, |
| "rewards/rejected": -0.001338455593213439, |
| "step": 3, |
| "train_speed(iter/s)": 0.005521 |
| }, |
| { |
| "epoch": 0.017897091722595078, |
| "grad_norm": 6.5741682052612305, |
| "learning_rate": 3.478260869565218e-05, |
| "logits/chosen": -0.40403854846954346, |
| "logits/rejected": -0.5402908325195312, |
| "logps/chosen": -6.373080253601074, |
| "logps/rejected": -16.791393280029297, |
| "loss": 1.222740888595581, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5450983047485352, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.010452846065163612, |
| "rewards/margins": 0.031955912709236145, |
| "rewards/rejected": -0.021503068506717682, |
| "step": 4, |
| "train_speed(iter/s)": 0.005551 |
| }, |
| { |
| "epoch": 0.02237136465324385, |
| "grad_norm": 3.046576738357544, |
| "learning_rate": 4.347826086956522e-05, |
| "logits/chosen": -0.3187038004398346, |
| "logits/rejected": -0.4441537857055664, |
| "logps/chosen": -9.070188522338867, |
| "logps/rejected": -17.255901336669922, |
| "loss": 1.2276911735534668, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5655408501625061, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.007649715524166822, |
| "rewards/margins": 0.07260076701641083, |
| "rewards/rejected": -0.08025049418210983, |
| "step": 5, |
| "train_speed(iter/s)": 0.005557 |
| }, |
| { |
| "epoch": 0.026845637583892617, |
| "grad_norm": 4.846445560455322, |
| "learning_rate": 5.217391304347826e-05, |
| "logits/chosen": -0.30176016688346863, |
| "logits/rejected": -0.45638489723205566, |
| "logps/chosen": -8.63841438293457, |
| "logps/rejected": -18.97266960144043, |
| "loss": 1.1484646797180176, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5371410250663757, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.024880727753043175, |
| "rewards/margins": 0.23170879483222961, |
| "rewards/rejected": -0.25658950209617615, |
| "step": 6, |
| "train_speed(iter/s)": 0.005568 |
| }, |
| { |
| "epoch": 0.03131991051454139, |
| "grad_norm": 7.562114715576172, |
| "learning_rate": 6.086956521739131e-05, |
| "logits/chosen": -0.3937205672264099, |
| "logits/rejected": -0.48791196942329407, |
| "logps/chosen": -10.315498352050781, |
| "logps/rejected": -20.046728134155273, |
| "loss": 1.3093780279159546, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6873191595077515, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.14371810853481293, |
| "rewards/margins": 0.2797008454799652, |
| "rewards/rejected": -0.42341893911361694, |
| "step": 7, |
| "train_speed(iter/s)": 0.005577 |
| }, |
| { |
| "epoch": 0.035794183445190156, |
| "grad_norm": 7.459152698516846, |
| "learning_rate": 6.956521739130436e-05, |
| "logits/chosen": -0.3540106415748596, |
| "logits/rejected": -0.43836480379104614, |
| "logps/chosen": -8.758136749267578, |
| "logps/rejected": -18.294340133666992, |
| "loss": 1.3215807676315308, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6817622184753418, |
| "rewards/accuracies": 0.640625, |
| "rewards/chosen": -0.19053705036640167, |
| "rewards/margins": 0.2214164435863495, |
| "rewards/rejected": -0.41195347905158997, |
| "step": 8, |
| "train_speed(iter/s)": 0.005585 |
| }, |
| { |
| "epoch": 0.040268456375838924, |
| "grad_norm": 3.0581202507019043, |
| "learning_rate": 7.82608695652174e-05, |
| "logits/chosen": -0.3466319143772125, |
| "logits/rejected": -0.4212740361690521, |
| "logps/chosen": -10.387115478515625, |
| "logps/rejected": -20.363929748535156, |
| "loss": 1.1448516845703125, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5626436471939087, |
| "rewards/accuracies": 0.703125, |
| "rewards/chosen": -0.06380077451467514, |
| "rewards/margins": 0.3404861092567444, |
| "rewards/rejected": -0.40428686141967773, |
| "step": 9, |
| "train_speed(iter/s)": 0.005583 |
| }, |
| { |
| "epoch": 0.0447427293064877, |
| "grad_norm": 3.9353108406066895, |
| "learning_rate": 8.695652173913044e-05, |
| "logits/chosen": -0.4188399314880371, |
| "logits/rejected": -0.5007613301277161, |
| "logps/chosen": -9.096799850463867, |
| "logps/rejected": -18.05609130859375, |
| "loss": 1.2927519083023071, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6729815006256104, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.060015156865119934, |
| "rewards/margins": 0.20498159527778625, |
| "rewards/rejected": -0.264996737241745, |
| "step": 10, |
| "train_speed(iter/s)": 0.00559 |
| }, |
| { |
| "epoch": 0.049217002237136466, |
| "grad_norm": 2.2350685596466064, |
| "learning_rate": 9.565217391304348e-05, |
| "logits/chosen": -0.41371461749076843, |
| "logits/rejected": -0.4829203188419342, |
| "logps/chosen": -7.282920837402344, |
| "logps/rejected": -15.015157699584961, |
| "loss": 1.167080044746399, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5672391653060913, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.07586651295423508, |
| "rewards/margins": 0.2256288230419159, |
| "rewards/rejected": -0.14976228773593903, |
| "step": 11, |
| "train_speed(iter/s)": 0.005596 |
| }, |
| { |
| "epoch": 0.053691275167785234, |
| "grad_norm": 2.6733052730560303, |
| "learning_rate": 0.00010434782608695653, |
| "logits/chosen": -0.38637417554855347, |
| "logits/rejected": -0.45029905438423157, |
| "logps/chosen": -11.042367935180664, |
| "logps/rejected": -16.67047882080078, |
| "loss": 1.2763197422027588, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6531897783279419, |
| "rewards/accuracies": 0.671875, |
| "rewards/chosen": 0.06360332667827606, |
| "rewards/margins": 0.17586028575897217, |
| "rewards/rejected": -0.1122569665312767, |
| "step": 12, |
| "train_speed(iter/s)": 0.005598 |
| }, |
| { |
| "epoch": 0.058165548098434, |
| "grad_norm": 1.9472479820251465, |
| "learning_rate": 0.00011304347826086956, |
| "logits/chosen": -0.4195230007171631, |
| "logits/rejected": -0.4734322428703308, |
| "logps/chosen": -7.205465316772461, |
| "logps/rejected": -13.222125053405762, |
| "loss": 1.1376935243606567, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4900382459163666, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.018632210791110992, |
| "rewards/margins": 0.11014734208583832, |
| "rewards/rejected": -0.09151513874530792, |
| "step": 13, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 0.06263982102908278, |
| "grad_norm": 1.828997015953064, |
| "learning_rate": 0.00012173913043478263, |
| "logits/chosen": -0.399485319852829, |
| "logits/rejected": -0.4682197570800781, |
| "logps/chosen": -8.005407333374023, |
| "logps/rejected": -16.06272315979004, |
| "loss": 1.1697134971618652, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5445835590362549, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0030609495006501675, |
| "rewards/margins": 0.16438230872154236, |
| "rewards/rejected": -0.16132137179374695, |
| "step": 14, |
| "train_speed(iter/s)": 0.005604 |
| }, |
| { |
| "epoch": 0.06711409395973154, |
| "grad_norm": 2.1229512691497803, |
| "learning_rate": 0.00013043478260869567, |
| "logits/chosen": -0.3835911750793457, |
| "logits/rejected": -0.4592927396297455, |
| "logps/chosen": -6.499156475067139, |
| "logps/rejected": -15.707530975341797, |
| "loss": 1.0722893476486206, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.503006637096405, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06762155145406723, |
| "rewards/margins": 0.31041574478149414, |
| "rewards/rejected": -0.24279415607452393, |
| "step": 15, |
| "train_speed(iter/s)": 0.005608 |
| }, |
| { |
| "epoch": 0.07158836689038031, |
| "grad_norm": 4.802647113800049, |
| "learning_rate": 0.0001391304347826087, |
| "logits/chosen": -0.3306896388530731, |
| "logits/rejected": -0.41153809428215027, |
| "logps/chosen": -10.432010650634766, |
| "logps/rejected": -21.58324432373047, |
| "loss": 1.3572790622711182, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7762681245803833, |
| "rewards/accuracies": 0.671875, |
| "rewards/chosen": -0.06817664206027985, |
| "rewards/margins": 0.363118052482605, |
| "rewards/rejected": -0.4312947690486908, |
| "step": 16, |
| "train_speed(iter/s)": 0.005608 |
| }, |
| { |
| "epoch": 0.07606263982102908, |
| "grad_norm": 4.122814178466797, |
| "learning_rate": 0.00014782608695652173, |
| "logits/chosen": -0.32593876123428345, |
| "logits/rejected": -0.4176383316516876, |
| "logps/chosen": -9.068134307861328, |
| "logps/rejected": -21.460546493530273, |
| "loss": 1.1545155048370361, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5955549478530884, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0028565023094415665, |
| "rewards/margins": 0.3992985486984253, |
| "rewards/rejected": -0.3964419960975647, |
| "step": 17, |
| "train_speed(iter/s)": 0.00561 |
| }, |
| { |
| "epoch": 0.08053691275167785, |
| "grad_norm": 1.9508388042449951, |
| "learning_rate": 0.0001565217391304348, |
| "logits/chosen": -0.3363662660121918, |
| "logits/rejected": -0.40299132466316223, |
| "logps/chosen": -7.306541919708252, |
| "logps/rejected": -15.145895004272461, |
| "loss": 1.1673789024353027, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5968848466873169, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13542409241199493, |
| "rewards/margins": 0.3165813684463501, |
| "rewards/rejected": -0.18115724623203278, |
| "step": 18, |
| "train_speed(iter/s)": 0.005611 |
| }, |
| { |
| "epoch": 0.08501118568232663, |
| "grad_norm": 4.076289176940918, |
| "learning_rate": 0.00016521739130434784, |
| "logits/chosen": -0.33677053451538086, |
| "logits/rejected": -0.42119455337524414, |
| "logps/chosen": -6.263703346252441, |
| "logps/rejected": -17.470884323120117, |
| "loss": 0.9588873982429504, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.42532750964164734, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.17247594892978668, |
| "rewards/margins": 0.41344738006591797, |
| "rewards/rejected": -0.24097144603729248, |
| "step": 19, |
| "train_speed(iter/s)": 0.005613 |
| }, |
| { |
| "epoch": 0.0894854586129754, |
| "grad_norm": 2.8480069637298584, |
| "learning_rate": 0.00017391304347826088, |
| "logits/chosen": -0.27281832695007324, |
| "logits/rejected": -0.3382987976074219, |
| "logps/chosen": -8.123124122619629, |
| "logps/rejected": -17.66653060913086, |
| "loss": 1.1668881177902222, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5924651622772217, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.00576989259570837, |
| "rewards/margins": 0.35684970021247864, |
| "rewards/rejected": -0.36261963844299316, |
| "step": 20, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.09395973154362416, |
| "grad_norm": 2.664872646331787, |
| "learning_rate": 0.00018260869565217392, |
| "logits/chosen": -0.2556913197040558, |
| "logits/rejected": -0.3474789261817932, |
| "logps/chosen": -5.220917701721191, |
| "logps/rejected": -23.547605514526367, |
| "loss": 0.9417474865913391, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4541093111038208, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.043988656252622604, |
| "rewards/margins": 0.6349199414253235, |
| "rewards/rejected": -0.5909312963485718, |
| "step": 21, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.09843400447427293, |
| "grad_norm": 2.325854539871216, |
| "learning_rate": 0.00019130434782608697, |
| "logits/chosen": -0.2636699676513672, |
| "logits/rejected": -0.33711570501327515, |
| "logps/chosen": -8.976820945739746, |
| "logps/rejected": -19.829130172729492, |
| "loss": 1.0729048252105713, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.595395028591156, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.09026271104812622, |
| "rewards/margins": 0.623466432094574, |
| "rewards/rejected": -0.5332037210464478, |
| "step": 22, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.1029082774049217, |
| "grad_norm": 2.684507369995117, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.24679847061634064, |
| "logits/rejected": -0.3378849923610687, |
| "logps/chosen": -7.14666223526001, |
| "logps/rejected": -23.644922256469727, |
| "loss": 0.9377666115760803, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.427379846572876, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.0773518830537796, |
| "rewards/margins": 0.6581345796585083, |
| "rewards/rejected": -0.7354865074157715, |
| "step": 23, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.10738255033557047, |
| "grad_norm": 2.6381781101226807, |
| "learning_rate": 0.00019999724204599747, |
| "logits/chosen": -0.246232807636261, |
| "logits/rejected": -0.3716464936733246, |
| "logps/chosen": -6.111550331115723, |
| "logps/rejected": -24.672321319580078, |
| "loss": 0.9850844740867615, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5655074119567871, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.027247514575719833, |
| "rewards/margins": 0.8617139458656311, |
| "rewards/rejected": -0.8344664573669434, |
| "step": 24, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.11185682326621924, |
| "grad_norm": 2.5178840160369873, |
| "learning_rate": 0.00019998896833611603, |
| "logits/chosen": -0.23005428910255432, |
| "logits/rejected": -0.3447636365890503, |
| "logps/chosen": -8.92835807800293, |
| "logps/rejected": -25.173357009887695, |
| "loss": 1.1135735511779785, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5982357263565063, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.07063186913728714, |
| "rewards/margins": 0.6180722117424011, |
| "rewards/rejected": -0.6887040138244629, |
| "step": 25, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.116331096196868, |
| "grad_norm": 1.9486284255981445, |
| "learning_rate": 0.0001999751793267259, |
| "logits/chosen": -0.17476646602153778, |
| "logits/rejected": -0.28910771012306213, |
| "logps/chosen": -6.982065200805664, |
| "logps/rejected": -25.235271453857422, |
| "loss": 0.9616149067878723, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4814175069332123, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.017612511292099953, |
| "rewards/margins": 0.7644082307815552, |
| "rewards/rejected": -0.7467955350875854, |
| "step": 26, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.12080536912751678, |
| "grad_norm": 2.024102210998535, |
| "learning_rate": 0.0001999558757784162, |
| "logits/chosen": -0.22182828187942505, |
| "logits/rejected": -0.33474183082580566, |
| "logps/chosen": -8.283013343811035, |
| "logps/rejected": -23.36388397216797, |
| "loss": 1.071007251739502, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5651244521141052, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.005644451826810837, |
| "rewards/margins": 0.549856424331665, |
| "rewards/rejected": -0.5555009245872498, |
| "step": 27, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.12527964205816555, |
| "grad_norm": 2.334916830062866, |
| "learning_rate": 0.0001999310587559529, |
| "logits/chosen": -0.1745743602514267, |
| "logits/rejected": -0.28861111402511597, |
| "logps/chosen": -7.148108959197998, |
| "logps/rejected": -22.50945472717285, |
| "loss": 1.0254650115966797, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5092177391052246, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.024923313409090042, |
| "rewards/margins": 0.6763253211975098, |
| "rewards/rejected": -0.6514019966125488, |
| "step": 28, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.1297539149888143, |
| "grad_norm": 4.392460823059082, |
| "learning_rate": 0.00019990072962822007, |
| "logits/chosen": -0.1607450246810913, |
| "logits/rejected": -0.30258849263191223, |
| "logps/chosen": -8.330341339111328, |
| "logps/rejected": -23.302955627441406, |
| "loss": 1.1018993854522705, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6234543919563293, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1296255737543106, |
| "rewards/margins": 0.7057831883430481, |
| "rewards/rejected": -0.5761575102806091, |
| "step": 29, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.1342281879194631, |
| "grad_norm": 1.9557713270187378, |
| "learning_rate": 0.00019986489006814452, |
| "logits/chosen": -0.15406475961208344, |
| "logits/rejected": -0.32140052318573, |
| "logps/chosen": -6.338605880737305, |
| "logps/rejected": -27.934860229492188, |
| "loss": 0.8741156458854675, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.43781277537345886, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.005505967885255814, |
| "rewards/margins": 0.9152470827102661, |
| "rewards/rejected": -0.9097411632537842, |
| "step": 30, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.13870246085011187, |
| "grad_norm": 3.570540189743042, |
| "learning_rate": 0.00019982354205260347, |
| "logits/chosen": -0.13595342636108398, |
| "logits/rejected": -0.24140222370624542, |
| "logps/chosen": -9.938725471496582, |
| "logps/rejected": -26.17535972595215, |
| "loss": 1.1157177686691284, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6451507806777954, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.06947211921215057, |
| "rewards/margins": 0.9398424029350281, |
| "rewards/rejected": -1.0093145370483398, |
| "step": 31, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.14317673378076062, |
| "grad_norm": 2.263465404510498, |
| "learning_rate": 0.00019977668786231534, |
| "logits/chosen": -0.13489112257957458, |
| "logits/rejected": -0.2546093165874481, |
| "logps/chosen": -9.485723495483398, |
| "logps/rejected": -30.659465789794922, |
| "loss": 1.1024084091186523, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7035719752311707, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.094739630818367, |
| "rewards/margins": 1.1513433456420898, |
| "rewards/rejected": -1.0566036701202393, |
| "step": 32, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.1476510067114094, |
| "grad_norm": 4.003927230834961, |
| "learning_rate": 0.00019972433008171416, |
| "logits/chosen": -0.2770022451877594, |
| "logits/rejected": -0.3445180356502533, |
| "logps/chosen": -7.854065895080566, |
| "logps/rejected": -17.47032356262207, |
| "loss": 1.1709215641021729, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.635529637336731, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.03749306499958038, |
| "rewards/margins": 0.6043045520782471, |
| "rewards/rejected": -0.5668113827705383, |
| "step": 33, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.15212527964205816, |
| "grad_norm": 2.930697202682495, |
| "learning_rate": 0.00019966647159880703, |
| "logits/chosen": -0.20737791061401367, |
| "logits/rejected": -0.3068307638168335, |
| "logps/chosen": -5.924644470214844, |
| "logps/rejected": -20.47597885131836, |
| "loss": 0.9971151947975159, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4416566789150238, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.04966810718178749, |
| "rewards/margins": 0.48412013053894043, |
| "rewards/rejected": -0.4344520568847656, |
| "step": 34, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.15659955257270694, |
| "grad_norm": 2.292167901992798, |
| "learning_rate": 0.00019960311560501454, |
| "logits/chosen": -0.20404092967510223, |
| "logits/rejected": -0.2819429636001587, |
| "logps/chosen": -9.679859161376953, |
| "logps/rejected": -22.75543785095215, |
| "loss": 1.1032215356826782, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6184647679328918, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.07092233002185822, |
| "rewards/margins": 0.8317320942878723, |
| "rewards/rejected": -0.7608097791671753, |
| "step": 35, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.1610738255033557, |
| "grad_norm": 1.5171147584915161, |
| "learning_rate": 0.0001995342655949951, |
| "logits/chosen": -0.18677955865859985, |
| "logits/rejected": -0.30433592200279236, |
| "logps/chosen": -7.317014217376709, |
| "logps/rejected": -23.58348274230957, |
| "loss": 1.0045303106307983, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5375936031341553, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13633473217487335, |
| "rewards/margins": 0.7853459119796753, |
| "rewards/rejected": -0.6490110754966736, |
| "step": 36, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.16554809843400448, |
| "grad_norm": 1.8760979175567627, |
| "learning_rate": 0.00019945992536645187, |
| "logits/chosen": -0.17943330109119415, |
| "logits/rejected": -0.27492690086364746, |
| "logps/chosen": -9.596693992614746, |
| "logps/rejected": -25.057254791259766, |
| "loss": 1.0518603324890137, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5895128846168518, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 4.468485713005066e-06, |
| "rewards/margins": 0.8271102905273438, |
| "rewards/rejected": -0.827105700969696, |
| "step": 37, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.17002237136465326, |
| "grad_norm": 2.1765875816345215, |
| "learning_rate": 0.0001993800990199235, |
| "logits/chosen": -0.17796888947486877, |
| "logits/rejected": -0.27494707703590393, |
| "logps/chosen": -8.854700088500977, |
| "logps/rejected": -22.589916229248047, |
| "loss": 1.1736024618148804, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6680553555488586, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.027395736426115036, |
| "rewards/margins": 0.702995777130127, |
| "rewards/rejected": -0.6756000518798828, |
| "step": 38, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.174496644295302, |
| "grad_norm": 2.0398788452148438, |
| "learning_rate": 0.0001992947909585578, |
| "logits/chosen": -0.15950845181941986, |
| "logits/rejected": -0.285598486661911, |
| "logps/chosen": -7.358231544494629, |
| "logps/rejected": -23.962072372436523, |
| "loss": 0.9235795736312866, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5008021593093872, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.1876223385334015, |
| "rewards/margins": 0.9989414215087891, |
| "rewards/rejected": -0.8113190531730652, |
| "step": 39, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.1789709172259508, |
| "grad_norm": 2.137016773223877, |
| "learning_rate": 0.000199204005887869, |
| "logits/chosen": -0.18114963173866272, |
| "logits/rejected": -0.23074296116828918, |
| "logps/chosen": -9.713254928588867, |
| "logps/rejected": -15.0142822265625, |
| "loss": 1.1612474918365479, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5743556618690491, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.08143825083971024, |
| "rewards/margins": 0.3889071047306061, |
| "rewards/rejected": -0.30746886134147644, |
| "step": 40, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.18344519015659955, |
| "grad_norm": 2.5089111328125, |
| "learning_rate": 0.000199107748815478, |
| "logits/chosen": -0.17036814987659454, |
| "logits/rejected": -0.25284841656684875, |
| "logps/chosen": -7.5790181159973145, |
| "logps/rejected": -22.757585525512695, |
| "loss": 0.9598883986473083, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5350608229637146, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.10258667171001434, |
| "rewards/margins": 0.982707679271698, |
| "rewards/rejected": -0.8801209926605225, |
| "step": 41, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.18791946308724833, |
| "grad_norm": 2.425978660583496, |
| "learning_rate": 0.00019900602505083648, |
| "logits/chosen": -0.17056672275066376, |
| "logits/rejected": -0.24402382969856262, |
| "logps/chosen": -9.11720085144043, |
| "logps/rejected": -26.321962356567383, |
| "loss": 0.9588575959205627, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5199980735778809, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.02633209154009819, |
| "rewards/margins": 1.148821234703064, |
| "rewards/rejected": -1.1751532554626465, |
| "step": 42, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.19239373601789708, |
| "grad_norm": 4.055136203765869, |
| "learning_rate": 0.0001988988402049336, |
| "logits/chosen": -0.06654427945613861, |
| "logits/rejected": -0.19404850900173187, |
| "logps/chosen": -10.976348876953125, |
| "logps/rejected": -35.80363464355469, |
| "loss": 1.205069661140442, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7504922151565552, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.11243688315153122, |
| "rewards/margins": 1.5014084577560425, |
| "rewards/rejected": -1.6138453483581543, |
| "step": 43, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.19686800894854586, |
| "grad_norm": 2.9351325035095215, |
| "learning_rate": 0.00019878620018998696, |
| "logits/chosen": -0.21978600323200226, |
| "logits/rejected": -0.32882797718048096, |
| "logps/chosen": -6.976373195648193, |
| "logps/rejected": -30.215131759643555, |
| "loss": 0.9490640759468079, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5300356149673462, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.11029693484306335, |
| "rewards/margins": 1.1996657848358154, |
| "rewards/rejected": -1.3099626302719116, |
| "step": 44, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.20134228187919462, |
| "grad_norm": 2.6874403953552246, |
| "learning_rate": 0.00019866811121911607, |
| "logits/chosen": -0.1936551034450531, |
| "logits/rejected": -0.284149706363678, |
| "logps/chosen": -6.9035515785217285, |
| "logps/rejected": -22.284351348876953, |
| "loss": 0.8805959820747375, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4068413972854614, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.07986758649349213, |
| "rewards/margins": 0.8386067152023315, |
| "rewards/rejected": -0.7587391138076782, |
| "step": 45, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.2058165548098434, |
| "grad_norm": 1.7893315553665161, |
| "learning_rate": 0.000198544579806, |
| "logits/chosen": -0.20961534976959229, |
| "logits/rejected": -0.33521464467048645, |
| "logps/chosen": -7.002056121826172, |
| "logps/rejected": -25.93248748779297, |
| "loss": 0.9666993618011475, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4805544912815094, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.07972076535224915, |
| "rewards/margins": 0.8979825377464294, |
| "rewards/rejected": -0.8182618021965027, |
| "step": 46, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.21029082774049218, |
