| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 34.35367990587695, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": -0.49797338247299194, | |
| "logits/rejected": -0.5135231018066406, | |
| "logps/chosen": -1.1745355129241943, | |
| "logps/rejected": -1.3596293926239014, | |
| "loss": 2.1735, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1745355129241943, | |
| "rewards/margins": 0.18509384989738464, | |
| "rewards/rejected": -1.3596293926239014, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 16.378216146989434, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": -0.521752655506134, | |
| "logits/rejected": -0.4988512396812439, | |
| "logps/chosen": -1.1591465473175049, | |
| "logps/rejected": -1.2624419927597046, | |
| "loss": 2.1407, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.1591465473175049, | |
| "rewards/margins": 0.10329560935497284, | |
| "rewards/rejected": -1.2624419927597046, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 22.732280640563598, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": -0.46235981583595276, | |
| "logits/rejected": -0.4507545530796051, | |
| "logps/chosen": -1.1068508625030518, | |
| "logps/rejected": -1.361823558807373, | |
| "loss": 2.1077, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1068508625030518, | |
| "rewards/margins": 0.2549726366996765, | |
| "rewards/rejected": -1.361823558807373, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 22.78289950681047, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -0.4406924247741699, | |
| "logits/rejected": -0.4528113007545471, | |
| "logps/chosen": -1.161055564880371, | |
| "logps/rejected": -1.2642455101013184, | |
| "loss": 2.167, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.161055564880371, | |
| "rewards/margins": 0.10318990051746368, | |
| "rewards/rejected": -1.2642455101013184, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 13.313947854898018, | |
| "learning_rate": 3.125e-07, | |
| "logits/chosen": -0.5021263360977173, | |
| "logits/rejected": -0.47814303636550903, | |
| "logps/chosen": -1.1769291162490845, | |
| "logps/rejected": -1.2403558492660522, | |
| "loss": 2.1418, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.1769291162490845, | |
| "rewards/margins": 0.06342674791812897, | |
| "rewards/rejected": -1.2403558492660522, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 21.17299953179056, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": -0.4899294972419739, | |
| "logits/rejected": -0.49411076307296753, | |
| "logps/chosen": -1.1576581001281738, | |
| "logps/rejected": -1.280582070350647, | |
| "loss": 2.1692, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.1576581001281738, | |
| "rewards/margins": 0.12292404472827911, | |
| "rewards/rejected": -1.280582070350647, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 21.15597636264796, | |
| "learning_rate": 4.3749999999999994e-07, | |
| "logits/chosen": -0.4856337904930115, | |
| "logits/rejected": -0.4433709979057312, | |
| "logps/chosen": -1.1427704095840454, | |
| "logps/rejected": -1.2787848711013794, | |
| "loss": 2.134, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1427704095840454, | |
| "rewards/margins": 0.13601449131965637, | |
| "rewards/rejected": -1.2787848711013794, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 29.038714028264685, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -0.4945921301841736, | |
| "logits/rejected": -0.4987305998802185, | |
| "logps/chosen": -1.0738334655761719, | |
| "logps/rejected": -1.39645516872406, | |
| "loss": 2.0884, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0738334655761719, | |
| "rewards/margins": 0.3226216435432434, | |
| "rewards/rejected": -1.39645516872406, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 30.530804134708777, | |
| "learning_rate": 5.625e-07, | |
| "logits/chosen": -0.45864447951316833, | |
| "logits/rejected": -0.4690025746822357, | |
| "logps/chosen": -1.090343952178955, | |
| "logps/rejected": -1.317134976387024, | |
| "loss": 2.1074, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.090343952178955, | |
| "rewards/margins": 0.22679109871387482, | |
| "rewards/rejected": -1.317134976387024, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 21.789634951246953, | |
| "learning_rate": 5.999678242522831e-07, | |
| "logits/chosen": -0.4777728021144867, | |
| "logits/rejected": -0.49264296889305115, | |
| "logps/chosen": -1.1642675399780273, | |
| "logps/rejected": -1.4595439434051514, | |
| "loss": 2.1327, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1642675399780273, | |