| "grad_norm": 2.2033731937408447, |
| "learning_rate": 0.0001984156127645178, |
| "logits/chosen": -0.21569889783859253, |
| "logits/rejected": -0.38949355483055115, |
| "logps/chosen": -5.476630687713623, |
| "logps/rejected": -27.572505950927734, |
| "loss": 0.8785698413848877, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.42979833483695984, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.17229261994361877, |
| "rewards/margins": 1.044838547706604, |
| "rewards/rejected": -0.8725459575653076, |
| "step": 47, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.21476510067114093, |
| "grad_norm": 2.1163032054901123, |
| "learning_rate": 0.00019828121720837286, |
| "logits/chosen": -0.34833359718322754, |
| "logits/rejected": -0.41568437218666077, |
| "logps/chosen": -7.711272239685059, |
| "logps/rejected": -25.60678482055664, |
| "loss": 0.9890228509902954, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.548301100730896, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.07654333114624023, |
| "rewards/margins": 1.2576009035110474, |
| "rewards/rejected": -1.1810575723648071, |
| "step": 48, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.21923937360178972, |
| "grad_norm": 1.987210988998413, |
| "learning_rate": 0.00019814140055070042, |
| "logits/chosen": -0.2835751473903656, |
| "logits/rejected": -0.38263481855392456, |
| "logps/chosen": -8.624626159667969, |
| "logps/rejected": -25.244403839111328, |
| "loss": 0.9971789121627808, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5687417387962341, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.09697108715772629, |
| "rewards/margins": 1.0362679958343506, |
| "rewards/rejected": -0.9392969608306885, |
| "step": 49, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.22371364653243847, |
| "grad_norm": 1.9501866102218628, |
| "learning_rate": 0.0001979961705036587, |
| "logits/chosen": -0.38833191990852356, |
| "logits/rejected": -0.4256289303302765, |
| "logps/chosen": -8.22580623626709, |
| "logps/rejected": -23.093332290649414, |
| "loss": 0.9298551082611084, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4849992096424103, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.04563959687948227, |
| "rewards/margins": 1.0999000072479248, |
| "rewards/rejected": -1.054260492324829, |
| "step": 50, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.22818791946308725, |
| "grad_norm": 2.496234893798828, |
| "learning_rate": 0.00019784553507800349, |
| "logits/chosen": -0.2710403501987457, |
| "logits/rejected": -0.41735631227493286, |
| "logps/chosen": -8.894371032714844, |
| "logps/rejected": -28.586936950683594, |
| "loss": 1.1125377416610718, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6875916719436646, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.10749680548906326, |
| "rewards/margins": 1.2817683219909668, |
| "rewards/rejected": -1.174271583557129, |
| "step": 51, |
| "train_speed(iter/s)": 0.005611 |
| }, |
| { |
| "epoch": 0.232662192393736, |
| "grad_norm": 3.274833917617798, |
| "learning_rate": 0.00019768950258264623, |
| "logits/chosen": -0.39167118072509766, |
| "logits/rejected": -0.4531523883342743, |
| "logps/chosen": -9.231929779052734, |
| "logps/rejected": -22.90334701538086, |
| "loss": 1.1200237274169922, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6103911399841309, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.16369852423667908, |
| "rewards/margins": 0.9332659244537354, |
| "rewards/rejected": -1.0969644784927368, |
| "step": 52, |
| "train_speed(iter/s)": 0.005612 |
| }, |
| { |
| "epoch": 0.2371364653243848, |
| "grad_norm": 2.9111413955688477, |
| "learning_rate": 0.0001975280816241959, |
| "logits/chosen": -0.3340080976486206, |
| "logits/rejected": -0.3975221514701843, |
| "logps/chosen": -11.631303787231445, |
| "logps/rejected": -26.26183319091797, |
| "loss": 1.1326872110366821, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6245000958442688, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.032076627016067505, |
| "rewards/margins": 0.9003015160560608, |
| "rewards/rejected": -0.9323781132698059, |
| "step": 53, |
| "train_speed(iter/s)": 0.005612 |
| }, |
| { |
| "epoch": 0.24161073825503357, |
| "grad_norm": 2.7828478813171387, |
| "learning_rate": 0.00019736128110648407, |
| "logits/chosen": -0.345708429813385, |
| "logits/rejected": -0.447677880525589, |
| "logps/chosen": -8.686080932617188, |
| "logps/rejected": -23.45962142944336, |
| "loss": 0.9978072047233582, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5644747018814087, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.18989036977291107, |
| "rewards/margins": 1.1108869314193726, |
| "rewards/rejected": -0.9209966659545898, |
| "step": 54, |
| "train_speed(iter/s)": 0.005613 |
| }, |
| { |
| "epoch": 0.24608501118568232, |
| "grad_norm": 2.1220943927764893, |
| "learning_rate": 0.0001971891102300738, |
| "logits/chosen": -0.3758760690689087, |
| "logits/rejected": -0.4768344461917877, |
| "logps/chosen": -7.041341304779053, |
| "logps/rejected": -22.334056854248047, |
| "loss": 1.0144665241241455, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5401796698570251, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1189827248454094, |
| "rewards/margins": 0.939737856388092, |
| "rewards/rejected": -0.820755124092102, |
| "step": 55, |
| "train_speed(iter/s)": 0.005613 |
| }, |
| { |
| "epoch": 0.2505592841163311, |
| "grad_norm": 1.843038558959961, |
| "learning_rate": 0.00019701157849175228, |
| "logits/chosen": -0.37703099846839905, |
| "logits/rejected": -0.46533748507499695, |
| "logps/chosen": -8.012426376342773, |
| "logps/rejected": -22.621742248535156, |
| "loss": 1.0123478174209595, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5707098245620728, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15585632622241974, |
| "rewards/margins": 0.9487887620925903, |
| "rewards/rejected": -0.792932391166687, |
| "step": 56, |
| "train_speed(iter/s)": 0.005614 |
| }, |
| { |
| "epoch": 0.2550335570469799, |
| "grad_norm": 2.230550527572632, |
| "learning_rate": 0.00019682869568400684, |
| "logits/chosen": -0.27077603340148926, |
| "logits/rejected": -0.43189266324043274, |
| "logps/chosen": -6.5583930015563965, |
| "logps/rejected": -29.555484771728516, |
| "loss": 0.9619537591934204, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5382556915283203, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.012836463749408722, |
| "rewards/margins": 1.0998661518096924, |
| "rewards/rejected": -1.0870296955108643, |
| "step": 57, |
| "train_speed(iter/s)": 0.005614 |
| }, |
| { |
| "epoch": 0.2595078299776286, |
| "grad_norm": 2.252734661102295, |
| "learning_rate": 0.00019664047189448493, |
| "logits/chosen": -0.30061304569244385, |
| "logits/rejected": -0.4390646517276764, |
| "logps/chosen": -5.5507659912109375, |
| "logps/rejected": -26.51093292236328, |
| "loss": 0.9082356095314026, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41620010137557983, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.027434132993221283, |
| "rewards/margins": 0.9383218288421631, |
| "rewards/rejected": -0.9657559394836426, |
| "step": 58, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.2639821029082774, |
| "grad_norm": 2.943512201309204, |
| "learning_rate": 0.00019644691750543767, |
| "logits/chosen": -0.2617231011390686, |
| "logits/rejected": -0.38011983036994934, |
| "logps/chosen": -10.003762245178223, |
| "logps/rejected": -27.785303115844727, |
| "loss": 1.1763017177581787, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6572035551071167, |
| "rewards/accuracies": 0.671875, |
| "rewards/chosen": -0.07653994858264923, |
| "rewards/margins": 0.9367327690124512, |
| "rewards/rejected": -1.0132726430892944, |
| "step": 59, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.2684563758389262, |
| "grad_norm": 1.736883521080017, |
| "learning_rate": 0.00019624804319314705, |
| "logits/chosen": -0.3555205464363098, |
| "logits/rejected": -0.4549139738082886, |
| "logps/chosen": -7.022237300872803, |
| "logps/rejected": -28.04250717163086, |
| "loss": 0.9159735441207886, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.47989311814308167, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.060450125485658646, |
| "rewards/margins": 1.1664522886276245, |
| "rewards/rejected": -1.1060020923614502, |
| "step": 60, |
| "train_speed(iter/s)": 0.005616 |
| }, |
| { |
| "epoch": 0.27293064876957496, |
| "grad_norm": 1.8152859210968018, |
| "learning_rate": 0.00019604385992733715, |
| "logits/chosen": -0.2845993638038635, |
| "logits/rejected": -0.3513396680355072, |
| "logps/chosen": -9.077548027038574, |
| "logps/rejected": -22.780668258666992, |
| "loss": 0.9680572748184204, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4960174560546875, |
| "rewards/accuracies": 0.734375, |
| "rewards/chosen": 0.06926427036523819, |
| "rewards/margins": 0.9312706589698792, |
| "rewards/rejected": -0.862006425857544, |
| "step": 61, |
| "train_speed(iter/s)": 0.005616 |
| }, |
| { |
| "epoch": 0.27740492170022374, |
| "grad_norm": 2.46378755569458, |
| "learning_rate": 0.00019583437897056915, |
| "logits/chosen": -0.3284386098384857, |
| "logits/rejected": -0.4009109139442444, |
| "logps/chosen": -8.893238067626953, |
| "logps/rejected": -21.27986717224121, |
| "loss": 1.128482699394226, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5518131256103516, |
| "rewards/accuracies": 0.609375, |
| "rewards/chosen": 0.010286218486726284, |
| "rewards/margins": 0.5274713635444641, |
| "rewards/rejected": -0.5171850919723511, |
| "step": 62, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.28187919463087246, |
| "grad_norm": 2.1732046604156494, |
| "learning_rate": 0.00019561961187761985, |
| "logits/chosen": -0.24422433972358704, |
| "logits/rejected": -0.375420480966568, |
| "logps/chosen": -6.340362071990967, |
| "logps/rejected": -20.184795379638672, |
| "loss": 0.987078070640564, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4728015065193176, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.168831467628479, |
| "rewards/margins": 0.6444362998008728, |
| "rewards/rejected": -0.4756048321723938, |
| "step": 63, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.28635346756152125, |
| "grad_norm": 2.231933355331421, |
| "learning_rate": 0.00019539957049484458, |
| "logits/chosen": -0.22632239758968353, |
| "logits/rejected": -0.3298073410987854, |
| "logps/chosen": -5.790369987487793, |
| "logps/rejected": -21.647865295410156, |
| "loss": 0.9348481893539429, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4110260307788849, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.11635202169418335, |
| "rewards/margins": 0.6466613411903381, |
| "rewards/rejected": -0.5303093194961548, |
| "step": 64, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.29082774049217003, |
| "grad_norm": 2.592883825302124, |
| "learning_rate": 0.00019517426695952358, |
| "logits/chosen": -0.2631714344024658, |
| "logits/rejected": -0.3835735321044922, |
| "logps/chosen": -8.772464752197266, |
| "logps/rejected": -23.13977813720703, |
| "loss": 1.0820575952529907, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6241056323051453, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.1670496016740799, |
| "rewards/margins": 0.9779586791992188, |
| "rewards/rejected": -0.8109091520309448, |
| "step": 65, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.2953020134228188, |
| "grad_norm": 2.1508402824401855, |
| "learning_rate": 0.0001949437136991925, |
| "logits/chosen": -0.28555068373680115, |
| "logits/rejected": -0.42386704683303833, |
| "logps/chosen": -4.788064002990723, |
| "logps/rejected": -26.989431381225586, |
| "loss": 0.7586299777030945, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3784462809562683, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09788702428340912, |
| "rewards/margins": 1.3373408317565918, |
| "rewards/rejected": -1.239453673362732, |
| "step": 66, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.29977628635346754, |
| "grad_norm": 4.065814971923828, |
| "learning_rate": 0.00019470792343095718, |
| "logits/chosen": -0.3336545526981354, |
| "logits/rejected": -0.3990672826766968, |
| "logps/chosen": -8.33244514465332, |
| "logps/rejected": -29.013874053955078, |
| "loss": 1.0058256387710571, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6062498092651367, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.020362887531518936, |
| "rewards/margins": 1.5735464096069336, |
| "rewards/rejected": -1.5939093828201294, |
| "step": 67, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.3042505592841163, |
| "grad_norm": 2.998826265335083, |
| "learning_rate": 0.0001944669091607919, |
| "logits/chosen": -0.17095105350017548, |
| "logits/rejected": -0.33402499556541443, |
| "logps/chosen": -7.172041893005371, |
| "logps/rejected": -36.20872497558594, |
| "loss": 0.8435176610946655, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5263906121253967, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.01939973793923855, |
| "rewards/margins": 2.0239157676696777, |
| "rewards/rejected": -2.004516124725342, |
| "step": 68, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.3087248322147651, |
| "grad_norm": 2.8316242694854736, |
| "learning_rate": 0.00019422068418282202, |
| "logits/chosen": -0.23822994530200958, |
| "logits/rejected": -0.3510916531085968, |
| "logps/chosen": -8.204334259033203, |
| "logps/rejected": -26.101930618286133, |
| "loss": 0.9929779171943665, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.592566967010498, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.07919630408287048, |
| "rewards/margins": 1.4216609001159668, |
| "rewards/rejected": -1.3424646854400635, |
| "step": 69, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.3131991051454139, |
| "grad_norm": 2.1727654933929443, |
| "learning_rate": 0.00019396926207859084, |
| "logits/chosen": -0.19429409503936768, |
| "logits/rejected": -0.33817699551582336, |
| "logps/chosen": -7.17151403427124, |
| "logps/rejected": -29.4168701171875, |
| "loss": 0.8446124196052551, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4697123169898987, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.09260577708482742, |
| "rewards/margins": 1.373153567314148, |
| "rewards/rejected": -1.280547857284546, |
| "step": 70, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.31767337807606266, |
| "grad_norm": 2.131971597671509, |
| "learning_rate": 0.00019371265671631037, |
| "logits/chosen": -0.2500344514846802, |
| "logits/rejected": -0.391252338886261, |
| "logps/chosen": -6.454396724700928, |
| "logps/rejected": -31.325042724609375, |
| "loss": 0.857275128364563, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45489931106567383, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.12299442291259766, |
| "rewards/margins": 1.4080641269683838, |
| "rewards/rejected": -1.2850697040557861, |
| "step": 71, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.3221476510067114, |
| "grad_norm": 2.6743502616882324, |
| "learning_rate": 0.00019345088225009626, |
| "logits/chosen": -0.20935188233852386, |
| "logits/rejected": -0.3504689633846283, |
| "logps/chosen": -8.028803825378418, |
| "logps/rejected": -33.473976135253906, |
| "loss": 0.951294481754303, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5516812205314636, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.002318259794265032, |
| "rewards/margins": 1.4767383337020874, |
| "rewards/rejected": -1.4744200706481934, |
| "step": 72, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.32662192393736017, |
| "grad_norm": 2.1057374477386475, |
| "learning_rate": 0.0001931839531191873, |
| "logits/chosen": -0.21616998314857483, |
| "logits/rejected": -0.3215841054916382, |
| "logps/chosen": -6.386109352111816, |
| "logps/rejected": -26.058643341064453, |
| "loss": 0.9168461561203003, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5115207433700562, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.16049177944660187, |
| "rewards/margins": 1.1623506546020508, |
| "rewards/rejected": -1.0018587112426758, |
| "step": 73, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.33109619686800895, |
| "grad_norm": 1.8313034772872925, |
| "learning_rate": 0.00019291188404714878, |
| "logits/chosen": -0.2762296795845032, |
| "logits/rejected": -0.3759133815765381, |
| "logps/chosen": -7.998234748840332, |
| "logps/rejected": -29.753068923950195, |
| "loss": 0.9865212440490723, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5828378796577454, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.04827752709388733, |
| "rewards/margins": 1.679243564605713, |
| "rewards/rejected": -1.6309659481048584, |
| "step": 74, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.33557046979865773, |
| "grad_norm": 2.7458698749542236, |
| "learning_rate": 0.0001926346900410604, |
| "logits/chosen": -0.26945143938064575, |
| "logits/rejected": -0.319683313369751, |
| "logps/chosen": -9.152531623840332, |
| "logps/rejected": -22.447328567504883, |
| "loss": 1.118042230606079, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6387289762496948, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.012468192726373672, |
| "rewards/margins": 0.904109537601471, |
| "rewards/rejected": -0.8916412591934204, |
| "step": 75, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.3400447427293065, |
| "grad_norm": 2.5607635974884033, |
| "learning_rate": 0.00019235238639068856, |
| "logits/chosen": -0.19932252168655396, |
| "logits/rejected": -0.37197503447532654, |
| "logps/chosen": -7.51588773727417, |
| "logps/rejected": -33.714637756347656, |
| "loss": 0.9608560800552368, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5613290667533875, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0078031644225120544, |
| "rewards/margins": 1.3550933599472046, |
| "rewards/rejected": -1.347290277481079, |
| "step": 76, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.34451901565995524, |
| "grad_norm": 2.152766227722168, |
| "learning_rate": 0.00019206498866764288, |
| "logits/chosen": -0.18406139314174652, |
| "logits/rejected": -0.36746495962142944, |
| "logps/chosen": -4.768266677856445, |
| "logps/rejected": -35.82173156738281, |
| "loss": 0.8752678632736206, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5213672518730164, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.024980902671813965, |
| "rewards/margins": 1.6603260040283203, |
| "rewards/rejected": -1.6853069067001343, |
| "step": 77, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.348993288590604, |
| "grad_norm": 1.957088589668274, |
| "learning_rate": 0.0001917725127245174, |
| "logits/chosen": -0.2368219494819641, |
| "logits/rejected": -0.31324779987335205, |
| "logps/chosen": -7.644411087036133, |
| "logps/rejected": -25.68874740600586, |
| "loss": 0.9143100380897522, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.47288480401039124, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.04239855334162712, |
| "rewards/margins": 0.925534188747406, |
| "rewards/rejected": -0.8831356167793274, |
| "step": 78, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.3534675615212528, |
| "grad_norm": 2.0980629920959473, |
| "learning_rate": 0.0001914749746940161, |
| "logits/chosen": -0.3016536831855774, |
| "logits/rejected": -0.3982566297054291, |
| "logps/chosen": -7.566556930541992, |
| "logps/rejected": -26.53971290588379, |
| "loss": 0.9413108825683594, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5161536931991577, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.1749323159456253, |
| "rewards/margins": 1.1360664367675781, |
| "rewards/rejected": -0.9611341953277588, |
| "step": 79, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.3579418344519016, |
| "grad_norm": 2.1753523349761963, |
| "learning_rate": 0.00019117239098806295, |
| "logits/chosen": -0.25202423334121704, |
| "logits/rejected": -0.3296201527118683, |
| "logps/chosen": -8.623554229736328, |
| "logps/rejected": -24.96216583251953, |
| "loss": 0.9156360626220703, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.521922767162323, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.17782656848430634, |
| "rewards/margins": 1.1749255657196045, |
| "rewards/rejected": -0.997098982334137, |
| "step": 80, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.3624161073825503, |
| "grad_norm": 3.2816970348358154, |
| "learning_rate": 0.00019086477829689685, |
| "logits/chosen": -0.22371181845664978, |
| "logits/rejected": -0.3009035885334015, |
| "logps/chosen": -10.19958782196045, |
| "logps/rejected": -26.129743576049805, |
| "loss": 1.161747932434082, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7456332445144653, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06418374180793762, |
| "rewards/margins": 1.1640924215316772, |
| "rewards/rejected": -1.099908709526062, |
| "step": 81, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.3668903803131991, |
| "grad_norm": 4.27427339553833, |
| "learning_rate": 0.0001905521535881509, |
| "logits/chosen": -0.15121029317378998, |
| "logits/rejected": -0.3036370873451233, |
| "logps/chosen": -7.66187858581543, |
| "logps/rejected": -30.927831649780273, |
| "loss": 1.1210309267044067, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6641408205032349, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.00867693591862917, |
| "rewards/margins": 1.272807002067566, |
| "rewards/rejected": -1.2814840078353882, |
| "step": 82, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.3713646532438479, |
| "grad_norm": 2.29978084564209, |
| "learning_rate": 0.00019023453410591635, |
| "logits/chosen": -0.21486631035804749, |
| "logits/rejected": -0.3746005892753601, |
| "logps/chosen": -9.569436073303223, |
| "logps/rejected": -32.21381378173828, |
| "loss": 1.1229475736618042, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6794871687889099, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.0131741464138031, |
| "rewards/margins": 1.4518442153930664, |
| "rewards/rejected": -1.4386701583862305, |
| "step": 83, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.37583892617449666, |
| "grad_norm": 2.6593234539031982, |
| "learning_rate": 0.00018991193736979175, |
| "logits/chosen": -0.12227392941713333, |
| "logits/rejected": -0.2726154327392578, |
| "logps/chosen": -7.078165054321289, |
| "logps/rejected": -30.666397094726562, |
| "loss": 0.7911984920501709, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4572717845439911, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.06052548438310623, |
| "rewards/margins": 1.428847074508667, |
| "rewards/rejected": -1.3683216571807861, |
| "step": 84, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.38031319910514544, |
| "grad_norm": 2.8958120346069336, |
| "learning_rate": 0.00018958438117391618, |
| "logits/chosen": -0.0972181037068367, |
| "logits/rejected": -0.2590752840042114, |
| "logps/chosen": -8.193891525268555, |
| "logps/rejected": -35.56235885620117, |
| "loss": 1.0837842226028442, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6906030774116516, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.06835082173347473, |
| "rewards/margins": 1.79606294631958, |
| "rewards/rejected": -1.8644137382507324, |
| "step": 85, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.38478747203579416, |
| "grad_norm": 3.5083513259887695, |
| "learning_rate": 0.00018925188358598813, |
| "logits/chosen": -0.17531399428844452, |
| "logits/rejected": -0.25995591282844543, |
| "logps/chosen": -8.639050483703613, |
| "logps/rejected": -29.226892471313477, |
| "loss": 1.0494576692581177, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6720495820045471, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.03724604845046997, |
| "rewards/margins": 1.4595208168029785, |
| "rewards/rejected": -1.4967668056488037, |
| "step": 86, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.38926174496644295, |
| "grad_norm": 2.0132222175598145, |
| "learning_rate": 0.00018891446294626866, |
| "logits/chosen": -0.05051232874393463, |
| "logits/rejected": -0.25516456365585327, |
| "logps/chosen": -4.587294578552246, |
| "logps/rejected": -36.072322845458984, |
| "loss": 0.602871298789978, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.32869917154312134, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.17169487476348877, |
| "rewards/margins": 2.032069444656372, |
| "rewards/rejected": -1.8603744506835938, |
| "step": 87, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.39373601789709173, |
| "grad_norm": 3.1251611709594727, |
| "learning_rate": 0.00018857213786656985, |
| "logits/chosen": -0.16258063912391663, |
| "logits/rejected": -0.2599830627441406, |
| "logps/chosen": -9.219761848449707, |
| "logps/rejected": -27.451026916503906, |
| "loss": 1.0880504846572876, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6885117292404175, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.04242437332868576, |
| "rewards/margins": 1.207942247390747, |
| "rewards/rejected": -1.1655179262161255, |
| "step": 88, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.3982102908277405, |
| "grad_norm": 2.0855207443237305, |
| "learning_rate": 0.0001882249272292282, |
| "logits/chosen": -0.1683093011379242, |
| "logits/rejected": -0.3001120686531067, |
| "logps/chosen": -7.336337089538574, |
| "logps/rejected": -24.294296264648438, |
| "loss": 0.9775981903076172, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.515565812587738, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.09023457765579224, |
| "rewards/margins": 0.9091135859489441, |
| "rewards/rejected": -0.8188790082931519, |
| "step": 89, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.40268456375838924, |
| "grad_norm": 1.8228232860565186, |
| "learning_rate": 0.00018787285018606297, |
| "logits/chosen": -0.1059221625328064, |
| "logits/rejected": -0.30855828523635864, |
| "logps/chosen": -3.3506364822387695, |
| "logps/rejected": -32.792762756347656, |
| "loss": 0.6121629476547241, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2924441993236542, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.11779851466417313, |
| "rewards/margins": 1.558281660079956, |
| "rewards/rejected": -1.4404832124710083, |
| "step": 90, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.407158836689038, |
| "grad_norm": 2.038564443588257, |
| "learning_rate": 0.00018751592615732005, |
| "logits/chosen": -0.23549917340278625, |
| "logits/rejected": -0.28893160820007324, |
| "logps/chosen": -7.049911975860596, |
| "logps/rejected": -23.307397842407227, |
| "loss": 0.8128443956375122, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.42805472016334534, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.235755056142807, |
| "rewards/margins": 1.322766900062561, |
| "rewards/rejected": -1.0870119333267212, |
| "step": 91, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.4116331096196868, |
| "grad_norm": 3.0157337188720703, |
| "learning_rate": 0.0001871541748306005, |
| "logits/chosen": -0.16576483845710754, |
| "logits/rejected": -0.3216527998447418, |
| "logps/chosen": -3.5259528160095215, |
| "logps/rejected": -34.49477005004883, |
| "loss": 0.6575559377670288, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.324476957321167, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.0956890657544136, |
| "rewards/margins": 2.1255316734313965, |
| "rewards/rejected": -2.0298423767089844, |
| "step": 92, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.4161073825503356, |
| "grad_norm": 4.184682846069336, |
| "learning_rate": 0.00018678761615977468, |
| "logits/chosen": -0.08753462135791779, |
| "logits/rejected": -0.2798512876033783, |
| "logps/chosen": -9.157657623291016, |
| "logps/rejected": -36.74687576293945, |
| "loss": 1.0873180627822876, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7294520139694214, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.07273296266794205, |
| "rewards/margins": 1.809075117111206, |
| "rewards/rejected": -1.736342430114746, |
| "step": 93, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.42058165548098436, |
| "grad_norm": 3.6453614234924316, |
| "learning_rate": 0.00018641627036388169, |
| "logits/chosen": -0.26587241888046265, |
| "logits/rejected": -0.35532236099243164, |
| "logps/chosen": -6.705877780914307, |
| "logps/rejected": -34.05841827392578, |
| "loss": 0.8682178258895874, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5556175112724304, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0699356198310852, |
| "rewards/margins": 2.082659959793091, |
| "rewards/rejected": -2.0127243995666504, |
| "step": 94, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.4250559284116331, |
| "grad_norm": 3.926342725753784, |
| "learning_rate": 0.00018604015792601396, |
| "logits/chosen": -0.18198826909065247, |
| "logits/rejected": -0.28424450755119324, |
| "logps/chosen": -9.228765487670898, |
| "logps/rejected": -28.494712829589844, |
| "loss": 1.100342869758606, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6449709534645081, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0022647292353212833, |
| "rewards/margins": 1.3801387548446655, |
| "rewards/rejected": -1.3778741359710693, |
| "step": 95, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.42953020134228187, |
| "grad_norm": 6.2491960525512695, |
| "learning_rate": 0.00018565929959218758, |
| "logits/chosen": -0.162868469953537, |
| "logits/rejected": -0.35664084553718567, |
| "logps/chosen": -5.505702495574951, |
| "logps/rejected": -40.82157516479492, |
| "loss": 0.8373050689697266, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.48504531383514404, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.03489302098751068, |
| "rewards/margins": 1.9277093410491943, |
| "rewards/rejected": -1.892816424369812, |
| "step": 96, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.43400447427293065, |
| "grad_norm": 3.155294895172119, |
| "learning_rate": 0.0001852737163701979, |
| "logits/chosen": -0.14200377464294434, |
| "logits/rejected": -0.3238810896873474, |
| "logps/chosen": -6.1366353034973145, |
| "logps/rejected": -36.68225860595703, |
| "loss": 0.7804915904998779, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.47483596205711365, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.05882677063345909, |
| "rewards/margins": 1.9737701416015625, |
| "rewards/rejected": -1.9149433374404907, |
| "step": 97, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.43847874720357943, |
| "grad_norm": 2.6832268238067627, |
| "learning_rate": 0.00018488342952846073, |
| "logits/chosen": -0.12941551208496094, |
| "logits/rejected": -0.2959509789943695, |
| "logps/chosen": -12.032288551330566, |
| "logps/rejected": -35.381710052490234, |
| "loss": 0.838473916053772, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.49494707584381104, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.032307837158441544, |
| "rewards/margins": 1.5512968301773071, |
| "rewards/rejected": -1.5189889669418335, |
| "step": 98, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.4429530201342282, |
| "grad_norm": 2.0462474822998047, |
| "learning_rate": 0.0001844884605948392, |
| "logits/chosen": -0.23558180034160614, |
| "logits/rejected": -0.3692886531352997, |
| "logps/chosen": -9.6895112991333, |
| "logps/rejected": -26.33759117126465, |
| "loss": 1.0117727518081665, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6607507467269897, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.148264542222023, |
| "rewards/margins": 1.4499516487121582, |
| "rewards/rejected": -1.3016871213912964, |
| "step": 99, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.44742729306487694, |
| "grad_norm": 2.056586503982544, |
| "learning_rate": 0.00018408883135545632, |
| "logits/chosen": -0.20895381271839142, |
| "logits/rejected": -0.350538432598114, |
| "logps/chosen": -7.372556209564209, |
| "logps/rejected": -30.752912521362305, |
| "loss": 0.8857133388519287, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.49547722935676575, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0997147411108017, |
| "rewards/margins": 1.3622983694076538, |
| "rewards/rejected": -1.262583613395691, |
| "step": 100, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.4519015659955257, |
| "grad_norm": 2.2281575202941895, |
| "learning_rate": 0.00018368456385349334, |
| "logits/chosen": -0.22281751036643982, |
| "logits/rejected": -0.34848061203956604, |
| "logps/chosen": -6.225516319274902, |
| "logps/rejected": -27.612606048583984, |
| "loss": 0.9511269330978394, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5547913312911987, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.09937570989131927, |
| "rewards/margins": 1.3068878650665283, |
| "rewards/rejected": -1.2075122594833374, |
| "step": 101, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.4563758389261745, |
| "grad_norm": 2.278520345687866, |
| "learning_rate": 0.0001832756803879737, |
| "logits/chosen": -0.20784702897071838, |
| "logits/rejected": -0.3433951139450073, |
| "logps/chosen": -9.43735122680664, |
| "logps/rejected": -25.318925857543945, |
| "loss": 0.8974735140800476, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5149669647216797, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.10113431513309479, |
| "rewards/margins": 1.3135309219360352, |
| "rewards/rejected": -1.2123966217041016, |
| "step": 102, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.4608501118568233, |
| "grad_norm": 2.4370529651641846, |
| "learning_rate": 0.0001828622035125332, |
| "logits/chosen": -0.21997469663619995, |
| "logits/rejected": -0.3634868562221527, |
| "logps/chosen": -6.314621925354004, |
| "logps/rejected": -29.342872619628906, |
| "loss": 0.8530542254447937, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5132930874824524, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12123695015907288, |
| "rewards/margins": 1.667711615562439, |
| "rewards/rejected": -1.546474575996399, |
| "step": 103, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.465324384787472, |
| "grad_norm": 2.014124870300293, |
| "learning_rate": 0.00018244415603417603, |
| "logits/chosen": -0.24184371531009674, |
| "logits/rejected": -0.36744868755340576, |
| "logps/chosen": -9.540416717529297, |
| "logps/rejected": -29.473955154418945, |
| "loss": 0.9970014095306396, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6495711803436279, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.11495855450630188, |
| "rewards/margins": 1.480215311050415, |
| "rewards/rejected": -1.365256667137146, |
| "step": 104, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.4697986577181208, |
| "grad_norm": 3.0877394676208496, |
| "learning_rate": 0.00018202156101201645, |
| "logits/chosen": -0.3040401041507721, |
| "logits/rejected": -0.3883563280105591, |
| "logps/chosen": -5.42517614364624, |
| "logps/rejected": -24.144115447998047, |
| "loss": 0.8570873737335205, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.447566956281662, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.056889522820711136, |
| "rewards/margins": 1.262117624282837, |
| "rewards/rejected": -1.2052280902862549, |
| "step": 105, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.4742729306487696, |
| "grad_norm": 2.858959674835205, |
| "learning_rate": 0.00018159444175600703, |
| "logits/chosen": -0.2462853342294693, |
| "logits/rejected": -0.3662327527999878, |
| "logps/chosen": -9.698307991027832, |
| "logps/rejected": -28.979686737060547, |
| "loss": 0.9490512609481812, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5636053681373596, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.21260645985603333, |
| "rewards/margins": 1.5004974603652954, |
| "rewards/rejected": -1.2878910303115845, |
| "step": 106, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.47874720357941836, |
| "grad_norm": 2.4962282180786133, |
| "learning_rate": 0.00018116282182565311, |
| "logits/chosen": -0.33677104115486145, |
| "logits/rejected": -0.3884970545768738, |
| "logps/chosen": -6.480227947235107, |
| "logps/rejected": -24.565458297729492, |
| "loss": 0.779257595539093, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4241567850112915, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.19994103908538818, |
| "rewards/margins": 1.3964340686798096, |
| "rewards/rejected": -1.196493148803711, |
| "step": 107, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.48322147651006714, |
| "grad_norm": 5.664783000946045, |
| "learning_rate": 0.00018072672502871296, |
| "logits/chosen": -0.25596198439598083, |
| "logits/rejected": -0.39605844020843506, |
| "logps/chosen": -5.608950138092041, |
| "logps/rejected": -35.774356842041016, |
| "loss": 0.7417677640914917, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4080236554145813, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.06679990887641907, |
| "rewards/margins": 1.873131275177002, |
| "rewards/rejected": -1.8063313961029053, |
| "step": 108, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.48769574944071586, |
| "grad_norm": 2.3131730556488037, |
| "learning_rate": 0.00018028617541988472, |
| "logits/chosen": -0.24975106120109558, |
| "logits/rejected": -0.3782438635826111, |
| "logps/chosen": -7.196217060089111, |
| "logps/rejected": -30.633535385131836, |
| "loss": 0.8902066946029663, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5435032844543457, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.005009647458791733, |
| "rewards/margins": 1.4205687046051025, |
| "rewards/rejected": -1.4255784749984741, |
| "step": 109, |
| "train_speed(iter/s)": 0.005618 |
| }, |
| { |
| "epoch": 0.49217002237136465, |
| "grad_norm": 1.9771348237991333, |
| "learning_rate": 0.00017984119729947944, |
| "logits/chosen": -0.24366146326065063, |
| "logits/rejected": -0.3742184340953827, |
| "logps/chosen": -7.5738325119018555, |
| "logps/rejected": -36.54035186767578, |
| "loss": 0.8050943613052368, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5340873003005981, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.21246597170829773, |
| "rewards/margins": 2.065727710723877, |
| "rewards/rejected": -1.853261947631836, |
| "step": 110, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.4966442953020134, |
| "grad_norm": 2.3881213665008545, |
| "learning_rate": 0.000179391815212081, |
| "logits/chosen": -0.2847563922405243, |
| "logits/rejected": -0.41209399700164795, |
| "logps/chosen": -7.6792683601379395, |
| "logps/rejected": -35.70549011230469, |
| "loss": 0.8006418347358704, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.482678085565567, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.05708705633878708, |
| "rewards/margins": 1.9757572412490845, |
| "rewards/rejected": -1.9186700582504272, |
| "step": 111, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5011185682326622, |
| "grad_norm": 2.492396116256714, |
| "learning_rate": 0.0001789380539451919, |
| "logits/chosen": -0.28943008184432983, |
| "logits/rejected": -0.36423763632774353, |
| "logps/chosen": -11.887781143188477, |
| "logps/rejected": -28.29849624633789, |
| "loss": 1.1252532005310059, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.730536162853241, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.11893969774246216, |
| "rewards/margins": 1.626894474029541, |
| "rewards/rejected": -1.5079548358917236, |
| "step": 112, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5055928411633109, |
| "grad_norm": 2.185915946960449, |
| "learning_rate": 0.0001784799385278661, |
| "logits/chosen": -0.27568331360816956, |
| "logits/rejected": -0.40015357732772827, |
| "logps/chosen": -6.916875839233398, |
| "logps/rejected": -33.39084243774414, |
| "loss": 0.7926216125488281, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4541390538215637, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.04535483196377754, |
| "rewards/margins": 1.5393131971359253, |
| "rewards/rejected": -1.4939584732055664, |
| "step": 113, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5100671140939598, |
| "grad_norm": 2.4158413410186768, |
| "learning_rate": 0.0001780174942293287, |
| "logits/chosen": -0.3091110587120056, |
| "logits/rejected": -0.40796011686325073, |
| "logps/chosen": -6.973065376281738, |
| "logps/rejected": -28.748104095458984, |
| "loss": 0.9577688574790955, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6081997156143188, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1668429970741272, |
| "rewards/margins": 1.543600082397461, |
| "rewards/rejected": -1.376757025718689, |
| "step": 114, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5145413870246085, |
| "grad_norm": 1.9534834623336792, |
| "learning_rate": 0.00017755074655758174, |
| "logits/chosen": -0.2860637605190277, |
| "logits/rejected": -0.407674103975296, |
| "logps/chosen": -7.297882556915283, |
| "logps/rejected": -31.946149826049805, |
| "loss": 0.8219610452651978, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5011268854141235, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18072012066841125, |
| "rewards/margins": 1.9877647161483765, |
| "rewards/rejected": -1.807044267654419, |
| "step": 115, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5190156599552572, |
| "grad_norm": 4.979251384735107, |
| "learning_rate": 0.00017707972125799735, |
| "logits/chosen": -0.3695046305656433, |
| "logits/rejected": -0.4696848392486572, |
| "logps/chosen": -6.748042583465576, |
| "logps/rejected": -28.866714477539062, |
| "loss": 0.8665784597396851, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5186231732368469, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.20859460532665253, |
| "rewards/margins": 1.632576584815979, |
| "rewards/rejected": -1.4239819049835205, |
| "step": 116, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5234899328859061, |
| "grad_norm": 2.385962963104248, |
| "learning_rate": 0.0001766044443118978, |
| "logits/chosen": -0.3143601715564728, |
| "logits/rejected": -0.3878296911716461, |
| "logps/chosen": -8.803234100341797, |
| "logps/rejected": -24.038074493408203, |
| "loss": 1.0012811422348022, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5274536609649658, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05610959604382515, |
| "rewards/margins": 1.1446653604507446, |
| "rewards/rejected": -1.0885558128356934, |
| "step": 117, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.5279642058165548, |
| "grad_norm": 4.134023189544678, |
| "learning_rate": 0.0001761249419351222, |
| "logits/chosen": -0.3163571357727051, |
| "logits/rejected": -0.4659003019332886, |
| "logps/chosen": -5.800329208374023, |
| "logps/rejected": -33.57259750366211, |
| "loss": 0.824342668056488, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4846615791320801, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.10410095751285553, |
| "rewards/margins": 1.8639620542526245, |
| "rewards/rejected": -1.7598611116409302, |
| "step": 118, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5324384787472036, |
| "grad_norm": 2.1616837978363037, |
| "learning_rate": 0.00017564124057658056, |
| "logits/chosen": -0.32389989495277405, |
| "logits/rejected": -0.4355754852294922, |
| "logps/chosen": -9.749198913574219, |
| "logps/rejected": -28.6905517578125, |
| "loss": 0.9396259784698486, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5555200576782227, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.15960067510604858, |
| "rewards/margins": 1.569477915763855, |
| "rewards/rejected": -1.4098771810531616, |
| "step": 119, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5369127516778524, |
| "grad_norm": 6.066991806030273, |
| "learning_rate": 0.00017515336691679477, |
| "logits/chosen": -0.3255332112312317, |
| "logits/rejected": -0.4361691474914551, |
| "logps/chosen": -5.765564918518066, |
| "logps/rejected": -29.787982940673828, |
| "loss": 0.8768988847732544, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4620920717716217, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.06438101083040237, |
| "rewards/margins": 1.6063313484191895, |
| "rewards/rejected": -1.5419504642486572, |
| "step": 120, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5413870246085011, |
| "grad_norm": 3.371220827102661, |
| "learning_rate": 0.0001746613478664271, |
| "logits/chosen": -0.3099900186061859, |
| "logits/rejected": -0.3865930438041687, |
| "logps/chosen": -9.693756103515625, |
| "logps/rejected": -29.932273864746094, |
| "loss": 0.9306422472000122, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5374558568000793, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.13387000560760498, |
| "rewards/margins": 1.695232629776001, |
| "rewards/rejected": -1.5613627433776855, |
| "step": 121, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5458612975391499, |
| "grad_norm": 2.195613384246826, |
| "learning_rate": 0.00017416521056479577, |
| "logits/chosen": -0.26501592993736267, |
| "logits/rejected": -0.42681097984313965, |
| "logps/chosen": -6.962400436401367, |
| "logps/rejected": -33.85113525390625, |
| "loss": 0.81517094373703, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4922808110713959, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1830664575099945, |
| "rewards/margins": 2.055619478225708, |
| "rewards/rejected": -1.8725531101226807, |
| "step": 122, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5503355704697986, |
| "grad_norm": 2.068668842315674, |
| "learning_rate": 0.0001736649823783779, |
| "logits/chosen": -0.30316585302352905, |
| "logits/rejected": -0.45119816064834595, |
| "logps/chosen": -7.222802639007568, |
| "logps/rejected": -36.19852066040039, |
| "loss": 0.8215571045875549, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5099644660949707, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.15236249566078186, |
| "rewards/margins": 2.028899669647217, |
| "rewards/rejected": -1.8765370845794678, |
| "step": 123, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.5548098434004475, |
| "grad_norm": 3.336085557937622, |
| "learning_rate": 0.00017316069089930007, |
| "logits/chosen": -0.30727842450141907, |
| "logits/rejected": -0.440543532371521, |
| "logps/chosen": -7.512927532196045, |
| "logps/rejected": -35.85497283935547, |
| "loss": 0.913392186164856, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5231627821922302, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.062082529067993164, |
| "rewards/margins": 1.9021214246749878, |
| "rewards/rejected": -1.840038776397705, |
| "step": 124, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5592841163310962, |
| "grad_norm": 2.1725635528564453, |
| "learning_rate": 0.00017265236394381633, |
| "logits/chosen": -0.32654696702957153, |
| "logits/rejected": -0.4225502014160156, |
| "logps/chosen": -7.848655700683594, |
| "logps/rejected": -28.869049072265625, |
| "loss": 0.9009624123573303, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4998895227909088, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.15675391256809235, |
| "rewards/margins": 1.4675108194351196, |
| "rewards/rejected": -1.3107569217681885, |
| "step": 125, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5637583892617449, |
| "grad_norm": 3.518026113510132, |
| "learning_rate": 0.00017214002955077393, |
| "logits/chosen": -0.28199663758277893, |
| "logits/rejected": -0.3950486481189728, |
| "logps/chosen": -6.079558849334717, |
| "logps/rejected": -29.816913604736328, |
| "loss": 0.8342326283454895, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.49542737007141113, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.1437724232673645, |
| "rewards/margins": 1.676330804824829, |
| "rewards/rejected": -1.5325582027435303, |
| "step": 126, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5682326621923938, |
| "grad_norm": 2.908057928085327, |
| "learning_rate": 0.00017162371598006666, |
| "logits/chosen": -0.2865302562713623, |
| "logits/rejected": -0.39790090918540955, |
| "logps/chosen": -8.84012508392334, |
| "logps/rejected": -25.36516761779785, |
| "loss": 0.9509469866752625, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5248587131500244, |
| "rewards/accuracies": 0.734375, |
| "rewards/chosen": 0.13465049862861633, |
| "rewards/margins": 1.1559473276138306, |
| "rewards/rejected": -1.0212968587875366, |
| "step": 127, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5727069351230425, |
| "grad_norm": 3.0232553482055664, |
| "learning_rate": 0.0001711034517110761, |
| "logits/chosen": -0.29277676343917847, |
| "logits/rejected": -0.42875176668167114, |
| "logps/chosen": -5.9850897789001465, |
| "logps/rejected": -23.161203384399414, |
| "loss": 0.9273078441619873, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.49909183382987976, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.23866456747055054, |
| "rewards/margins": 1.1174073219299316, |
| "rewards/rejected": -0.8787426352500916, |
| "step": 128, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5771812080536913, |
| "grad_norm": 2.585862636566162, |
| "learning_rate": 0.0001705792654411007, |
| "logits/chosen": -0.28234124183654785, |
| "logits/rejected": -0.43345722556114197, |
| "logps/chosen": -6.964176177978516, |
| "logps/rejected": -29.49521827697754, |
| "loss": 0.8882912397384644, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5149023532867432, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.13531282544136047, |
| "rewards/margins": 1.5991878509521484, |
| "rewards/rejected": -1.4638750553131104, |
| "step": 129, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.5816554809843401, |
| "grad_norm": 1.718246579170227, |
| "learning_rate": 0.00017005118608377288, |
| "logits/chosen": -0.25894033908843994, |
| "logits/rejected": -0.35234570503234863, |
| "logps/chosen": -7.487059116363525, |
| "logps/rejected": -31.42106056213379, |
| "loss": 0.8475052714347839, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4482796788215637, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.10318747907876968, |
| "rewards/margins": 1.6407992839813232, |
| "rewards/rejected": -1.537611722946167, |
| "step": 130, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.5861297539149888, |
| "grad_norm": 1.766929268836975, |
| "learning_rate": 0.00016951924276746425, |
| "logits/chosen": -0.2688104212284088, |
| "logits/rejected": -0.3982466757297516, |
| "logps/chosen": -4.793079853057861, |
| "logps/rejected": -32.5130615234375, |
| "loss": 0.6818744540214539, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3626863360404968, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.21576498448848724, |
| "rewards/margins": 2.161926746368408, |
| "rewards/rejected": -1.9461617469787598, |
| "step": 131, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.5906040268456376, |
| "grad_norm": 2.9491937160491943, |
| "learning_rate": 0.00016898346483367867, |
| "logits/chosen": -0.21997107565402985, |
| "logits/rejected": -0.4120599627494812, |
| "logps/chosen": -5.662740230560303, |
| "logps/rejected": -35.41563034057617, |
| "loss": 0.8450470566749573, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4608619809150696, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.16963335871696472, |
| "rewards/margins": 2.1595871448516846, |
| "rewards/rejected": -1.989953637123108, |
| "step": 132, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.5950782997762863, |
| "grad_norm": 3.147080659866333, |
| "learning_rate": 0.00016844388183543418, |
| "logits/chosen": -0.24002769589424133, |
| "logits/rejected": -0.40396854281425476, |
| "logps/chosen": -4.948556423187256, |
| "logps/rejected": -46.72270965576172, |
| "loss": 0.6112457513809204, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.35614243149757385, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.04689760506153107, |
| "rewards/margins": 3.0640969276428223, |
| "rewards/rejected": -3.0171995162963867, |
| "step": 133, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.5995525727069351, |
| "grad_norm": 3.0668933391571045, |
| "learning_rate": 0.00016790052353563253, |
| "logits/chosen": -0.23955634236335754, |
| "logits/rejected": -0.4039704203605652, |
| "logps/chosen": -7.274659156799316, |
| "logps/rejected": -42.970394134521484, |
| "loss": 0.8683923482894897, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5945435762405396, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08974725753068924, |
| "rewards/margins": 2.8205695152282715, |
| "rewards/rejected": -2.7308225631713867, |
| "step": 134, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.6040268456375839, |
| "grad_norm": 5.152528762817383, |
| "learning_rate": 0.00016735341990541764, |
| "logits/chosen": -0.2829990088939667, |
| "logits/rejected": -0.3635609447956085, |
| "logps/chosen": -13.53715705871582, |
| "logps/rejected": -37.386192321777344, |
| "loss": 1.2928409576416016, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.8582869172096252, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.11900941282510757, |
| "rewards/margins": 1.9026870727539062, |
| "rewards/rejected": -2.0216965675354004, |
| "step": 135, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.6085011185682326, |
| "grad_norm": 3.047576665878296, |
| "learning_rate": 0.0001668026011225225, |
| "logits/chosen": -0.1925339251756668, |
| "logits/rejected": -0.3699888586997986, |
| "logps/chosen": -7.0672407150268555, |
| "logps/rejected": -38.66764831542969, |
| "loss": 0.8304687738418579, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5318548083305359, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.06892228871583939, |
| "rewards/margins": 2.3139286041259766, |
| "rewards/rejected": -2.245006561279297, |
| "step": 136, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.6129753914988815, |
| "grad_norm": 2.2745347023010254, |
| "learning_rate": 0.00016624809756960444, |
| "logits/chosen": -0.11182728409767151, |
| "logits/rejected": -0.2984737455844879, |
| "logps/chosen": -5.8051066398620605, |
| "logps/rejected": -31.151941299438477, |
| "loss": 0.7885230779647827, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.38472679257392883, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11600922793149948, |
| "rewards/margins": 1.5462919473648071, |
| "rewards/rejected": -1.430282711982727, |
| "step": 137, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.6174496644295302, |
| "grad_norm": 2.21187162399292, |
| "learning_rate": 0.0001656899398325693, |
| "logits/chosen": -0.17814204096794128, |
| "logits/rejected": -0.28958842158317566, |
| "logps/chosen": -11.657821655273438, |
| "logps/rejected": -28.357452392578125, |
| "loss": 1.0776360034942627, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6752557754516602, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.15066924691200256, |
| "rewards/margins": 1.488224744796753, |
| "rewards/rejected": -1.3375554084777832, |
| "step": 138, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.6219239373601789, |
| "grad_norm": 2.826819896697998, |
| "learning_rate": 0.0001651281586988844, |
| "logits/chosen": -0.2112751007080078, |
| "logits/rejected": -0.3410148024559021, |
| "logps/chosen": -6.036767482757568, |
| "logps/rejected": -22.628429412841797, |
| "loss": 0.9106030464172363, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4903098940849304, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.15719373524188995, |
| "rewards/margins": 1.1676839590072632, |
| "rewards/rejected": -1.0104901790618896, |
| "step": 139, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.6263982102908278, |
| "grad_norm": 1.9086319208145142, |
| "learning_rate": 0.00016456278515588024, |
| "logits/chosen": -0.17578895390033722, |
| "logits/rejected": -0.2394012212753296, |
| "logps/chosen": -7.572815895080566, |
| "logps/rejected": -27.15605926513672, |
| "loss": 0.9325375556945801, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5454995036125183, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.1564534306526184, |
| "rewards/margins": 1.4798097610473633, |
| "rewards/rejected": -1.3233562707901, |
| "step": 140, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.6308724832214765, |
| "grad_norm": 2.0129153728485107, |
| "learning_rate": 0.00016399385038904138, |
| "logits/chosen": -0.15921340882778168, |
| "logits/rejected": -0.3579505681991577, |
| "logps/chosen": -4.274261951446533, |
| "logps/rejected": -33.25766372680664, |
| "loss": 0.767472505569458, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40454328060150146, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.07659953087568283, |
| "rewards/margins": 1.6252936124801636, |
| "rewards/rejected": -1.5486938953399658, |
| "step": 141, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6353467561521253, |
| "grad_norm": 2.377077579498291, |
| "learning_rate": 0.00016342138578028613, |
| "logits/chosen": -0.0892738476395607, |
| "logits/rejected": -0.2357676923274994, |
| "logps/chosen": -7.721086502075195, |
| "logps/rejected": -30.33781623840332, |
| "loss": 0.948746919631958, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5438387393951416, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.15941840410232544, |
| "rewards/margins": 1.4621765613555908, |
| "rewards/rejected": -1.3027582168579102, |
| "step": 142, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.639821029082774, |
| "grad_norm": 2.5466043949127197, |
| "learning_rate": 0.00016284542290623567, |
| "logits/chosen": -0.2347170114517212, |
| "logits/rejected": -0.37242478132247925, |
| "logps/chosen": -6.517569065093994, |
| "logps/rejected": -34.21914291381836, |
| "loss": 0.7467401027679443, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4813250005245209, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.12270703911781311, |
| "rewards/margins": 2.2491090297698975, |
| "rewards/rejected": -2.1264021396636963, |
| "step": 143, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6442953020134228, |
| "grad_norm": 2.4996299743652344, |
| "learning_rate": 0.00016226599353647228, |
| "logits/chosen": -0.21299152076244354, |
| "logits/rejected": -0.3434564769268036, |
| "logps/chosen": -8.105118751525879, |
| "logps/rejected": -33.501346588134766, |
| "loss": 0.8418401479721069, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4816139042377472, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0775795578956604, |
| "rewards/margins": 1.7935457229614258, |
| "rewards/rejected": -1.7159662246704102, |
| "step": 144, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6487695749440716, |
| "grad_norm": 2.852815866470337, |
| "learning_rate": 0.00016168312963178697, |
| "logits/chosen": -0.14896008372306824, |
| "logits/rejected": -0.29324355721473694, |
| "logps/chosen": -8.402843475341797, |
| "logps/rejected": -30.785614013671875, |
| "loss": 0.966815173625946, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6201918721199036, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.06825412809848785, |
| "rewards/margins": 1.6479313373565674, |
| "rewards/rejected": -1.5796774625778198, |
| "step": 145, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6532438478747203, |
| "grad_norm": 3.7648990154266357, |
| "learning_rate": 0.00016109686334241655, |
| "logits/chosen": -0.19154362380504608, |
| "logits/rejected": -0.3218660056591034, |
| "logps/chosen": -10.407678604125977, |
| "logps/rejected": -36.2245979309082, |
| "loss": 0.9397022128105164, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6243451833724976, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.07954104989767075, |
| "rewards/margins": 2.1638622283935547, |
| "rewards/rejected": -2.0843210220336914, |
| "step": 146, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6577181208053692, |
| "grad_norm": 5.735119819641113, |
| "learning_rate": 0.00016050722700627012, |
| "logits/chosen": -0.18544235825538635, |
| "logits/rejected": -0.38329729437828064, |
| "logps/chosen": -5.036187171936035, |
| "logps/rejected": -37.17473602294922, |
| "loss": 0.7683266401290894, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4119165539741516, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.026729077100753784, |
| "rewards/margins": 1.973728895187378, |
| "rewards/rejected": -1.9470000267028809, |
| "step": 147, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6621923937360179, |
| "grad_norm": 2.958143472671509, |
| "learning_rate": 0.0001599142531471456, |
| "logits/chosen": -0.22263091802597046, |
| "logits/rejected": -0.3673292100429535, |
| "logps/chosen": -8.627145767211914, |
| "logps/rejected": -34.2421875, |
| "loss": 0.8043665885925293, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4987104833126068, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.11155197024345398, |
| "rewards/margins": 1.846522569656372, |
| "rewards/rejected": -1.7349706888198853, |
| "step": 148, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 1.810313105583191, |
| "learning_rate": 0.00015931797447293552, |
| "logits/chosen": -0.25553733110427856, |
| "logits/rejected": -0.38629549741744995, |
| "logps/chosen": -5.752464294433594, |
| "logps/rejected": -25.610708236694336, |
| "loss": 0.7433949708938599, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3995458483695984, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.1192181408405304, |
| "rewards/margins": 1.3859959840774536, |
| "rewards/rejected": -1.266777753829956, |
| "step": 149, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 0.6711409395973155, |
| "grad_norm": 4.102981090545654, |
| "learning_rate": 0.00015871842387382305, |
| "logits/chosen": -0.2806379497051239, |
| "logits/rejected": -0.390668660402298, |
| "logps/chosen": -15.568105697631836, |
| "logps/rejected": -25.584888458251953, |
| "loss": 1.1545464992523193, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6800742745399475, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.030628615990281105, |
| "rewards/margins": 1.1961755752563477, |
| "rewards/rejected": -1.1655468940734863, |
| "step": 150, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 0.6756152125279642, |
| "grad_norm": 2.4007022380828857, |
| "learning_rate": 0.00015811563442046767, |
| "logits/chosen": -0.21862031519412994, |
| "logits/rejected": -0.4358934164047241, |
| "logps/chosen": -7.3923821449279785, |
| "logps/rejected": -38.984134674072266, |
| "loss": 0.8094646334648132, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.510733425617218, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.1613074541091919, |
| "rewards/margins": 2.380244016647339, |
| "rewards/rejected": -2.2189366817474365, |
| "step": 151, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.680089485458613, |
| "grad_norm": 1.7411408424377441, |
| "learning_rate": 0.00015750963936218105, |
| "logits/chosen": -0.31933140754699707, |
| "logits/rejected": -0.47491735219955444, |
| "logps/chosen": -6.779457092285156, |
| "logps/rejected": -43.27799987792969, |
| "loss": 0.8203473091125488, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4936409592628479, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09017263352870941, |
| "rewards/margins": 2.683919906616211, |
| "rewards/rejected": -2.593747138977051, |
| "step": 152, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.6845637583892618, |
| "grad_norm": 4.146765232086182, |
| "learning_rate": 0.00015690047212509316, |
| "logits/chosen": -0.37495219707489014, |
| "logits/rejected": -0.4795547127723694, |
| "logps/chosen": -7.767773628234863, |
| "logps/rejected": -37.96726608276367, |
| "loss": 0.8648288249969482, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5447298884391785, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08953765779733658, |
| "rewards/margins": 2.2388553619384766, |
| "rewards/rejected": -2.149317741394043, |
| "step": 153, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.6890380313199105, |
| "grad_norm": 2.1461992263793945, |
| "learning_rate": 0.00015628816631030836, |
| "logits/chosen": -0.32161566615104675, |
| "logits/rejected": -0.43094250559806824, |
| "logps/chosen": -8.232239723205566, |
| "logps/rejected": -36.47972106933594, |
| "loss": 0.7230132222175598, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.391425758600235, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12491930276155472, |
| "rewards/margins": 2.1603236198425293, |
| "rewards/rejected": -2.035404682159424, |
| "step": 154, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.6935123042505593, |
| "grad_norm": 4.374569416046143, |
| "learning_rate": 0.00015567275569205218, |
| "logits/chosen": -0.25436070561408997, |
| "logits/rejected": -0.45995038747787476, |
| "logps/chosen": -8.461811065673828, |
| "logps/rejected": -44.51617431640625, |
| "loss": 0.9842564463615417, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6817278861999512, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.040151700377464294, |
| "rewards/margins": 2.510289430618286, |
| "rewards/rejected": -2.47013783454895, |
| "step": 155, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.697986577181208, |
| "grad_norm": 4.502596378326416, |
| "learning_rate": 0.00015505427421580808, |
| "logits/chosen": -0.4160749316215515, |
| "logits/rejected": -0.506524384021759, |
| "logps/chosen": -8.063937187194824, |
| "logps/rejected": -32.82113265991211, |
| "loss": 0.8426942825317383, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5383368730545044, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.11457733064889908, |
| "rewards/margins": 2.155301094055176, |
| "rewards/rejected": -2.0407238006591797, |
| "step": 156, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 0.7024608501118568, |
| "grad_norm": 3.176142692565918, |
| "learning_rate": 0.00015443275599644538, |
| "logits/chosen": -0.42277562618255615, |
| "logits/rejected": -0.5972986817359924, |
| "logps/chosen": -5.387449264526367, |
| "logps/rejected": -45.159637451171875, |
| "loss": 0.6948708295822144, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4455086886882782, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1367429941892624, |
| "rewards/margins": 3.0624241828918457, |
| "rewards/rejected": -2.9256811141967773, |
| "step": 157, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7069351230425056, |
| "grad_norm": 3.3241329193115234, |
| "learning_rate": 0.00015380823531633729, |
| "logits/chosen": -0.3996984660625458, |
| "logits/rejected": -0.5501067042350769, |
| "logps/chosen": -6.42928409576416, |
| "logps/rejected": -34.65045166015625, |
| "loss": 0.7898067235946655, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.474100798368454, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.12420983612537384, |
| "rewards/margins": 2.1587226390838623, |
| "rewards/rejected": -2.034512758255005, |
| "step": 158, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7114093959731543, |
| "grad_norm": 3.497190237045288, |
| "learning_rate": 0.00015318074662346994, |
| "logits/chosen": -0.41886377334594727, |
| "logits/rejected": -0.5915812849998474, |
| "logps/chosen": -7.062689781188965, |
| "logps/rejected": -42.2303466796875, |
| "loss": 0.7893521785736084, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.46673545241355896, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.0015753507614135742, |
| "rewards/margins": 2.453634738922119, |
| "rewards/rejected": -2.452059745788574, |
| "step": 159, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7158836689038032, |
| "grad_norm": 2.7327327728271484, |
| "learning_rate": 0.00015255032452954245, |
| "logits/chosen": -0.3371175527572632, |
| "logits/rejected": -0.5264066457748413, |
| "logps/chosen": -7.435405731201172, |
| "logps/rejected": -50.61220932006836, |
| "loss": 0.7223600149154663, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4396481215953827, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.025311682373285294, |
| "rewards/margins": 3.0541183948516846, |
| "rewards/rejected": -3.028806447982788, |
| "step": 160, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7203579418344519, |
| "grad_norm": 4.352429389953613, |
| "learning_rate": 0.00015191700380805752, |
| "logits/chosen": -0.44862544536590576, |
| "logits/rejected": -0.5382689833641052, |
| "logps/chosen": -10.93178939819336, |
| "logps/rejected": -35.30765151977539, |
| "loss": 0.9672499895095825, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5727404356002808, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.02203945443034172, |
| "rewards/margins": 2.1007864475250244, |
| "rewards/rejected": -2.122826099395752, |
| "step": 161, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7248322147651006, |
| "grad_norm": 4.718836784362793, |
| "learning_rate": 0.00015128081939240357, |
| "logits/chosen": -0.3789413869380951, |
| "logits/rejected": -0.5269034504890442, |
| "logps/chosen": -6.491231441497803, |
| "logps/rejected": -39.26808547973633, |
| "loss": 0.8075002431869507, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4541192948818207, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.02522008866071701, |
| "rewards/margins": 2.450805187225342, |
| "rewards/rejected": -2.4255850315093994, |
| "step": 162, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7293064876957495, |
| "grad_norm": 3.2440578937530518, |
| "learning_rate": 0.00015064180637392764, |
| "logits/chosen": -0.3883664608001709, |
| "logits/rejected": -0.5545072555541992, |
| "logps/chosen": -4.190845012664795, |
| "logps/rejected": -28.679349899291992, |
| "loss": 0.7410864233970642, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3713621199131012, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.13914553821086884, |
| "rewards/margins": 1.7679030895233154, |
| "rewards/rejected": -1.6287575960159302, |
| "step": 163, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7337807606263982, |
| "grad_norm": 2.2508671283721924, |
| "learning_rate": 0.00015000000000000001, |
| "logits/chosen": -0.41260772943496704, |
| "logits/rejected": -0.5085133910179138, |
| "logps/chosen": -9.365473747253418, |
| "logps/rejected": -26.65607452392578, |
| "loss": 0.8695790767669678, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5267202258110046, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.19111280143260956, |
| "rewards/margins": 1.517545461654663, |
| "rewards/rejected": -1.3264328241348267, |
| "step": 164, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.738255033557047, |
| "grad_norm": 1.987011194229126, |
| "learning_rate": 0.00014935543567206984, |
| "logits/chosen": -0.4246593117713928, |
| "logits/rejected": -0.5221028327941895, |
| "logps/chosen": -7.673718452453613, |
| "logps/rejected": -29.984445571899414, |
| "loss": 0.9060878157615662, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5306745171546936, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.12876029312610626, |
| "rewards/margins": 1.761435627937317, |
| "rewards/rejected": -1.6326754093170166, |
| "step": 165, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7427293064876958, |
| "grad_norm": 2.439631700515747, |
| "learning_rate": 0.00014870814894371245, |
| "logits/chosen": -0.4619531035423279, |
| "logits/rejected": -0.5249654054641724, |
| "logps/chosen": -10.422987937927246, |
| "logps/rejected": -28.858501434326172, |
| "loss": 0.9730715155601501, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6119397282600403, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.0721518024802208, |
| "rewards/margins": 1.6050055027008057, |
| "rewards/rejected": -1.532853603363037, |
| "step": 166, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7472035794183445, |
| "grad_norm": 2.184906482696533, |
| "learning_rate": 0.00014805817551866838, |
| "logits/chosen": -0.4090336859226227, |
| "logits/rejected": -0.5729665160179138, |
| "logps/chosen": -6.691109657287598, |
| "logps/rejected": -31.740489959716797, |
| "loss": 0.7506605386734009, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4180562496185303, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.16018937528133392, |
| "rewards/margins": 1.805705189704895, |
| "rewards/rejected": -1.6455157995224, |
| "step": 167, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7516778523489933, |
| "grad_norm": 3.0194168090820312, |
| "learning_rate": 0.00014740555124887375, |
| "logits/chosen": -0.4370940029621124, |
| "logits/rejected": -0.5650765895843506, |
| "logps/chosen": -8.651978492736816, |
| "logps/rejected": -27.381715774536133, |
| "loss": 1.0154253244400024, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5850539207458496, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.10986748337745667, |
| "rewards/margins": 1.2420449256896973, |
| "rewards/rejected": -1.1321773529052734, |
| "step": 168, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.756152125279642, |
| "grad_norm": 2.3874940872192383, |
| "learning_rate": 0.00014675031213248296, |
| "logits/chosen": -0.47348707914352417, |
| "logits/rejected": -0.6095517873764038, |
| "logps/chosen": -6.874096870422363, |
| "logps/rejected": -32.03157424926758, |
| "loss": 0.9490684866905212, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5718879699707031, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.11709172278642654, |
| "rewards/margins": 1.9521950483322144, |
| "rewards/rejected": -1.8351032733917236, |
| "step": 169, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7606263982102909, |
| "grad_norm": 2.2490291595458984, |
| "learning_rate": 0.00014609249431188278, |
| "logits/chosen": -0.47727060317993164, |
| "logits/rejected": -0.629499077796936, |
| "logps/chosen": -9.573102951049805, |
| "logps/rejected": -40.56207275390625, |
| "loss": 0.8739838600158691, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5675899982452393, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.07386627048254013, |
| "rewards/margins": 2.4693443775177, |
| "rewards/rejected": -2.3954782485961914, |
| "step": 170, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7651006711409396, |
| "grad_norm": 2.2746880054473877, |
| "learning_rate": 0.0001454321340716992, |
| "logits/chosen": -0.4521310031414032, |
| "logits/rejected": -0.6028440594673157, |
| "logps/chosen": -6.765651702880859, |
| "logps/rejected": -34.42289352416992, |
| "loss": 0.8103550672531128, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4473325312137604, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.11047148704528809, |
| "rewards/margins": 2.000337600708008, |
| "rewards/rejected": -1.8898662328720093, |
| "step": 171, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7695749440715883, |
| "grad_norm": 3.201479196548462, |
| "learning_rate": 0.00014476926783679538, |
| "logits/chosen": -0.5274945497512817, |
| "logits/rejected": -0.6449275612831116, |
| "logps/chosen": -7.808686256408691, |
| "logps/rejected": -29.541404724121094, |
| "loss": 0.8972499370574951, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5542759895324707, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13792771100997925, |
| "rewards/margins": 1.9822808504104614, |
| "rewards/rejected": -1.8443530797958374, |
| "step": 172, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7740492170022372, |
| "grad_norm": 2.665815830230713, |
| "learning_rate": 0.00014410393217026318, |
| "logits/chosen": -0.4524228870868683, |
| "logits/rejected": -0.6132454872131348, |
| "logps/chosen": -10.15006160736084, |
| "logps/rejected": -41.52241516113281, |
| "loss": 0.8796108365058899, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5965855717658997, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.20403775572776794, |
| "rewards/margins": 2.6676454544067383, |
| "rewards/rejected": -2.4636075496673584, |
| "step": 173, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7785234899328859, |
| "grad_norm": 3.2595131397247314, |
| "learning_rate": 0.00014343616377140582, |
| "logits/chosen": -0.5274872779846191, |
| "logits/rejected": -0.6744863986968994, |
| "logps/chosen": -8.180425643920898, |
| "logps/rejected": -43.29051971435547, |
| "loss": 0.7626029253005981, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4893677234649658, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07820628583431244, |
| "rewards/margins": 2.6948142051696777, |
| "rewards/rejected": -2.616607904434204, |
| "step": 174, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7829977628635347, |
| "grad_norm": 3.828737735748291, |
| "learning_rate": 0.00014276599947371388, |
| "logits/chosen": -0.49092575907707214, |
| "logits/rejected": -0.6571353673934937, |
| "logps/chosen": -5.790119647979736, |
| "logps/rejected": -36.81107711791992, |
| "loss": 0.7012597918510437, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41794249415397644, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11556562781333923, |
| "rewards/margins": 2.311169385910034, |
| "rewards/rejected": -2.195603847503662, |
| "step": 175, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7874720357941835, |
| "grad_norm": 8.090876579284668, |
| "learning_rate": 0.0001420934762428335, |
| "logits/chosen": -0.47237420082092285, |
| "logits/rejected": -0.6282271146774292, |
| "logps/chosen": -8.150810241699219, |
| "logps/rejected": -43.93907928466797, |
| "loss": 0.943342924118042, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6251461505889893, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.03655552491545677, |
| "rewards/margins": 2.5700528621673584, |
| "rewards/rejected": -2.6066083908081055, |
| "step": 176, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.7919463087248322, |
| "grad_norm": 2.371922016143799, |
| "learning_rate": 0.00014141863117452745, |
| "logits/chosen": -0.5399549007415771, |
| "logits/rejected": -0.6778814196586609, |
| "logps/chosen": -5.152353763580322, |
| "logps/rejected": -44.89213562011719, |
| "loss": 0.6812291741371155, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40526461601257324, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.09105440229177475, |
| "rewards/margins": 3.134105682373047, |
| "rewards/rejected": -3.043051242828369, |
| "step": 177, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.796420581655481, |
| "grad_norm": 3.8354110717773438, |
| "learning_rate": 0.0001407415014926288, |
| "logits/chosen": -0.5333717465400696, |
| "logits/rejected": -0.6879783868789673, |
| "logps/chosen": -8.797846794128418, |
| "logps/rejected": -42.53145980834961, |
| "loss": 0.8743501901626587, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6057224273681641, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.014064118266105652, |
| "rewards/margins": 2.73359751701355, |
| "rewards/rejected": -2.7195334434509277, |
| "step": 178, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.8008948545861297, |
| "grad_norm": 2.9632174968719482, |
| "learning_rate": 0.00014006212454698797, |
| "logits/chosen": -0.5371872782707214, |
| "logits/rejected": -0.6701904535293579, |
| "logps/chosen": -8.69965648651123, |
| "logps/rejected": -35.25392150878906, |
| "loss": 0.9904757738113403, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7375240325927734, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.1597575545310974, |
| "rewards/margins": 2.451249599456787, |
| "rewards/rejected": -2.291491746902466, |
| "step": 179, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.8053691275167785, |
| "grad_norm": 3.797736406326294, |
| "learning_rate": 0.00013938053781141222, |
| "logits/chosen": -0.4403976798057556, |
| "logits/rejected": -0.6186584234237671, |
| "logps/chosen": -6.943508148193359, |
| "logps/rejected": -43.18886947631836, |
| "loss": 0.7113572955131531, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.47942519187927246, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.0965544655919075, |
| "rewards/margins": 2.6151046752929688, |
| "rewards/rejected": -2.518550157546997, |
| "step": 180, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 0.8098434004474273, |
| "grad_norm": 2.0015857219696045, |
| "learning_rate": 0.00013869677888159887, |
| "logits/chosen": -0.4414394497871399, |
| "logits/rejected": -0.5980672240257263, |
| "logps/chosen": -6.416171550750732, |
| "logps/rejected": -34.68084716796875, |
| "loss": 0.7498472929000854, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4237560033798218, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1907028704881668, |
| "rewards/margins": 1.996950626373291, |
| "rewards/rejected": -1.8062477111816406, |
| "step": 181, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.814317673378076, |
| "grad_norm": 3.0869336128234863, |
| "learning_rate": 0.00013801088547306148, |
| "logits/chosen": -0.38094401359558105, |
| "logits/rejected": -0.5588433742523193, |
| "logps/chosen": -6.33656120300293, |
| "logps/rejected": -34.21479797363281, |
| "loss": 0.7289522886276245, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.42591309547424316, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.25475332140922546, |
| "rewards/margins": 2.0227596759796143, |
| "rewards/rejected": -1.7680062055587769, |
| "step": 182, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8187919463087249, |
| "grad_norm": 3.356872081756592, |
| "learning_rate": 0.00013732289541904948, |
| "logits/chosen": -0.4178617596626282, |
| "logits/rejected": -0.6289569139480591, |
| "logps/chosen": -7.5165696144104, |
| "logps/rejected": -40.291988372802734, |
| "loss": 0.7296820282936096, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.44231122732162476, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.10255959630012512, |
| "rewards/margins": 2.4512202739715576, |
| "rewards/rejected": -2.348660707473755, |
| "step": 183, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8232662192393736, |
| "grad_norm": 3.37188720703125, |
| "learning_rate": 0.00013663284666846134, |
| "logits/chosen": -0.4483550786972046, |
| "logits/rejected": -0.6116241812705994, |
| "logps/chosen": -7.468384742736816, |
| "logps/rejected": -36.9821662902832, |
| "loss": 0.8177734017372131, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5343176126480103, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13721218705177307, |
| "rewards/margins": 2.2863237857818604, |
| "rewards/rejected": -2.14911150932312, |
| "step": 184, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8277404921700223, |
| "grad_norm": 5.6198577880859375, |
| "learning_rate": 0.00013594077728375128, |
| "logits/chosen": -0.40004104375839233, |
| "logits/rejected": -0.6040188074111938, |
| "logps/chosen": -4.515679359436035, |
| "logps/rejected": -34.73591232299805, |
| "loss": 0.7695606350898743, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.43972742557525635, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13223202526569366, |
| "rewards/margins": 2.1134281158447266, |
| "rewards/rejected": -1.9811959266662598, |
| "step": 185, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8322147651006712, |
| "grad_norm": 2.3837716579437256, |
| "learning_rate": 0.00013524672543882996, |
| "logits/chosen": -0.36424440145492554, |
| "logits/rejected": -0.5849202871322632, |
| "logps/chosen": -5.077719688415527, |
| "logps/rejected": -37.2129020690918, |
| "loss": 0.7024149894714355, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.419405460357666, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.14517350494861603, |
| "rewards/margins": 2.114969253540039, |
| "rewards/rejected": -1.9697959423065186, |
| "step": 186, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8366890380313199, |
| "grad_norm": 1.8797435760498047, |
| "learning_rate": 0.00013455072941695863, |
| "logits/chosen": -0.46017298102378845, |
| "logits/rejected": -0.6270426511764526, |
| "logps/chosen": -5.515599250793457, |
| "logps/rejected": -35.95061492919922, |
| "loss": 0.6589202284812927, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.38542789220809937, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.22691603004932404, |
| "rewards/margins": 2.2506327629089355, |
| "rewards/rejected": -2.023716688156128, |
| "step": 187, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8411633109619687, |
| "grad_norm": 3.041163206100464, |
| "learning_rate": 0.00013385282760863758, |
| "logits/chosen": -0.4110674560070038, |
| "logits/rejected": -0.5675216317176819, |
| "logps/chosen": -8.159733772277832, |
| "logps/rejected": -37.882545471191406, |
| "loss": 0.8134464025497437, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5169687867164612, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.21555905044078827, |
| "rewards/margins": 2.2340266704559326, |
| "rewards/rejected": -2.018467664718628, |
| "step": 188, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8456375838926175, |
| "grad_norm": 4.433757781982422, |
| "learning_rate": 0.00013315305850948846, |
| "logits/chosen": -0.42173799872398376, |
| "logits/rejected": -0.577892005443573, |
| "logps/chosen": -6.5279011726379395, |
| "logps/rejected": -32.7200927734375, |
| "loss": 0.7939549684524536, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45606493949890137, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06638915091753006, |
| "rewards/margins": 1.937835931777954, |
| "rewards/rejected": -1.8714466094970703, |
| "step": 189, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8501118568232662, |
| "grad_norm": 4.739625453948975, |
| "learning_rate": 0.00013245146071813114, |
| "logits/chosen": -0.325115442276001, |
| "logits/rejected": -0.5505204796791077, |
| "logps/chosen": -5.985814571380615, |
| "logps/rejected": -38.72880172729492, |
| "loss": 0.8021903038024902, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45717212557792664, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.034357890486717224, |
| "rewards/margins": 2.132763147354126, |
| "rewards/rejected": -2.098405122756958, |
| "step": 190, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.854586129753915, |
| "grad_norm": 2.447476625442505, |
| "learning_rate": 0.00013174807293405428, |
| "logits/chosen": -0.4180159270763397, |
| "logits/rejected": -0.5316296219825745, |
| "logps/chosen": -9.652694702148438, |
| "logps/rejected": -40.23739242553711, |
| "loss": 1.0005970001220703, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6700254678726196, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08515516668558121, |
| "rewards/margins": 2.6961252689361572, |
| "rewards/rejected": -2.6109700202941895, |
| "step": 191, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8590604026845637, |
| "grad_norm": 1.7246663570404053, |
| "learning_rate": 0.00013104293395548098, |
| "logits/chosen": -0.3204466998577118, |
| "logits/rejected": -0.5718462467193604, |
| "logps/chosen": -4.133203029632568, |
| "logps/rejected": -42.92612075805664, |
| "loss": 0.6551041603088379, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4175110161304474, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2626938819885254, |
| "rewards/margins": 2.859964370727539, |
| "rewards/rejected": -2.5972704887390137, |
| "step": 192, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8635346756152126, |
| "grad_norm": 2.3944220542907715, |
| "learning_rate": 0.00013033608267722858, |
| "logits/chosen": -0.2919933497905731, |
| "logits/rejected": -0.521720290184021, |
| "logps/chosen": -5.744174003601074, |
| "logps/rejected": -42.762874603271484, |
| "loss": 0.7408826351165771, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4414190948009491, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16091150045394897, |
| "rewards/margins": 2.5740907192230225, |
| "rewards/rejected": -2.413179397583008, |
| "step": 193, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8680089485458613, |
| "grad_norm": 2.7282540798187256, |
| "learning_rate": 0.00012962755808856342, |
| "logits/chosen": -0.3888423442840576, |
| "logits/rejected": -0.5426087379455566, |
| "logps/chosen": -9.648605346679688, |
| "logps/rejected": -37.68000793457031, |
| "loss": 0.9185003638267517, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.596512496471405, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.0373542495071888, |
| "rewards/margins": 2.42608904838562, |
| "rewards/rejected": -2.388734817504883, |
| "step": 194, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.87248322147651, |
| "grad_norm": 1.9054832458496094, |
| "learning_rate": 0.0001289173992710499, |
| "logits/chosen": -0.3781818151473999, |
| "logits/rejected": -0.4687708914279938, |
| "logps/chosen": -6.976923942565918, |
| "logps/rejected": -33.893470764160156, |
| "loss": 0.7810498476028442, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4621986150741577, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1968202143907547, |
| "rewards/margins": 2.0472464561462402, |
| "rewards/rejected": -1.8504258394241333, |
| "step": 195, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8769574944071589, |
| "grad_norm": 4.798694610595703, |
| "learning_rate": 0.00012820564539639512, |
| "logits/chosen": -0.2861044704914093, |
| "logits/rejected": -0.47894322872161865, |
| "logps/chosen": -6.239299774169922, |
| "logps/rejected": -34.055992126464844, |
| "loss": 0.9598355889320374, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5919869542121887, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.1414777934551239, |
| "rewards/margins": 2.0753138065338135, |
| "rewards/rejected": -1.9338361024856567, |
| "step": 196, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8814317673378076, |
| "grad_norm": 2.859259843826294, |
| "learning_rate": 0.00012749233572428804, |
| "logits/chosen": -0.3430550694465637, |
| "logits/rejected": -0.5082629323005676, |
| "logps/chosen": -5.32232666015625, |
| "logps/rejected": -38.88724136352539, |
| "loss": 0.6508762240409851, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4472533166408539, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.28352218866348267, |
| "rewards/margins": 2.8338985443115234, |
| "rewards/rejected": -2.5503764152526855, |
| "step": 197, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8859060402684564, |
| "grad_norm": 3.0246400833129883, |
| "learning_rate": 0.00012677750960023396, |
| "logits/chosen": -0.32787883281707764, |
| "logits/rejected": -0.4495905637741089, |
| "logps/chosen": -10.30122184753418, |
| "logps/rejected": -42.09465026855469, |
| "loss": 0.866395115852356, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.574297308921814, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.13904830813407898, |
| "rewards/margins": 2.8916144371032715, |
| "rewards/rejected": -2.75256609916687, |
| "step": 198, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8903803131991052, |
| "grad_norm": 4.209658622741699, |
| "learning_rate": 0.0001260612064533843, |
| "logits/chosen": -0.2963576018810272, |
| "logits/rejected": -0.5197300314903259, |
| "logps/chosen": -7.585211753845215, |
| "logps/rejected": -38.58053970336914, |
| "loss": 0.9138990640640259, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6033953428268433, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.2075960785150528, |
| "rewards/margins": 2.4362571239471436, |
| "rewards/rejected": -2.228661298751831, |
| "step": 199, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8948545861297539, |
| "grad_norm": 5.18996524810791, |
| "learning_rate": 0.0001253434657943616, |
| "logits/chosen": -0.2213418185710907, |
| "logits/rejected": -0.45573991537094116, |
| "logps/chosen": -8.23147201538086, |
| "logps/rejected": -50.46209716796875, |
| "loss": 0.974190890789032, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6812121868133545, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15396177768707275, |
| "rewards/margins": 3.1692397594451904, |
| "rewards/rejected": -3.015277862548828, |
| "step": 200, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.8993288590604027, |
| "grad_norm": 6.953693389892578, |
| "learning_rate": 0.0001246243272130804, |
| "logits/chosen": -0.3670811057090759, |
| "logits/rejected": -0.5695959329605103, |
| "logps/chosen": -4.348814964294434, |
| "logps/rejected": -44.83324432373047, |
| "loss": 0.6207172870635986, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4008825123310089, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.020788919180631638, |
| "rewards/margins": 3.0340218544006348, |
| "rewards/rejected": -3.013232469558716, |
| "step": 201, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.9038031319910514, |
| "grad_norm": 2.0709781646728516, |
| "learning_rate": 0.00012390383037656327, |
| "logits/chosen": -0.2781023383140564, |
| "logits/rejected": -0.5001654624938965, |
| "logps/chosen": -5.481650352478027, |
| "logps/rejected": -45.52299499511719, |
| "loss": 0.6335952877998352, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41790321469306946, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.14746525883674622, |
| "rewards/margins": 3.0953919887542725, |
| "rewards/rejected": -2.9479269981384277, |
| "step": 202, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.9082774049217002, |
| "grad_norm": 3.936102867126465, |
| "learning_rate": 0.00012318201502675285, |
| "logits/chosen": -0.4156622588634491, |
| "logits/rejected": -0.49767816066741943, |
| "logps/chosen": -9.21568775177002, |
| "logps/rejected": -37.2863883972168, |
| "loss": 0.8709239959716797, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.568230390548706, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.08218996226787567, |
| "rewards/margins": 2.5877761840820312, |
| "rewards/rejected": -2.5055861473083496, |
| "step": 203, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.912751677852349, |
| "grad_norm": 3.1774184703826904, |
| "learning_rate": 0.00012245892097831982, |
| "logits/chosen": -0.3529960513114929, |
| "logits/rejected": -0.5200711488723755, |
| "logps/chosen": -5.73190450668335, |
| "logps/rejected": -33.8062858581543, |
| "loss": 0.7089971303939819, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4285321831703186, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.19104257225990295, |
| "rewards/margins": 2.2398736476898193, |
| "rewards/rejected": -2.0488312244415283, |
| "step": 204, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9172259507829977, |
| "grad_norm": 5.392505168914795, |
| "learning_rate": 0.0001217345881164667, |
| "logits/chosen": -0.26341456174850464, |
| "logits/rejected": -0.4720497131347656, |
| "logps/chosen": -5.948171138763428, |
| "logps/rejected": -39.315799713134766, |
| "loss": 0.7282085418701172, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4318191409111023, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.0930575281381607, |
| "rewards/margins": 2.5829875469207764, |
| "rewards/rejected": -2.4899299144744873, |
| "step": 205, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9217002237136466, |
| "grad_norm": 1.9011708498001099, |
| "learning_rate": 0.00012100905639472779, |
| "logits/chosen": -0.30916672945022583, |
| "logits/rejected": -0.4543130397796631, |
| "logps/chosen": -6.238247394561768, |
| "logps/rejected": -37.510353088378906, |
| "loss": 0.6386366486549377, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4259726405143738, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22653764486312866, |
| "rewards/margins": 2.5835399627685547, |
| "rewards/rejected": -2.3570022583007812, |
| "step": 206, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9261744966442953, |
| "grad_norm": 2.14213228225708, |
| "learning_rate": 0.00012028236583276542, |
| "logits/chosen": -0.2883036136627197, |
| "logits/rejected": -0.41744503378868103, |
| "logps/chosen": -8.626989364624023, |
| "logps/rejected": -35.49958038330078, |
| "loss": 0.8952888250350952, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5360022783279419, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08197470754384995, |
| "rewards/margins": 1.8212658166885376, |
| "rewards/rejected": -1.7392911911010742, |
| "step": 207, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.930648769574944, |
| "grad_norm": 4.2299604415893555, |
| "learning_rate": 0.00011955455651416246, |
| "logits/chosen": -0.27569687366485596, |
| "logits/rejected": -0.46719983220100403, |
| "logps/chosen": -6.777812480926514, |
| "logps/rejected": -37.963653564453125, |
| "loss": 0.6860868334770203, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4499008357524872, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.2759566307067871, |
| "rewards/margins": 2.4637763500213623, |
| "rewards/rejected": -2.187819480895996, |
| "step": 208, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9351230425055929, |
| "grad_norm": 2.7029130458831787, |
| "learning_rate": 0.00011882566858421135, |
| "logits/chosen": -0.3265005946159363, |
| "logits/rejected": -0.4556988775730133, |
| "logps/chosen": -8.206356048583984, |
| "logps/rejected": -35.63490676879883, |
| "loss": 0.8148421049118042, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5701502561569214, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.17454853653907776, |
| "rewards/margins": 2.2981252670288086, |
| "rewards/rejected": -2.1235766410827637, |
| "step": 209, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9395973154362416, |
| "grad_norm": 3.4696102142333984, |
| "learning_rate": 0.00011809574224769981, |
| "logits/chosen": -0.30080342292785645, |
| "logits/rejected": -0.5164678692817688, |
| "logps/chosen": -8.895907402038574, |
| "logps/rejected": -46.112876892089844, |
| "loss": 0.830041766166687, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6021772623062134, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.16136810183525085, |
| "rewards/margins": 3.008272171020508, |
| "rewards/rejected": -2.8469040393829346, |
| "step": 210, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9440715883668904, |
| "grad_norm": 2.8616750240325928, |
| "learning_rate": 0.00011736481776669306, |
| "logits/chosen": -0.31334638595581055, |
| "logits/rejected": -0.45864373445510864, |
| "logps/chosen": -5.779447555541992, |
| "logps/rejected": -32.93850326538086, |
| "loss": 0.6704469919204712, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3824280798435211, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.16728241741657257, |
| "rewards/margins": 2.098851442337036, |
| "rewards/rejected": -1.9315690994262695, |
| "step": 211, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9485458612975392, |
| "grad_norm": 2.4147164821624756, |
| "learning_rate": 0.00011663293545831302, |
| "logits/chosen": -0.30772557854652405, |
| "logits/rejected": -0.5058435201644897, |
| "logps/chosen": -9.233728408813477, |
| "logps/rejected": -48.16905975341797, |
| "loss": 0.8766963481903076, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6797882318496704, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.21987132728099823, |
| "rewards/margins": 3.1963741779327393, |
| "rewards/rejected": -2.9765028953552246, |
| "step": 212, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9530201342281879, |
| "grad_norm": 2.5486626625061035, |
| "learning_rate": 0.00011590013569251457, |
| "logits/chosen": -0.29882898926734924, |
| "logits/rejected": -0.43773284554481506, |
| "logps/chosen": -6.394394874572754, |
| "logps/rejected": -34.92805862426758, |
| "loss": 0.6322169303894043, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3928750455379486, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24000707268714905, |
| "rewards/margins": 2.3915932178497314, |
| "rewards/rejected": -2.1515860557556152, |
| "step": 213, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.9574944071588367, |
| "grad_norm": 2.4576425552368164, |
| "learning_rate": 0.0001151664588898586, |
| "logits/chosen": -0.35491204261779785, |
| "logits/rejected": -0.584037721157074, |
| "logps/chosen": -5.498895645141602, |
| "logps/rejected": -50.16877365112305, |
| "loss": 0.6617028117179871, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45270809531211853, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.15284521877765656, |
| "rewards/margins": 3.232778310775757, |
| "rewards/rejected": -3.0799331665039062, |
| "step": 214, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.9619686800894854, |
| "grad_norm": 2.428880214691162, |
| "learning_rate": 0.00011443194551928266, |
| "logits/chosen": -0.4301077425479889, |
| "logits/rejected": -0.5748926401138306, |
| "logps/chosen": -5.792450904846191, |
| "logps/rejected": -41.20557403564453, |
| "loss": 0.6007997989654541, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.35522550344467163, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.08507393300533295, |
| "rewards/margins": 2.962244987487793, |
| "rewards/rejected": -2.8771708011627197, |
| "step": 215, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.9664429530201343, |
| "grad_norm": 5.455321788787842, |
| "learning_rate": 0.00011369663609586854, |
| "logits/chosen": -0.4297526478767395, |
| "logits/rejected": -0.4847024381160736, |
| "logps/chosen": -8.678275108337402, |
| "logps/rejected": -29.509740829467773, |
| "loss": 0.8770935535430908, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5399242639541626, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08710916340351105, |
| "rewards/margins": 1.9487600326538086, |
| "rewards/rejected": -1.8616507053375244, |
| "step": 216, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.970917225950783, |
| "grad_norm": 3.3811686038970947, |
| "learning_rate": 0.00011296057117860759, |
| "logits/chosen": -0.31279319524765015, |
| "logits/rejected": -0.5541050434112549, |
| "logps/chosen": -9.380926132202148, |
| "logps/rejected": -47.00410461425781, |
| "loss": 0.8845710158348083, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6761906147003174, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.0901801735162735, |
| "rewards/margins": 3.160248041152954, |
| "rewards/rejected": -3.0700676441192627, |
| "step": 217, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9753914988814317, |
| "grad_norm": 7.690444469451904, |
| "learning_rate": 0.00011222379136816345, |
| "logits/chosen": -0.3524364233016968, |
| "logits/rejected": -0.5617326498031616, |
| "logps/chosen": -10.020103454589844, |
| "logps/rejected": -47.92871856689453, |
| "loss": 0.9561252593994141, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6848851442337036, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15353024005889893, |
| "rewards/margins": 3.196129083633423, |
| "rewards/rejected": -3.0425989627838135, |
| "step": 218, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9798657718120806, |
| "grad_norm": 2.4010486602783203, |
| "learning_rate": 0.00011148633730463273, |
| "logits/chosen": -0.33735328912734985, |
| "logits/rejected": -0.6036848425865173, |
| "logps/chosen": -5.477067470550537, |
| "logps/rejected": -46.78923034667969, |
| "loss": 0.6661429405212402, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45900487899780273, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.16355790197849274, |
| "rewards/margins": 3.1154041290283203, |
| "rewards/rejected": -2.9518463611602783, |
| "step": 219, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9843400447427293, |
| "grad_norm": 3.927243232727051, |
| "learning_rate": 0.00011074824966530312, |
| "logits/chosen": -0.37031301856040955, |
| "logits/rejected": -0.5702263712882996, |
| "logps/chosen": -7.515779495239258, |
| "logps/rejected": -40.670631408691406, |
| "loss": 0.9489113688468933, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5751282572746277, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.11724560707807541, |
| "rewards/margins": 2.502910614013672, |
| "rewards/rejected": -2.385664939880371, |
| "step": 220, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9888143176733781, |
| "grad_norm": 6.59840726852417, |
| "learning_rate": 0.00011000956916240985, |
| "logits/chosen": -0.41073113679885864, |
| "logits/rejected": -0.6303737759590149, |
| "logps/chosen": -6.878203392028809, |
| "logps/rejected": -45.99734878540039, |
| "loss": 0.7122887372970581, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5226024389266968, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.11505140364170074, |
| "rewards/margins": 3.235778570175171, |
| "rewards/rejected": -3.1207275390625, |
| "step": 221, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9932885906040269, |
| "grad_norm": 2.5927021503448486, |
| "learning_rate": 0.00010927033654088983, |
| "logits/chosen": -0.3365384638309479, |
| "logits/rejected": -0.5594428777694702, |
| "logps/chosen": -8.96605396270752, |
| "logps/rejected": -43.94550323486328, |
| "loss": 1.002770185470581, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.6635525226593018, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.014833219349384308, |
| "rewards/margins": 2.641629219055176, |
| "rewards/rejected": -2.6564626693725586, |
| "step": 222, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 0.9977628635346756, |
| "grad_norm": 2.235867977142334, |
| "learning_rate": 0.00010853059257613448, |
| "logits/chosen": -0.3795309066772461, |
| "logits/rejected": -0.5900553464889526, |
| "logps/chosen": -6.937759876251221, |
| "logps/rejected": -40.11079406738281, |
| "loss": 0.5907022953033447, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33843478560447693, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.07233436405658722, |
| "rewards/margins": 2.356292724609375, |
| "rewards/rejected": -2.2839584350585938, |
| "step": 223, |
| "train_speed(iter/s)": 0.00562 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 2.235867977142334, |
| "learning_rate": 0.00010779037807174033, |
| "logits/chosen": -0.38307642936706543, |
| "logits/rejected": -0.575429379940033, |
| "logps/chosen": -10.307394981384277, |
| "logps/rejected": -36.11070251464844, |
| "loss": 0.5518274903297424, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.7534576058387756, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.2281292825937271, |
| "rewards/margins": 2.14729905128479, |
| "rewards/rejected": -1.9191696643829346, |
| "step": 224, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 1.0044742729306488, |
| "grad_norm": 2.484886407852173, |
| "learning_rate": 0.00010704973385725851, |
| "logits/chosen": -0.4444534182548523, |
| "logits/rejected": -0.5337757468223572, |
| "logps/chosen": -7.323111534118652, |
| "logps/rejected": -35.633766174316406, |
| "loss": 0.6172289848327637, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.39472728967666626, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.29545480012893677, |
| "rewards/margins": 2.4496452808380127, |
| "rewards/rejected": -2.1541905403137207, |
| "step": 225, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0089485458612975, |
| "grad_norm": 2.8453030586242676, |
| "learning_rate": 0.00010630870078594249, |
| "logits/chosen": -0.3669445812702179, |
| "logits/rejected": -0.5369352698326111, |
| "logps/chosen": -6.009266376495361, |
| "logps/rejected": -43.218292236328125, |
| "loss": 0.47680002450942993, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31007498502731323, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4155438244342804, |
| "rewards/margins": 3.0658767223358154, |
| "rewards/rejected": -2.6503331661224365, |
| "step": 226, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0134228187919463, |
| "grad_norm": 4.40702486038208, |
| "learning_rate": 0.00010556731973249485, |
| "logits/chosen": -0.3794156312942505, |
| "logits/rejected": -0.5609134435653687, |
| "logps/chosen": -6.1214189529418945, |
| "logps/rejected": -38.3588981628418, |
| "loss": 0.5431226491928101, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3845002055168152, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.422393262386322, |
| "rewards/margins": 2.9117181301116943, |
| "rewards/rejected": -2.4893248081207275, |
| "step": 227, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0178970917225951, |
| "grad_norm": 1.7455840110778809, |
| "learning_rate": 0.00010482563159081238, |
| "logits/chosen": -0.3268698453903198, |
| "logits/rejected": -0.6033288836479187, |
| "logps/chosen": -5.496387004852295, |
| "logps/rejected": -50.4907112121582, |
| "loss": 0.5993185639381409, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4379004240036011, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21979132294654846, |
| "rewards/margins": 3.0987210273742676, |
| "rewards/rejected": -2.878929615020752, |
| "step": 228, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0223713646532437, |
| "grad_norm": 3.708115339279175, |
| "learning_rate": 0.00010408367727173067, |
| "logits/chosen": -0.335446834564209, |
| "logits/rejected": -0.6248437166213989, |
| "logps/chosen": -1.8182162046432495, |
| "logps/rejected": -49.31162643432617, |
| "loss": 0.30696067214012146, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.16442172229290009, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.1987520307302475, |
| "rewards/margins": 3.270106792449951, |
| "rewards/rejected": -3.071354866027832, |
| "step": 229, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0268456375838926, |
| "grad_norm": 1.4245268106460571, |
| "learning_rate": 0.00010334149770076747, |
| "logits/chosen": -0.4321066439151764, |
| "logits/rejected": -0.5824412107467651, |
| "logps/chosen": -5.075246810913086, |
| "logps/rejected": -35.597496032714844, |
| "loss": 0.527398943901062, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3494449853897095, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3942062258720398, |
| "rewards/margins": 2.7244553565979004, |
| "rewards/rejected": -2.330249071121216, |
| "step": 230, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0313199105145414, |
| "grad_norm": 2.983712673187256, |
| "learning_rate": 0.0001025991338158651, |
| "logits/chosen": -0.3966450095176697, |
| "logits/rejected": -0.5776999592781067, |
| "logps/chosen": -3.8128936290740967, |
| "logps/rejected": -38.07972717285156, |
| "loss": 0.46247074007987976, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.28802502155303955, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2343234121799469, |
| "rewards/margins": 2.575010061264038, |
| "rewards/rejected": -2.340686559677124, |
| "step": 231, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 1.0357941834451903, |
| "grad_norm": 1.7103675603866577, |
| "learning_rate": 0.00010185662656513251, |
| "logits/chosen": -0.33908575773239136, |
| "logits/rejected": -0.543910801410675, |
| "logps/chosen": -6.781057834625244, |
| "logps/rejected": -40.72597885131836, |
| "loss": 0.5812743902206421, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.39541077613830566, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.32517990469932556, |
| "rewards/margins": 2.686272144317627, |
| "rewards/rejected": -2.3610920906066895, |
| "step": 232, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0402684563758389, |
| "grad_norm": 1.6416723728179932, |
| "learning_rate": 0.00010111401690458654, |
| "logits/chosen": -0.37221670150756836, |
| "logits/rejected": -0.5374215841293335, |
| "logps/chosen": -5.313752174377441, |
| "logps/rejected": -39.40772247314453, |
| "loss": 0.47626128792762756, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3095937967300415, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.38376766443252563, |
| "rewards/margins": 2.872239828109741, |
| "rewards/rejected": -2.4884719848632812, |
| "step": 233, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0447427293064877, |
| "grad_norm": 1.6093069314956665, |
| "learning_rate": 0.00010037134579589302, |
| "logits/chosen": -0.4185706377029419, |
| "logits/rejected": -0.6050505638122559, |
| "logps/chosen": -7.153165817260742, |
| "logps/rejected": -43.014652252197266, |
| "loss": 0.6177042126655579, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4057208001613617, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.21249428391456604, |
| "rewards/margins": 3.0852949619293213, |
| "rewards/rejected": -2.872800588607788, |
| "step": 234, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0492170022371365, |
| "grad_norm": 2.2540111541748047, |
| "learning_rate": 9.962865420410701e-05, |
| "logits/chosen": -0.3561246991157532, |
| "logits/rejected": -0.5401315093040466, |
| "logps/chosen": -6.133809566497803, |
| "logps/rejected": -40.84746170043945, |
| "loss": 0.4942439794540405, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3529200553894043, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2851286828517914, |
| "rewards/margins": 2.9554595947265625, |
| "rewards/rejected": -2.6703310012817383, |
| "step": 235, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0536912751677852, |
| "grad_norm": 2.0389819145202637, |
| "learning_rate": 9.888598309541347e-05, |
| "logits/chosen": -0.3859875798225403, |
| "logits/rejected": -0.6725212335586548, |
| "logps/chosen": -4.989171981811523, |
| "logps/rejected": -44.53968811035156, |
| "loss": 0.5320685505867004, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3883105516433716, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3034836947917938, |
| "rewards/margins": 3.2500219345092773, |
| "rewards/rejected": -2.946537971496582, |
| "step": 236, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.058165548098434, |
| "grad_norm": 11.425275802612305, |
| "learning_rate": 9.814337343486754e-05, |
| "logits/chosen": -0.41677045822143555, |
| "logits/rejected": -0.681226909160614, |
| "logps/chosen": -6.433794021606445, |
| "logps/rejected": -47.453330993652344, |
| "loss": 0.6826303601264954, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5648487210273743, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.19153328239917755, |
| "rewards/margins": 3.4548466205596924, |
| "rewards/rejected": -3.2633132934570312, |
| "step": 237, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0626398210290828, |
| "grad_norm": 2.6108782291412354, |
| "learning_rate": 9.740086618413495e-05, |
| "logits/chosen": -0.35633864998817444, |
| "logits/rejected": -0.6819493174552917, |
| "logps/chosen": -4.880098819732666, |
| "logps/rejected": -58.097110748291016, |
| "loss": 0.44784390926361084, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.37838631868362427, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.29481351375579834, |
| "rewards/margins": 4.305008411407471, |
| "rewards/rejected": -4.010194778442383, |
| "step": 238, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0671140939597314, |
| "grad_norm": 4.4547438621521, |
| "learning_rate": 9.665850229923258e-05, |
| "logits/chosen": -0.43452098965644836, |
| "logits/rejected": -0.6579560041427612, |
| "logps/chosen": -4.716073513031006, |
| "logps/rejected": -48.85848617553711, |
| "loss": 0.49553921818733215, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3607966899871826, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2058188021183014, |
| "rewards/margins": 3.5923969745635986, |
| "rewards/rejected": -3.38657808303833, |
| "step": 239, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0715883668903803, |
| "grad_norm": 3.9305038452148438, |
| "learning_rate": 9.591632272826934e-05, |
| "logits/chosen": -0.3088409900665283, |
| "logits/rejected": -0.6176849603652954, |
| "logps/chosen": -4.434784889221191, |
| "logps/rejected": -54.108062744140625, |
| "loss": 0.4914236068725586, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3651493787765503, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24546122550964355, |
| "rewards/margins": 3.803105354309082, |
| "rewards/rejected": -3.5576443672180176, |
| "step": 240, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0760626398210291, |
| "grad_norm": 2.7532975673675537, |
| "learning_rate": 9.517436840918766e-05, |
| "logits/chosen": -0.41035979986190796, |
| "logits/rejected": -0.6430486440658569, |
| "logps/chosen": -5.0145182609558105, |
| "logps/rejected": -46.19696807861328, |
| "loss": 0.5373749732971191, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.362954318523407, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.17752307653427124, |
| "rewards/margins": 3.215165138244629, |
| "rewards/rejected": -3.037642002105713, |
| "step": 241, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0805369127516777, |
| "grad_norm": 2.9417192935943604, |
| "learning_rate": 9.44326802675052e-05, |
| "logits/chosen": -0.3006266951560974, |
| "logits/rejected": -0.5898244976997375, |
| "logps/chosen": -5.02177619934082, |
| "logps/rejected": -47.1878662109375, |
| "loss": 0.44721511006355286, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.32431328296661377, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.39047926664352417, |
| "rewards/margins": 3.3298583030700684, |
| "rewards/rejected": -2.9393789768218994, |
| "step": 242, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0850111856823266, |
| "grad_norm": 2.234625816345215, |
| "learning_rate": 9.369129921405754e-05, |
| "logits/chosen": -0.3297862112522125, |
| "logits/rejected": -0.6249620318412781, |
| "logps/chosen": -4.588490009307861, |
| "logps/rejected": -43.63856506347656, |
| "loss": 0.44313275814056396, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.30556169152259827, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.2911630868911743, |
| "rewards/margins": 3.1379594802856445, |
| "rewards/rejected": -2.846796989440918, |
| "step": 243, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 1.0894854586129754, |
| "grad_norm": 2.9266371726989746, |
| "learning_rate": 9.295026614274152e-05, |
| "logits/chosen": -0.33700501918792725, |
| "logits/rejected": -0.6508822441101074, |
| "logps/chosen": -7.202463150024414, |
| "logps/rejected": -48.011070251464844, |
| "loss": 0.5501227974891663, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41064733266830444, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.466447114944458, |
| "rewards/margins": 3.4353857040405273, |
| "rewards/rejected": -2.9689383506774902, |
| "step": 244, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.0939597315436242, |
| "grad_norm": 1.520435094833374, |
| "learning_rate": 9.220962192825968e-05, |
| "logits/chosen": -0.2550147771835327, |
| "logits/rejected": -0.5666214823722839, |
| "logps/chosen": -4.967221736907959, |
| "logps/rejected": -48.09080123901367, |
| "loss": 0.42909708619117737, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33424443006515503, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.5164206027984619, |
| "rewards/margins": 3.6842832565307617, |
| "rewards/rejected": -3.167862892150879, |
| "step": 245, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.0984340044742729, |
| "grad_norm": 1.9133579730987549, |
| "learning_rate": 9.146940742386553e-05, |
| "logits/chosen": -0.3395351767539978, |
| "logits/rejected": -0.6194986701011658, |
| "logps/chosen": -3.245999574661255, |
| "logps/rejected": -48.67194747924805, |
| "loss": 0.40019088983535767, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2735483944416046, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2056996375322342, |
| "rewards/margins": 3.698310375213623, |
| "rewards/rejected": -3.492610454559326, |
| "step": 246, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.1029082774049217, |
| "grad_norm": 3.2224693298339844, |
| "learning_rate": 9.072966345911019e-05, |
| "logits/chosen": -0.29714155197143555, |
| "logits/rejected": -0.5788121223449707, |
| "logps/chosen": -5.69664192199707, |
| "logps/rejected": -56.27704620361328, |
| "loss": 0.5205512642860413, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.42394617199897766, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.22148963809013367, |
| "rewards/margins": 4.184146881103516, |
| "rewards/rejected": -3.9626572132110596, |
| "step": 247, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.1073825503355705, |
| "grad_norm": 3.369208335876465, |
| "learning_rate": 8.999043083759017e-05, |
| "logits/chosen": -0.3414077162742615, |
| "logits/rejected": -0.6023144721984863, |
| "logps/chosen": -5.8809943199157715, |
| "logps/rejected": -46.12065505981445, |
| "loss": 0.5429152250289917, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4344760477542877, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.37368032336235046, |
| "rewards/margins": 3.512446403503418, |
| "rewards/rejected": -3.138766288757324, |
| "step": 248, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.1118568232662192, |
| "grad_norm": 2.0529885292053223, |
| "learning_rate": 8.925175033469688e-05, |
| "logits/chosen": -0.24800914525985718, |
| "logits/rejected": -0.5255987644195557, |
| "logps/chosen": -6.151613235473633, |
| "logps/rejected": -52.948631286621094, |
| "loss": 0.5082452893257141, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.352613627910614, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3551517128944397, |
| "rewards/margins": 4.031224250793457, |
| "rewards/rejected": -3.676072359085083, |
| "step": 249, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.116331096196868, |
| "grad_norm": 2.5124740600585938, |
| "learning_rate": 8.851366269536729e-05, |
| "logits/chosen": -0.26961249113082886, |
| "logits/rejected": -0.5675927996635437, |
| "logps/chosen": -5.100187301635742, |
| "logps/rejected": -49.04716110229492, |
| "loss": 0.5566443800926208, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.44095391035079956, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.4250047504901886, |
| "rewards/margins": 4.00430965423584, |
| "rewards/rejected": -3.5793049335479736, |
| "step": 250, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 1.1208053691275168, |
| "grad_norm": 3.4243483543395996, |
| "learning_rate": 8.777620863183657e-05, |
| "logits/chosen": -0.34957340359687805, |
| "logits/rejected": -0.6483319997787476, |
| "logps/chosen": -6.03017520904541, |
| "logps/rejected": -50.83137130737305, |
| "loss": 0.6758670210838318, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.47578099370002747, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.1781749427318573, |
| "rewards/margins": 3.6373672485351562, |
| "rewards/rejected": -3.4591920375823975, |
| "step": 251, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 1.1252796420581657, |
| "grad_norm": 3.0111682415008545, |
| "learning_rate": 8.703942882139245e-05, |
| "logits/chosen": -0.34100106358528137, |
| "logits/rejected": -0.5722084045410156, |
| "logps/chosen": -5.588647365570068, |
| "logps/rejected": -47.537479400634766, |
| "loss": 0.47489383816719055, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33142027258872986, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4165104031562805, |
| "rewards/margins": 3.5505828857421875, |
| "rewards/rejected": -3.1340725421905518, |
| "step": 252, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 1.1297539149888143, |
| "grad_norm": 1.6837427616119385, |
| "learning_rate": 8.630336390413147e-05, |
| "logits/chosen": -0.318341463804245, |
| "logits/rejected": -0.602148175239563, |
| "logps/chosen": -4.762698650360107, |
| "logps/rejected": -57.818058013916016, |
| "loss": 0.41530388593673706, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3407127857208252, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3078892230987549, |
| "rewards/margins": 4.42728853225708, |
| "rewards/rejected": -4.119399070739746, |
| "step": 253, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 1.1342281879194631, |
| "grad_norm": 1.7377208471298218, |
| "learning_rate": 8.556805448071735e-05, |
| "logits/chosen": -0.2768436372280121, |
| "logits/rejected": -0.5286397933959961, |
| "logps/chosen": -5.270069122314453, |
| "logps/rejected": -44.92286682128906, |
| "loss": 0.5719718933105469, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40346431732177734, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.35468900203704834, |
| "rewards/margins": 3.2416746616363525, |
| "rewards/rejected": -2.8869855403900146, |
| "step": 254, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 1.138702460850112, |
| "grad_norm": 2.1543498039245605, |
| "learning_rate": 8.483354111014141e-05, |
| "logits/chosen": -0.4536677300930023, |
| "logits/rejected": -0.6746389865875244, |
| "logps/chosen": -6.907723903656006, |
| "logps/rejected": -43.56280517578125, |
| "loss": 0.5296320915222168, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41071781516075134, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4868796467781067, |
| "rewards/margins": 3.4957876205444336, |
| "rewards/rejected": -3.0089080333709717, |
| "step": 255, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 1.1431767337807606, |
| "grad_norm": 1.8656744956970215, |
| "learning_rate": 8.409986430748545e-05, |
| "logits/chosen": -0.39294540882110596, |
| "logits/rejected": -0.6587871313095093, |
| "logps/chosen": -5.210620880126953, |
| "logps/rejected": -46.816986083984375, |
| "loss": 0.5252071619033813, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3719002604484558, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3239804208278656, |
| "rewards/margins": 3.495149612426758, |
| "rewards/rejected": -3.1711692810058594, |
| "step": 256, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1476510067114094, |
| "grad_norm": 2.010622024536133, |
| "learning_rate": 8.336706454168701e-05, |
| "logits/chosen": -0.32442259788513184, |
| "logits/rejected": -0.6206884384155273, |
| "logps/chosen": -6.981322765350342, |
| "logps/rejected": -54.52667236328125, |
| "loss": 0.4600639343261719, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3734227120876312, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3917311429977417, |
| "rewards/margins": 4.217864513397217, |
| "rewards/rejected": -3.8261334896087646, |
| "step": 257, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1521252796420582, |
| "grad_norm": 1.5634533166885376, |
| "learning_rate": 8.263518223330697e-05, |
| "logits/chosen": -0.44610998034477234, |
| "logits/rejected": -0.6434642672538757, |
| "logps/chosen": -7.3278350830078125, |
| "logps/rejected": -48.34819412231445, |
| "loss": 0.5413830876350403, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4026271402835846, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.5039491057395935, |
| "rewards/margins": 3.8213727474212646, |
| "rewards/rejected": -3.3174238204956055, |
| "step": 258, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1565995525727069, |
| "grad_norm": 4.007801532745361, |
| "learning_rate": 8.190425775230021e-05, |
| "logits/chosen": -0.3476124405860901, |
| "logits/rejected": -0.6711318492889404, |
| "logps/chosen": -5.098440170288086, |
| "logps/rejected": -48.90070724487305, |
| "loss": 0.42464467883110046, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3378307819366455, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4362422227859497, |
| "rewards/margins": 3.8473596572875977, |
| "rewards/rejected": -3.4111177921295166, |
| "step": 259, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1610738255033557, |
| "grad_norm": 2.5361435413360596, |
| "learning_rate": 8.117433141578866e-05, |
| "logits/chosen": -0.4177584946155548, |
| "logits/rejected": -0.5863391160964966, |
| "logps/chosen": -6.138072967529297, |
| "logps/rejected": -46.63839340209961, |
| "loss": 0.5487118363380432, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40140300989151, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2436431646347046, |
| "rewards/margins": 3.2071709632873535, |
| "rewards/rejected": -2.9635279178619385, |
| "step": 260, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1655480984340045, |
| "grad_norm": 3.6130049228668213, |
| "learning_rate": 8.044544348583755e-05, |
| "logits/chosen": -0.38566017150878906, |
| "logits/rejected": -0.6825127005577087, |
| "logps/chosen": -3.661987781524658, |
| "logps/rejected": -47.40394592285156, |
| "loss": 0.40936416387557983, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.28705930709838867, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.34029075503349304, |
| "rewards/margins": 3.613295316696167, |
| "rewards/rejected": -3.2730045318603516, |
| "step": 261, |
| "train_speed(iter/s)": 0.005626 |
| }, |
| { |
| "epoch": 1.1700223713646531, |
| "grad_norm": 2.4069151878356934, |
| "learning_rate": 7.971763416723459e-05, |
| "logits/chosen": -0.32102227210998535, |
| "logits/rejected": -0.6520596146583557, |
| "logps/chosen": -7.717752933502197, |
| "logps/rejected": -56.516021728515625, |
| "loss": 0.5944215059280396, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.46641111373901367, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.31582900881767273, |
| "rewards/margins": 4.05399227142334, |
| "rewards/rejected": -3.738163709640503, |
| "step": 262, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.174496644295302, |
| "grad_norm": 2.279827833175659, |
| "learning_rate": 7.89909436052722e-05, |
| "logits/chosen": -0.4377071261405945, |
| "logits/rejected": -0.6339465975761414, |
| "logps/chosen": -6.656128406524658, |
| "logps/rejected": -49.15144348144531, |
| "loss": 0.6149309873580933, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45076847076416016, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3392454981803894, |
| "rewards/margins": 3.6513326168060303, |
| "rewards/rejected": -3.312087059020996, |
| "step": 263, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.1789709172259508, |
| "grad_norm": 3.1244442462921143, |
| "learning_rate": 7.826541188353329e-05, |
| "logits/chosen": -0.35084378719329834, |
| "logits/rejected": -0.6840541958808899, |
| "logps/chosen": -3.9023451805114746, |
| "logps/rejected": -53.137367248535156, |
| "loss": 0.43266385793685913, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3279496729373932, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3510107398033142, |
| "rewards/margins": 3.936466693878174, |
| "rewards/rejected": -3.585455894470215, |
| "step": 264, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.1834451901565997, |
| "grad_norm": 2.2550525665283203, |
| "learning_rate": 7.754107902168019e-05, |
| "logits/chosen": -0.40471333265304565, |
| "logits/rejected": -0.6776788234710693, |
| "logps/chosen": -6.384232044219971, |
| "logps/rejected": -54.759212493896484, |
| "loss": 0.4860294759273529, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.394418865442276, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.5021213293075562, |
| "rewards/margins": 4.259507656097412, |
| "rewards/rejected": -3.7573862075805664, |
| "step": 265, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.1879194630872483, |
| "grad_norm": 1.6058971881866455, |
| "learning_rate": 7.681798497324716e-05, |
| "logits/chosen": -0.4261001944541931, |
| "logits/rejected": -0.6693729758262634, |
| "logps/chosen": -6.607486248016357, |
| "logps/rejected": -48.35670471191406, |
| "loss": 0.6735593676567078, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5240046381950378, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.26564663648605347, |
| "rewards/margins": 3.774982452392578, |
| "rewards/rejected": -3.509335994720459, |
| "step": 266, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.192393736017897, |
| "grad_norm": 3.3279082775115967, |
| "learning_rate": 7.609616962343675e-05, |
| "logits/chosen": -0.49879682064056396, |
| "logits/rejected": -0.7210312485694885, |
| "logps/chosen": -4.761632919311523, |
| "logps/rejected": -49.54342269897461, |
| "loss": 0.45162299275398254, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3338029980659485, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.34260720014572144, |
| "rewards/margins": 3.9146251678466797, |
| "rewards/rejected": -3.5720183849334717, |
| "step": 267, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.196868008948546, |
| "grad_norm": 2.1273367404937744, |
| "learning_rate": 7.537567278691964e-05, |
| "logits/chosen": -0.3661321699619293, |
| "logits/rejected": -0.6873822808265686, |
| "logps/chosen": -5.206943988800049, |
| "logps/rejected": -56.860939025878906, |
| "loss": 0.4889792203903198, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41767942905426025, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.383396714925766, |
| "rewards/margins": 4.5527167320251465, |
| "rewards/rejected": -4.169320583343506, |
| "step": 268, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.2013422818791946, |
| "grad_norm": 4.435976505279541, |
| "learning_rate": 7.465653420563845e-05, |
| "logits/chosen": -0.42771661281585693, |
| "logits/rejected": -0.689077615737915, |
| "logps/chosen": -4.431490898132324, |
| "logps/rejected": -53.931156158447266, |
| "loss": 0.40404367446899414, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3060677647590637, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.30835211277008057, |
| "rewards/margins": 4.412530899047852, |
| "rewards/rejected": -4.104178428649902, |
| "step": 269, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.2058165548098434, |
| "grad_norm": 9.26548957824707, |
| "learning_rate": 7.393879354661577e-05, |
| "logits/chosen": -0.2901456356048584, |
| "logits/rejected": -0.6541563272476196, |
| "logps/chosen": -4.378477573394775, |
| "logps/rejected": -62.90247344970703, |
| "loss": 0.44348278641700745, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31372037529945374, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3390202820301056, |
| "rewards/margins": 4.9046478271484375, |
| "rewards/rejected": -4.565627098083496, |
| "step": 270, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 1.2102908277404922, |
| "grad_norm": 3.4660370349884033, |
| "learning_rate": 7.322249039976608e-05, |
| "logits/chosen": -0.35300642251968384, |
| "logits/rejected": -0.6308537721633911, |
| "logps/chosen": -4.3834004402160645, |
| "logps/rejected": -47.50883483886719, |
| "loss": 0.48561790585517883, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3273915648460388, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.20902855694293976, |
| "rewards/margins": 3.6998932361602783, |
| "rewards/rejected": -3.4908647537231445, |
| "step": 271, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2147651006711409, |
| "grad_norm": 3.6232869625091553, |
| "learning_rate": 7.2507664275712e-05, |
| "logits/chosen": -0.3756940960884094, |
| "logits/rejected": -0.6699195504188538, |
| "logps/chosen": -6.058266639709473, |
| "logps/rejected": -52.071563720703125, |
| "loss": 0.5045337677001953, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40500137209892273, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.5043036937713623, |
| "rewards/margins": 4.029274940490723, |
| "rewards/rejected": -3.5249714851379395, |
| "step": 272, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2192393736017897, |
| "grad_norm": 1.5248574018478394, |
| "learning_rate": 7.179435460360491e-05, |
| "logits/chosen": -0.467898964881897, |
| "logits/rejected": -0.6951407194137573, |
| "logps/chosen": -4.229708194732666, |
| "logps/rejected": -45.6329345703125, |
| "loss": 0.46140700578689575, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31180113554000854, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2676628828048706, |
| "rewards/margins": 3.5363547801971436, |
| "rewards/rejected": -3.2686920166015625, |
| "step": 273, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2237136465324385, |
| "grad_norm": 2.608729839324951, |
| "learning_rate": 7.108260072895013e-05, |
| "logits/chosen": -0.4148178696632385, |
| "logits/rejected": -0.63880455493927, |
| "logps/chosen": -4.667982578277588, |
| "logps/rejected": -43.82627868652344, |
| "loss": 0.4676215350627899, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.32163432240486145, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3947795331478119, |
| "rewards/margins": 3.6294405460357666, |
| "rewards/rejected": -3.2346608638763428, |
| "step": 274, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2281879194630871, |
| "grad_norm": 2.44980525970459, |
| "learning_rate": 7.037244191143661e-05, |
| "logits/chosen": -0.35544446110725403, |
| "logits/rejected": -0.6155003309249878, |
| "logps/chosen": -3.1806983947753906, |
| "logps/rejected": -42.958126068115234, |
| "loss": 0.42853832244873047, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.26983052492141724, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21684399247169495, |
| "rewards/margins": 3.123112678527832, |
| "rewards/rejected": -2.906268358230591, |
| "step": 275, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.232662192393736, |
| "grad_norm": 3.2948148250579834, |
| "learning_rate": 6.966391732277143e-05, |
| "logits/chosen": -0.25492265820503235, |
| "logits/rejected": -0.6005708575248718, |
| "logps/chosen": -5.198720932006836, |
| "logps/rejected": -50.45551300048828, |
| "loss": 0.4813903570175171, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3565835952758789, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.27139896154403687, |
| "rewards/margins": 3.685420036315918, |
| "rewards/rejected": -3.4140210151672363, |
| "step": 276, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2371364653243848, |
| "grad_norm": 2.477715253829956, |
| "learning_rate": 6.895706604451905e-05, |
| "logits/chosen": -0.33635056018829346, |
| "logits/rejected": -0.6376497149467468, |
| "logps/chosen": -5.271233081817627, |
| "logps/rejected": -45.175846099853516, |
| "loss": 0.5958554744720459, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.43258097767829895, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.41053637862205505, |
| "rewards/margins": 3.3894155025482178, |
| "rewards/rejected": -2.97887921333313, |
| "step": 277, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2416107382550337, |
| "grad_norm": 3.417055606842041, |
| "learning_rate": 6.825192706594575e-05, |
| "logits/chosen": -0.4261041283607483, |
| "logits/rejected": -0.7041194438934326, |
| "logps/chosen": -3.475921154022217, |
| "logps/rejected": -47.301570892333984, |
| "loss": 0.4375145137310028, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.26725247502326965, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3121051788330078, |
| "rewards/margins": 3.574404239654541, |
| "rewards/rejected": -3.262299060821533, |
| "step": 278, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.2460850111856823, |
| "grad_norm": 2.0796444416046143, |
| "learning_rate": 6.75485392818689e-05, |
| "logits/chosen": -0.29555225372314453, |
| "logits/rejected": -0.5412930250167847, |
| "logps/chosen": -6.336879730224609, |
| "logps/rejected": -48.47580337524414, |
| "loss": 0.5192568898200989, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.36876535415649414, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18509003520011902, |
| "rewards/margins": 3.489081382751465, |
| "rewards/rejected": -3.3039915561676025, |
| "step": 279, |
| "train_speed(iter/s)": 0.005624 |
| }, |
| { |
| "epoch": 1.250559284116331, |
| "grad_norm": 4.371717929840088, |
| "learning_rate": 6.684694149051156e-05, |
| "logits/chosen": -0.3555789589881897, |
| "logits/rejected": -0.6074813008308411, |
| "logps/chosen": -6.713379383087158, |
| "logps/rejected": -49.20836639404297, |
| "loss": 0.49521225690841675, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3682730793952942, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.340753436088562, |
| "rewards/margins": 3.6066648960113525, |
| "rewards/rejected": -3.265911340713501, |
| "step": 280, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 1.25503355704698, |
| "grad_norm": 1.6855332851409912, |
| "learning_rate": 6.614717239136246e-05, |
| "logits/chosen": -0.36132338643074036, |
| "logits/rejected": -0.6780315041542053, |
| "logps/chosen": -5.581424236297607, |
| "logps/rejected": -60.914573669433594, |
| "loss": 0.45032861828804016, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.35733604431152344, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3296295404434204, |
| "rewards/margins": 4.67379093170166, |
| "rewards/rejected": -4.344161510467529, |
| "step": 281, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 1.2595078299776286, |
| "grad_norm": 3.808645009994507, |
| "learning_rate": 6.54492705830414e-05, |
| "logits/chosen": -0.35267043113708496, |
| "logits/rejected": -0.6105719208717346, |
| "logps/chosen": -5.265847206115723, |
| "logps/rejected": -49.1428337097168, |
| "loss": 0.5204509496688843, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3694938123226166, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4101027548313141, |
| "rewards/margins": 3.825331926345825, |
| "rewards/rejected": -3.415228843688965, |
| "step": 282, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 1.2639821029082774, |
| "grad_norm": 6.0424370765686035, |
| "learning_rate": 6.475327456117005e-05, |
| "logits/chosen": -0.3812498152256012, |
| "logits/rejected": -0.5841005444526672, |
| "logps/chosen": -7.5697174072265625, |
| "logps/rejected": -48.44361114501953, |
| "loss": 0.6550671458244324, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.48223644495010376, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.31617283821105957, |
| "rewards/margins": 3.6579878330230713, |
| "rewards/rejected": -3.3418149948120117, |
| "step": 283, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 1.2684563758389262, |
| "grad_norm": 1.3673754930496216, |
| "learning_rate": 6.405922271624874e-05, |
| "logits/chosen": -0.15991437435150146, |
| "logits/rejected": -0.6190045475959778, |
| "logps/chosen": -5.2127227783203125, |
| "logps/rejected": -69.4470443725586, |
| "loss": 0.3656357526779175, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31946197152137756, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.46399012207984924, |
| "rewards/margins": 5.099308013916016, |
| "rewards/rejected": -4.635317802429199, |
| "step": 284, |
| "train_speed(iter/s)": 0.005623 |
| }, |
| { |
| "epoch": 1.272930648769575, |
| "grad_norm": 1.2922435998916626, |
| "learning_rate": 6.336715333153869e-05, |
| "logits/chosen": -0.3058968782424927, |
| "logits/rejected": -0.6286692023277283, |
| "logps/chosen": -4.774682521820068, |
| "logps/rejected": -51.09849548339844, |
| "loss": 0.4398406445980072, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33861246705055237, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.39336544275283813, |
| "rewards/margins": 3.7960190773010254, |
| "rewards/rejected": -3.402653455734253, |
| "step": 285, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.2774049217002237, |
| "grad_norm": 1.9561717510223389, |
| "learning_rate": 6.267710458095053e-05, |
| "logits/chosen": -0.2568581700325012, |
| "logits/rejected": -0.6096608638763428, |
| "logps/chosen": -3.851630210876465, |
| "logps/rejected": -60.34761047363281, |
| "loss": 0.3585345447063446, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2737765312194824, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.33957624435424805, |
| "rewards/margins": 4.589919090270996, |
| "rewards/rejected": -4.250343322753906, |
| "step": 286, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.2818791946308725, |
| "grad_norm": 2.467003583908081, |
| "learning_rate": 6.198911452693853e-05, |
| "logits/chosen": -0.2836165428161621, |
| "logits/rejected": -0.6040266156196594, |
| "logps/chosen": -3.4633231163024902, |
| "logps/rejected": -57.385013580322266, |
| "loss": 0.36413243412971497, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2521785497665405, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.29489317536354065, |
| "rewards/margins": 4.329642295837402, |
| "rewards/rejected": -4.0347490310668945, |
| "step": 287, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.2863534675615211, |
| "grad_norm": 2.8934900760650635, |
| "learning_rate": 6.130322111840114e-05, |
| "logits/chosen": -0.33327656984329224, |
| "logits/rejected": -0.6482805013656616, |
| "logps/chosen": -5.532173156738281, |
| "logps/rejected": -54.11648178100586, |
| "loss": 0.5029698610305786, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3735355734825134, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.26377081871032715, |
| "rewards/margins": 4.123575687408447, |
| "rewards/rejected": -3.85980486869812, |
| "step": 288, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.29082774049217, |
| "grad_norm": 4.349852085113525, |
| "learning_rate": 6.0619462188587793e-05, |
| "logits/chosen": -0.4272937774658203, |
| "logits/rejected": -0.6839829087257385, |
| "logps/chosen": -6.46608304977417, |
| "logps/rejected": -53.94236755371094, |
| "loss": 0.5734564065933228, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.45104509592056274, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.33807680010795593, |
| "rewards/margins": 4.335888385772705, |
| "rewards/rejected": -3.9978113174438477, |
| "step": 289, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.2953020134228188, |
| "grad_norm": 7.4762115478515625, |
| "learning_rate": 5.993787545301204e-05, |
| "logits/chosen": -0.3716397285461426, |
| "logits/rejected": -0.7264353632926941, |
| "logps/chosen": -6.0507588386535645, |
| "logps/rejected": -65.62486267089844, |
| "loss": 0.5227332711219788, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4141680598258972, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22040605545043945, |
| "rewards/margins": 5.065747261047363, |
| "rewards/rejected": -4.845340728759766, |
| "step": 290, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 1.2997762863534676, |
| "grad_norm": 1.7472515106201172, |
| "learning_rate": 5.9258498507371194e-05, |
| "logits/chosen": -0.33567675948143005, |
| "logits/rejected": -0.672459602355957, |
| "logps/chosen": -4.622915744781494, |
| "logps/rejected": -59.684303283691406, |
| "loss": 0.49214908480644226, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3552026152610779, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2619546055793762, |
| "rewards/margins": 4.604626655578613, |
| "rewards/rejected": -4.342671871185303, |
| "step": 291, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3042505592841163, |
| "grad_norm": 3.101167917251587, |
| "learning_rate": 5.8581368825472585e-05, |
| "logits/chosen": -0.404299259185791, |
| "logits/rejected": -0.6993152499198914, |
| "logps/chosen": -3.4533581733703613, |
| "logps/rejected": -59.574981689453125, |
| "loss": 0.36029624938964844, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2589876055717468, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.31231242418289185, |
| "rewards/margins": 4.646473407745361, |
| "rewards/rejected": -4.334160804748535, |
| "step": 292, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.308724832214765, |
| "grad_norm": 1.6495614051818848, |
| "learning_rate": 5.790652375716652e-05, |
| "logits/chosen": -0.34474000334739685, |
| "logits/rejected": -0.6941490173339844, |
| "logps/chosen": -4.03609037399292, |
| "logps/rejected": -66.86280822753906, |
| "loss": 0.4014691114425659, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33410951495170593, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22682754695415497, |
| "rewards/margins": 5.202803611755371, |
| "rewards/rejected": -4.975976467132568, |
| "step": 293, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.313199105145414, |
| "grad_norm": 3.4665303230285645, |
| "learning_rate": 5.7234000526286156e-05, |
| "logits/chosen": -0.3931652903556824, |
| "logits/rejected": -0.6202120780944824, |
| "logps/chosen": -3.0910351276397705, |
| "logps/rejected": -51.20895767211914, |
| "loss": 0.3959859311580658, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2664206922054291, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.27823901176452637, |
| "rewards/margins": 4.091604709625244, |
| "rewards/rejected": -3.8133652210235596, |
| "step": 294, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3176733780760626, |
| "grad_norm": 2.0953636169433594, |
| "learning_rate": 5.656383622859418e-05, |
| "logits/chosen": -0.38693806529045105, |
| "logits/rejected": -0.6490756273269653, |
| "logps/chosen": -4.278903484344482, |
| "logps/rejected": -51.5899772644043, |
| "loss": 0.41964665055274963, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2831048369407654, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.31565162539482117, |
| "rewards/margins": 4.059335231781006, |
| "rewards/rejected": -3.7436835765838623, |
| "step": 295, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3221476510067114, |
| "grad_norm": 2.943864345550537, |
| "learning_rate": 5.589606782973683e-05, |
| "logits/chosen": -0.3606775104999542, |
| "logits/rejected": -0.5533517599105835, |
| "logps/chosen": -3.7958004474639893, |
| "logps/rejected": -53.00364685058594, |
| "loss": 0.39812713861465454, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2592059075832367, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3119865357875824, |
| "rewards/margins": 4.129364490509033, |
| "rewards/rejected": -3.817378044128418, |
| "step": 296, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3266219239373602, |
| "grad_norm": 3.774707794189453, |
| "learning_rate": 5.5230732163204615e-05, |
| "logits/chosen": -0.34767642617225647, |
| "logits/rejected": -0.6499403715133667, |
| "logps/chosen": -3.6022303104400635, |
| "logps/rejected": -54.93785095214844, |
| "loss": 0.40875381231307983, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2905837595462799, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.36702224612236023, |
| "rewards/margins": 4.16538667678833, |
| "rewards/rejected": -3.7983648777008057, |
| "step": 297, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.331096196868009, |
| "grad_norm": 1.5885363817214966, |
| "learning_rate": 5.456786592830083e-05, |
| "logits/chosen": -0.3518737554550171, |
| "logits/rejected": -0.6097044944763184, |
| "logps/chosen": -4.608407020568848, |
| "logps/rejected": -53.15043258666992, |
| "loss": 0.40737074613571167, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31507447361946106, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.35934901237487793, |
| "rewards/margins": 4.229493618011475, |
| "rewards/rejected": -3.870144844055176, |
| "step": 298, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3355704697986577, |
| "grad_norm": 4.016380786895752, |
| "learning_rate": 5.39075056881172e-05, |
| "logits/chosen": -0.4338231086730957, |
| "logits/rejected": -0.659954309463501, |
| "logps/chosen": -5.361896514892578, |
| "logps/rejected": -52.52497863769531, |
| "loss": 0.4883747100830078, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40210235118865967, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.23951874673366547, |
| "rewards/margins": 4.1094865798950195, |
| "rewards/rejected": -3.8699681758880615, |
| "step": 299, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3400447427293065, |
| "grad_norm": 3.0947515964508057, |
| "learning_rate": 5.3249687867517095e-05, |
| "logits/chosen": -0.3721660077571869, |
| "logits/rejected": -0.6082777380943298, |
| "logps/chosen": -5.387287139892578, |
| "logps/rejected": -47.353939056396484, |
| "loss": 0.649885892868042, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4394390881061554, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.209368497133255, |
| "rewards/margins": 3.4814958572387695, |
| "rewards/rejected": -3.272127628326416, |
| "step": 300, |
| "train_speed(iter/s)": 0.005621 |
| }, |
| { |
| "epoch": 1.3400447427293065, |
| "eval_logits/chosen": -0.3397250771522522, |
| "eval_logits/rejected": -0.614118754863739, |
| "eval_logps/chosen": -7.453134536743164, |
| "eval_logps/rejected": -45.51069641113281, |
| "eval_loss": 0.7943270206451416, |
| "eval_nll_loss": 0.5665009617805481, |
| "eval_rewards/accuracies": 0.9027777910232544, |
| "eval_rewards/chosen": 0.16647693514823914, |
| "eval_rewards/margins": 3.2694599628448486, |
| "eval_rewards/rejected": -3.102982997894287, |
| "eval_runtime": 158.2199, |
| "eval_samples_per_second": 0.91, |
| "eval_steps_per_second": 0.228, |
| "step": 300 |
| }, |
| { |
| "epoch": 1.3445190156599551, |
| "grad_norm": 1.8513437509536743, |
| "learning_rate": 5.259444875112624e-05, |
| "logits/chosen": -0.35166898369789124, |
| "logits/rejected": -0.5892441868782043, |
| "logps/chosen": -6.206982612609863, |
| "logps/rejected": -53.98065185546875, |
| "loss": 0.522982120513916, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.41661083698272705, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.4109624922275543, |
| "rewards/margins": 4.339663028717041, |
| "rewards/rejected": -3.9287004470825195, |
| "step": 301, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.348993288590604, |
| "grad_norm": 2.262685537338257, |
| "learning_rate": 5.1941824481331626e-05, |
| "logits/chosen": -0.27198654413223267, |
| "logits/rejected": -0.5579496026039124, |
| "logps/chosen": -5.262180328369141, |
| "logps/rejected": -52.79492950439453, |
| "loss": 0.5066659450531006, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.40973570942878723, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.36156827211380005, |
| "rewards/margins": 4.104053497314453, |
| "rewards/rejected": -3.742485523223877, |
| "step": 302, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3534675615212528, |
| "grad_norm": 2.9432220458984375, |
| "learning_rate": 5.129185105628756e-05, |
| "logits/chosen": -0.33754950761795044, |
| "logits/rejected": -0.6656984686851501, |
| "logps/chosen": -5.836930274963379, |
| "logps/rejected": -57.983577728271484, |
| "loss": 0.5136628150939941, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.411702424287796, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2325628250837326, |
| "rewards/margins": 4.218040466308594, |
| "rewards/rejected": -3.9854774475097656, |
| "step": 303, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3579418344519016, |
| "grad_norm": 1.5537747144699097, |
| "learning_rate": 5.064456432793019e-05, |
| "logits/chosen": -0.3143869936466217, |
| "logits/rejected": -0.6705507040023804, |
| "logps/chosen": -4.21674108505249, |
| "logps/rejected": -56.43550491333008, |
| "loss": 0.37259286642074585, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2988947033882141, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.32924148440361023, |
| "rewards/margins": 4.115330696105957, |
| "rewards/rejected": -3.7860894203186035, |
| "step": 304, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3624161073825503, |
| "grad_norm": 5.257437705993652, |
| "learning_rate": 5.000000000000002e-05, |
| "logits/chosen": -0.2882038950920105, |
| "logits/rejected": -0.57389897108078, |
| "logps/chosen": -4.055515289306641, |
| "logps/rejected": -53.850440979003906, |
| "loss": 0.3756740987300873, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2755494713783264, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2226015329360962, |
| "rewards/margins": 4.168307304382324, |
| "rewards/rejected": -3.9457051753997803, |
| "step": 305, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.366890380313199, |
| "grad_norm": 2.4108705520629883, |
| "learning_rate": 4.93581936260724e-05, |
| "logits/chosen": -0.3027994632720947, |
| "logits/rejected": -0.5452630519866943, |
| "logps/chosen": -5.911576271057129, |
| "logps/rejected": -46.840003967285156, |
| "loss": 0.6861900091171265, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.46356502175331116, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.27359738945961, |
| "rewards/margins": 3.5590786933898926, |
| "rewards/rejected": -3.2854814529418945, |
| "step": 306, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.371364653243848, |
| "grad_norm": 3.5625786781311035, |
| "learning_rate": 4.8719180607596484e-05, |
| "logits/chosen": -0.3174867033958435, |
| "logits/rejected": -0.5927554368972778, |
| "logps/chosen": -5.7471771240234375, |
| "logps/rejected": -49.149932861328125, |
| "loss": 0.6849573254585266, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5026857256889343, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.1782091110944748, |
| "rewards/margins": 3.6787338256835938, |
| "rewards/rejected": -3.5005247592926025, |
| "step": 307, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3758389261744965, |
| "grad_norm": 2.4612863063812256, |
| "learning_rate": 4.808299619194251e-05, |
| "logits/chosen": -0.29022541642189026, |
| "logits/rejected": -0.5778020620346069, |
| "logps/chosen": -4.631112098693848, |
| "logps/rejected": -52.577415466308594, |
| "loss": 0.4348449110984802, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.30078357458114624, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.29566630721092224, |
| "rewards/margins": 3.845643997192383, |
| "rewards/rejected": -3.5499777793884277, |
| "step": 308, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3803131991051454, |
| "grad_norm": 1.6603471040725708, |
| "learning_rate": 4.744967547045754e-05, |
| "logits/chosen": -0.4165666103363037, |
| "logits/rejected": -0.6750180125236511, |
| "logps/chosen": -4.548413276672363, |
| "logps/rejected": -48.195499420166016, |
| "loss": 0.46350663900375366, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.36306896805763245, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3399573862552643, |
| "rewards/margins": 3.9113879203796387, |
| "rewards/rejected": -3.5714306831359863, |
| "step": 309, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3847874720357942, |
| "grad_norm": 6.359286308288574, |
| "learning_rate": 4.681925337653006e-05, |
| "logits/chosen": -0.4898450970649719, |
| "logits/rejected": -0.6976213455200195, |
| "logps/chosen": -2.3535208702087402, |
| "logps/rejected": -46.78605270385742, |
| "loss": 0.345736563205719, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.21656113862991333, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.34265750646591187, |
| "rewards/margins": 3.7292532920837402, |
| "rewards/rejected": -3.3865957260131836, |
| "step": 310, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.389261744966443, |
| "grad_norm": 1.8393584489822388, |
| "learning_rate": 4.6191764683662744e-05, |
| "logits/chosen": -0.3028990924358368, |
| "logits/rejected": -0.6557048559188843, |
| "logps/chosen": -3.447323799133301, |
| "logps/rejected": -54.03047180175781, |
| "loss": 0.34364035725593567, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.25607097148895264, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3466619849205017, |
| "rewards/margins": 4.0746564865112305, |
| "rewards/rejected": -3.727994918823242, |
| "step": 311, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3937360178970917, |
| "grad_norm": 1.5197970867156982, |
| "learning_rate": 4.5567244003554645e-05, |
| "logits/chosen": -0.3376431465148926, |
| "logits/rejected": -0.6943186521530151, |
| "logps/chosen": -5.426862716674805, |
| "logps/rejected": -52.74493408203125, |
| "loss": 0.5423266291618347, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4564194083213806, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.417688250541687, |
| "rewards/margins": 4.0157670974731445, |
| "rewards/rejected": -3.598078727722168, |
| "step": 312, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.3982102908277405, |
| "grad_norm": 1.6577972173690796, |
| "learning_rate": 4.494572578419194e-05, |
| "logits/chosen": -0.44612032175064087, |
| "logits/rejected": -0.5799017548561096, |
| "logps/chosen": -3.894256830215454, |
| "logps/rejected": -40.50112533569336, |
| "loss": 0.4608542323112488, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3280629813671112, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.35617274045944214, |
| "rewards/margins": 3.3939948081970215, |
| "rewards/rejected": -3.0378222465515137, |
| "step": 313, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4026845637583891, |
| "grad_norm": 2.6486778259277344, |
| "learning_rate": 4.432724430794786e-05, |
| "logits/chosen": -0.33609506487846375, |
| "logits/rejected": -0.6439554691314697, |
| "logps/chosen": -5.650619029998779, |
| "logps/rejected": -49.782135009765625, |
| "loss": 0.53228360414505, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3895391523838043, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.3346465826034546, |
| "rewards/margins": 3.677081346511841, |
| "rewards/rejected": -3.3424346446990967, |
| "step": 314, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.407158836689038, |
| "grad_norm": 7.691840648651123, |
| "learning_rate": 4.371183368969165e-05, |
| "logits/chosen": -0.4097428023815155, |
| "logits/rejected": -0.7250416278839111, |
| "logps/chosen": -3.7532975673675537, |
| "logps/rejected": -60.11164093017578, |
| "loss": 0.38747715950012207, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31893986463546753, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.16259630024433136, |
| "rewards/margins": 4.627762317657471, |
| "rewards/rejected": -4.465166091918945, |
| "step": 315, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4116331096196868, |
| "grad_norm": 1.6380422115325928, |
| "learning_rate": 4.309952787490689e-05, |
| "logits/chosen": -0.41353487968444824, |
| "logits/rejected": -0.656099796295166, |
| "logps/chosen": -7.336640357971191, |
| "logps/rejected": -49.86144256591797, |
| "loss": 0.6947516798973083, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5703275799751282, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.5092569589614868, |
| "rewards/margins": 4.116418361663818, |
| "rewards/rejected": -3.607161283493042, |
| "step": 316, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4161073825503356, |
| "grad_norm": 6.258267402648926, |
| "learning_rate": 4.249036063781896e-05, |
| "logits/chosen": -0.3748295307159424, |
| "logits/rejected": -0.6187299489974976, |
| "logps/chosen": -3.833024024963379, |
| "logps/rejected": -52.662418365478516, |
| "loss": 0.3898906409740448, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2913508415222168, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3392498195171356, |
| "rewards/margins": 4.161679267883301, |
| "rewards/rejected": -3.8224291801452637, |
| "step": 317, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4205816554809845, |
| "grad_norm": 1.4691777229309082, |
| "learning_rate": 4.1884365579532346e-05, |
| "logits/chosen": -0.3068940043449402, |
| "logits/rejected": -0.69428551197052, |
| "logps/chosen": -4.452873229980469, |
| "logps/rejected": -57.829078674316406, |
| "loss": 0.3776450753211975, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.28578072786331177, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3617071211338043, |
| "rewards/margins": 4.398006439208984, |
| "rewards/rejected": -4.036299228668213, |
| "step": 318, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.425055928411633, |
| "grad_norm": 2.0084006786346436, |
| "learning_rate": 4.128157612617696e-05, |
| "logits/chosen": -0.41073668003082275, |
| "logits/rejected": -0.6919273734092712, |
| "logps/chosen": -7.601768493652344, |
| "logps/rejected": -57.00600051879883, |
| "loss": 0.7179478406906128, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5513160824775696, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22752675414085388, |
| "rewards/margins": 4.380159378051758, |
| "rewards/rejected": -4.152632236480713, |
| "step": 319, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.429530201342282, |
| "grad_norm": 4.723047733306885, |
| "learning_rate": 4.0682025527064486e-05, |
| "logits/chosen": -0.14715087413787842, |
| "logits/rejected": -0.5763663053512573, |
| "logps/chosen": -5.163531303405762, |
| "logps/rejected": -71.21774291992188, |
| "loss": 0.5922179818153381, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4672166705131531, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.12609480321407318, |
| "rewards/margins": 4.96981954574585, |
| "rewards/rejected": -4.843725204467773, |
| "step": 320, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4340044742729305, |
| "grad_norm": 3.720151424407959, |
| "learning_rate": 4.008574685285442e-05, |
| "logits/chosen": -0.34143364429473877, |
| "logits/rejected": -0.6579993963241577, |
| "logps/chosen": -5.25741720199585, |
| "logps/rejected": -58.68061447143555, |
| "loss": 0.467029333114624, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.320791095495224, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.32364290952682495, |
| "rewards/margins": 4.543745040893555, |
| "rewards/rejected": -4.220102310180664, |
| "step": 321, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4384787472035794, |
| "grad_norm": 2.8483550548553467, |
| "learning_rate": 3.94927729937299e-05, |
| "logits/chosen": -0.3293909430503845, |
| "logits/rejected": -0.5969005823135376, |
| "logps/chosen": -6.698765277862549, |
| "logps/rejected": -58.93670654296875, |
| "loss": 0.5324967503547668, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3875662386417389, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.3495822548866272, |
| "rewards/margins": 4.754834175109863, |
| "rewards/rejected": -4.405251979827881, |
| "step": 322, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.4429530201342282, |
| "grad_norm": 1.6832650899887085, |
| "learning_rate": 3.890313665758348e-05, |
| "logits/chosen": -0.4338662624359131, |
| "logits/rejected": -0.7030037641525269, |
| "logps/chosen": -5.404772758483887, |
| "logps/rejected": -54.58758544921875, |
| "loss": 0.5070828795433044, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4381164610385895, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.32039517164230347, |
| "rewards/margins": 4.310638427734375, |
| "rewards/rejected": -3.9902429580688477, |
| "step": 323, |
| "train_speed(iter/s)": 0.005602 |
| }, |
| { |
| "epoch": 1.447427293064877, |
| "grad_norm": 1.7722690105438232, |
| "learning_rate": 3.8316870368213e-05, |
| "logits/chosen": -0.3775503933429718, |
| "logits/rejected": -0.5909088850021362, |
| "logps/chosen": -6.948042392730713, |
| "logps/rejected": -53.3364143371582, |
| "loss": 0.513221025466919, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.4226687550544739, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.4537815451622009, |
| "rewards/margins": 4.239094257354736, |
| "rewards/rejected": -3.785313129425049, |
| "step": 324, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4519015659955257, |
| "grad_norm": 3.676236629486084, |
| "learning_rate": 3.773400646352769e-05, |
| "logits/chosen": -0.36781182885169983, |
| "logits/rejected": -0.6920539140701294, |
| "logps/chosen": -3.714851140975952, |
| "logps/rejected": -61.90922546386719, |
| "loss": 0.40670472383499146, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3368794023990631, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.34268397092819214, |
| "rewards/margins": 4.873111724853516, |
| "rewards/rejected": -4.530427932739258, |
| "step": 325, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4563758389261745, |
| "grad_norm": 1.8188880681991577, |
| "learning_rate": 3.7154577093764334e-05, |
| "logits/chosen": -0.395844966173172, |
| "logits/rejected": -0.6281816363334656, |
| "logps/chosen": -7.8523454666137695, |
| "logps/rejected": -50.93449020385742, |
| "loss": 0.486368328332901, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.382522851228714, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4690604507923126, |
| "rewards/margins": 4.0102763175964355, |
| "rewards/rejected": -3.541215658187866, |
| "step": 326, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4608501118568233, |
| "grad_norm": 1.347862958908081, |
| "learning_rate": 3.657861421971388e-05, |
| "logits/chosen": -0.3429858386516571, |
| "logits/rejected": -0.6542443037033081, |
| "logps/chosen": -3.7786176204681396, |
| "logps/rejected": -60.83111572265625, |
| "loss": 0.41075778007507324, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31700214743614197, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2717721462249756, |
| "rewards/margins": 4.375649929046631, |
| "rewards/rejected": -4.103877544403076, |
| "step": 327, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.465324384787472, |
| "grad_norm": 2.8427813053131104, |
| "learning_rate": 3.6006149610958625e-05, |
| "logits/chosen": -0.3848859965801239, |
| "logits/rejected": -0.5358583331108093, |
| "logps/chosen": -4.8809051513671875, |
| "logps/rejected": -46.17042541503906, |
| "loss": 0.45546215772628784, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3384712338447571, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4213294982910156, |
| "rewards/margins": 3.781106472015381, |
| "rewards/rejected": -3.3597772121429443, |
| "step": 328, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4697986577181208, |
| "grad_norm": 3.1482083797454834, |
| "learning_rate": 3.543721484411976e-05, |
| "logits/chosen": -0.2849827706813812, |
| "logits/rejected": -0.6585791707038879, |
| "logps/chosen": -4.360767364501953, |
| "logps/rejected": -55.06772994995117, |
| "loss": 0.46504268050193787, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.36682000756263733, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.38763877749443054, |
| "rewards/margins": 4.205320835113525, |
| "rewards/rejected": -3.8176825046539307, |
| "step": 329, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4742729306487696, |
| "grad_norm": 3.2574269771575928, |
| "learning_rate": 3.487184130111562e-05, |
| "logits/chosen": -0.3076302111148834, |
| "logits/rejected": -0.6256691217422485, |
| "logps/chosen": -4.316102981567383, |
| "logps/rejected": -58.86445617675781, |
| "loss": 0.4626595377922058, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.335349440574646, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3723066747188568, |
| "rewards/margins": 4.587831020355225, |
| "rewards/rejected": -4.215524196624756, |
| "step": 330, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4787472035794185, |
| "grad_norm": 1.6450964212417603, |
| "learning_rate": 3.4310060167430725e-05, |
| "logits/chosen": -0.32500436902046204, |
| "logits/rejected": -0.624412477016449, |
| "logps/chosen": -4.937263488769531, |
| "logps/rejected": -59.63166809082031, |
| "loss": 0.39044544100761414, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.30855754017829895, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2391117811203003, |
| "rewards/margins": 4.519699573516846, |
| "rewards/rejected": -4.280588150024414, |
| "step": 331, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.483221476510067, |
| "grad_norm": 1.7565689086914062, |
| "learning_rate": 3.375190243039556e-05, |
| "logits/chosen": -0.3314094543457031, |
| "logits/rejected": -0.6345155239105225, |
| "logps/chosen": -4.823784828186035, |
| "logps/rejected": -51.84589767456055, |
| "loss": 0.4534035921096802, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3457861542701721, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22532111406326294, |
| "rewards/margins": 3.867464542388916, |
| "rewards/rejected": -3.642143726348877, |
| "step": 332, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.487695749440716, |
| "grad_norm": 4.223317623138428, |
| "learning_rate": 3.319739887747752e-05, |
| "logits/chosen": -0.5204321146011353, |
| "logits/rejected": -0.7036163210868835, |
| "logps/chosen": -4.016556262969971, |
| "logps/rejected": -48.14368438720703, |
| "loss": 0.4340854287147522, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31488320231437683, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.32344233989715576, |
| "rewards/margins": 3.877997875213623, |
| "rewards/rejected": -3.5545554161071777, |
| "step": 333, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4921700223713645, |
| "grad_norm": 1.824302077293396, |
| "learning_rate": 3.264658009458239e-05, |
| "logits/chosen": -0.3357248306274414, |
| "logits/rejected": -0.5842665433883667, |
| "logps/chosen": -8.244744300842285, |
| "logps/rejected": -52.15079116821289, |
| "loss": 0.6133258938789368, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.5364583730697632, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4413433074951172, |
| "rewards/margins": 4.141359806060791, |
| "rewards/rejected": -3.700016498565674, |
| "step": 334, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.4966442953020134, |
| "grad_norm": 2.2434675693511963, |
| "learning_rate": 3.209947646436752e-05, |
| "logits/chosen": -0.2935329079627991, |
| "logits/rejected": -0.5861437916755676, |
| "logps/chosen": -5.096604824066162, |
| "logps/rejected": -55.1965446472168, |
| "loss": 0.5088061094284058, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.37970277667045593, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2670452892780304, |
| "rewards/margins": 4.1344499588012695, |
| "rewards/rejected": -3.8674044609069824, |
| "step": 335, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.5011185682326622, |
| "grad_norm": 5.45697546005249, |
| "learning_rate": 3.155611816456586e-05, |
| "logits/chosen": -0.39318567514419556, |
| "logits/rejected": -0.5622078776359558, |
| "logps/chosen": -6.846230506896973, |
| "logps/rejected": -50.98385238647461, |
| "loss": 0.49888622760772705, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.38763007521629333, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.3062732219696045, |
| "rewards/margins": 4.060297012329102, |
| "rewards/rejected": -3.754023551940918, |
| "step": 336, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.505592841163311, |
| "grad_norm": 1.7538950443267822, |
| "learning_rate": 3.1016535166321356e-05, |
| "logits/chosen": -0.3007366359233856, |
| "logits/rejected": -0.5629429817199707, |
| "logps/chosen": -4.308128833770752, |
| "logps/rejected": -55.02139663696289, |
| "loss": 0.3758111298084259, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2836534082889557, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3867635428905487, |
| "rewards/margins": 4.165569305419922, |
| "rewards/rejected": -3.778805732727051, |
| "step": 337, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.5100671140939599, |
| "grad_norm": 1.6262391805648804, |
| "learning_rate": 3.0480757232535772e-05, |
| "logits/chosen": -0.3634895980358124, |
| "logits/rejected": -0.6413396596908569, |
| "logps/chosen": -5.209499359130859, |
| "logps/rejected": -53.341461181640625, |
| "loss": 0.46990200877189636, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3806995153427124, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3192530572414398, |
| "rewards/margins": 4.083569526672363, |
| "rewards/rejected": -3.7643163204193115, |
| "step": 338, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.5145413870246085, |
| "grad_norm": 5.5195088386535645, |
| "learning_rate": 2.9948813916227115e-05, |
| "logits/chosen": -0.3451216220855713, |
| "logits/rejected": -0.6123970150947571, |
| "logps/chosen": -5.359556198120117, |
| "logps/rejected": -54.252960205078125, |
| "loss": 0.46342599391937256, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3530768156051636, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24546122550964355, |
| "rewards/margins": 4.225433826446533, |
| "rewards/rejected": -3.9799728393554688, |
| "step": 339, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.5190156599552571, |
| "grad_norm": 3.048055648803711, |
| "learning_rate": 2.9420734558899322e-05, |
| "logits/chosen": -0.38990306854248047, |
| "logits/rejected": -0.6969528198242188, |
| "logps/chosen": -5.715635299682617, |
| "logps/rejected": -51.79830551147461, |
| "loss": 0.4567548930644989, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3587338328361511, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3831934928894043, |
| "rewards/margins": 4.281648635864258, |
| "rewards/rejected": -3.8984546661376953, |
| "step": 340, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.523489932885906, |
| "grad_norm": 2.501689910888672, |
| "learning_rate": 2.889654828892393e-05, |
| "logits/chosen": -0.24420952796936035, |
| "logits/rejected": -0.578873336315155, |
| "logps/chosen": -5.0676751136779785, |
| "logps/rejected": -59.5233154296875, |
| "loss": 0.37155669927597046, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.31278476119041443, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4591803550720215, |
| "rewards/margins": 4.813053607940674, |
| "rewards/rejected": -4.353872776031494, |
| "step": 341, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.5279642058165548, |
| "grad_norm": 1.5354278087615967, |
| "learning_rate": 2.8376284019933373e-05, |
| "logits/chosen": -0.4540112018585205, |
| "logits/rejected": -0.7229375243186951, |
| "logps/chosen": -4.83169412612915, |
| "logps/rejected": -54.2523193359375, |
| "loss": 0.45802274346351624, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3309043347835541, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.35657066106796265, |
| "rewards/margins": 4.4327545166015625, |
| "rewards/rejected": -4.076183795928955, |
| "step": 342, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.5324384787472036, |
| "grad_norm": 2.8110909461975098, |
| "learning_rate": 2.7859970449226104e-05, |
| "logits/chosen": -0.3509393334388733, |
| "logits/rejected": -0.6647713780403137, |
| "logps/chosen": -2.810865640640259, |
| "logps/rejected": -52.086273193359375, |
| "loss": 0.3939270079135895, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.2697364091873169, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3155217170715332, |
| "rewards/margins": 3.9285857677459717, |
| "rewards/rejected": -3.6130642890930176, |
| "step": 343, |
| "train_speed(iter/s)": 0.005601 |
| }, |
| { |
| "epoch": 1.5369127516778525, |
| "grad_norm": 1.7518740892410278, |
| "learning_rate": 2.73476360561837e-05, |
| "logits/chosen": -0.2257290929555893, |
| "logits/rejected": -0.5468841791152954, |
| "logps/chosen": -3.983034133911133, |
| "logps/rejected": -57.0892219543457, |
| "loss": 0.4137822091579437, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.30985069274902344, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4407218098640442, |
| "rewards/margins": 4.506251335144043, |
| "rewards/rejected": -4.0655293464660645, |
| "step": 344, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.541387024608501, |
| "grad_norm": 1.7124443054199219, |
| "learning_rate": 2.6839309100699973e-05, |
| "logits/chosen": -0.3101522922515869, |
| "logits/rejected": -0.6017225384712219, |
| "logps/chosen": -4.0843119621276855, |
| "logps/rejected": -56.81201934814453, |
| "loss": 0.4287090301513672, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3240818381309509, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2648446261882782, |
| "rewards/margins": 4.452305316925049, |
| "rewards/rejected": -4.187460899353027, |
| "step": 345, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.54586129753915, |
| "grad_norm": 2.6213417053222656, |
| "learning_rate": 2.6335017621622116e-05, |
| "logits/chosen": -0.40039491653442383, |
| "logits/rejected": -0.699411153793335, |
| "logps/chosen": -4.307460308074951, |
| "logps/rejected": -52.542236328125, |
| "loss": 0.5415531992912292, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.35593974590301514, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2844778001308441, |
| "rewards/margins": 3.960409641265869, |
| "rewards/rejected": -3.675931453704834, |
| "step": 346, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.5503355704697985, |
| "grad_norm": 2.4221184253692627, |
| "learning_rate": 2.5834789435204243e-05, |
| "logits/chosen": -0.28662967681884766, |
| "logits/rejected": -0.5893158912658691, |
| "logps/chosen": -5.251837730407715, |
| "logps/rejected": -55.46297836303711, |
| "loss": 0.46211305260658264, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3805118203163147, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.30533987283706665, |
| "rewards/margins": 4.2420220375061035, |
| "rewards/rejected": -3.9366822242736816, |
| "step": 347, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.5548098434004474, |
| "grad_norm": 3.0144851207733154, |
| "learning_rate": 2.5338652133572915e-05, |
| "logits/chosen": -0.3970368504524231, |
| "logits/rejected": -0.5789538621902466, |
| "logps/chosen": -5.7425737380981445, |
| "logps/rejected": -50.25794219970703, |
| "loss": 0.4803181290626526, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3716895580291748, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3145495057106018, |
| "rewards/margins": 4.102305889129639, |
| "rewards/rejected": -3.7877564430236816, |
| "step": 348, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.5592841163310962, |
| "grad_norm": 4.8794660568237305, |
| "learning_rate": 2.4846633083205263e-05, |
| "logits/chosen": -0.3288257420063019, |
| "logits/rejected": -0.6307476758956909, |
| "logps/chosen": -4.90744161605835, |
| "logps/rejected": -60.52650451660156, |
| "loss": 0.5285750031471252, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.3855237364768982, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.20134314894676208, |
| "rewards/margins": 4.785792827606201, |
| "rewards/rejected": -4.584449291229248, |
| "step": 349, |
| "train_speed(iter/s)": 0.0056 |
| }, |
| { |
| "epoch": 1.563758389261745, |
| "grad_norm": 2.8724992275238037, |
| "learning_rate": 2.4358759423419474e-05, |
| "logits/chosen": -0.30530738830566406, |
| "logits/rejected": -0.5869967937469482, |
| "logps/chosen": -5.618663311004639, |
| "logps/rejected": -61.05699157714844, |
| "loss": 0.4239829182624817, |
| "memory(GiB)": 46.57, |
| "nll_loss": 0.33604365587234497, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.5380984544754028, |
| "rewards/margins": 4.721798896789551, |
| "rewards/rejected": -4.183700084686279, |
| "step": 350, |
| "train_speed(iter/s)": 0.0056 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 446, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.0847082159026995e+17, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