| "rewards/margins": 0.2952764332294464, | |
| "rewards/rejected": -1.4595439434051514, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 175.8483724549221, | |
| "learning_rate": 5.996059263493219e-07, | |
| "logits/chosen": -0.4417606294155121, | |
| "logits/rejected": -0.4309404492378235, | |
| "logps/chosen": -1.1178853511810303, | |
| "logps/rejected": -1.343202829360962, | |
| "loss": 2.106, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.1178853511810303, | |
| "rewards/margins": 0.22531744837760925, | |
| "rewards/rejected": -1.343202829360962, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 18.6141832401793, | |
| "learning_rate": 5.988423976115163e-07, | |
| "logits/chosen": -0.48267728090286255, | |
| "logits/rejected": -0.47969865798950195, | |
| "logps/chosen": -1.2083253860473633, | |
| "logps/rejected": -1.33084237575531, | |
| "loss": 2.0877, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.2083253860473633, | |
| "rewards/margins": 0.12251707166433334, | |
| "rewards/rejected": -1.33084237575531, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 32.36335168763983, | |
| "learning_rate": 5.976782615723061e-07, | |
| "logits/chosen": -0.44251489639282227, | |
| "logits/rejected": -0.40677833557128906, | |
| "logps/chosen": -1.102402925491333, | |
| "logps/rejected": -1.6289558410644531, | |
| "loss": 2.0703, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.102402925491333, | |
| "rewards/margins": 0.5265528559684753, | |
| "rewards/rejected": -1.6289558410644531, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 48.08034129839621, | |
| "learning_rate": 5.961150787913738e-07, | |
| "logits/chosen": -0.3722413182258606, | |
| "logits/rejected": -0.3657040297985077, | |
| "logps/chosen": -1.1620112657546997, | |
| "logps/rejected": -1.426941990852356, | |
| "loss": 2.0511, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.1620112657546997, | |
| "rewards/margins": 0.2649305760860443, | |
| "rewards/rejected": -1.426941990852356, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 59.84416296380324, | |
| "learning_rate": 5.941549447626671e-07, | |
| "logits/chosen": -0.37373992800712585, | |
| "logits/rejected": -0.3564635217189789, | |
| "logps/chosen": -1.1539630889892578, | |
| "logps/rejected": -1.505507230758667, | |
| "loss": 2.0661, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1539630889892578, | |
| "rewards/margins": 0.3515442907810211, | |
| "rewards/rejected": -1.505507230758667, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 25.67817680738418, | |
| "learning_rate": 5.918004871053251e-07, | |
| "logits/chosen": -0.4051768183708191, | |
| "logits/rejected": -0.38645023107528687, | |
| "logps/chosen": -1.1455223560333252, | |
| "logps/rejected": -1.5474026203155518, | |
| "loss": 2.0988, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1455223560333252, | |
| "rewards/margins": 0.40188026428222656, | |
| "rewards/rejected": -1.5474026203155518, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 34.17749794765947, | |
| "learning_rate": 5.890548620412763e-07, | |
| "logits/chosen": -0.45642200112342834, | |
| "logits/rejected": -0.42394012212753296, | |
| "logps/chosen": -1.1208285093307495, | |
| "logps/rejected": -1.4425103664398193, | |
| "loss": 2.0996, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1208285093307495, | |
| "rewards/margins": 0.3216818571090698, | |
| "rewards/rejected": -1.4425103664398193, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 10.063019778918575, | |
| "learning_rate": 5.859217501642258e-07, | |
| "logits/chosen": -0.4428345561027527, | |
| "logits/rejected": -0.4144333004951477, | |
| "logps/chosen": -1.1374626159667969, | |
| "logps/rejected": -1.4302767515182495, | |
| "loss": 2.0461, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1374626159667969, | |
| "rewards/margins": 0.2928140461444855, | |
| "rewards/rejected": -1.4302767515182495, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 20.74891634135048, | |
| "learning_rate": 5.824053515057091e-07, | |
| "logits/chosen": -0.479747474193573, | |
| "logits/rejected": -0.39375734329223633, | |
| "logps/chosen": -1.159339189529419, | |
| "logps/rejected": -1.4628812074661255, | |
| "loss": 2.0797, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.159339189529419, | |
| "rewards/margins": 0.30354195833206177, | |
| "rewards/rejected": -1.4628812074661255, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 33.44699335449945, | |
| "learning_rate": 5.785103799048218e-07, | |
| "logits/chosen": -0.4181644022464752, | |
| "logits/rejected": -0.38857489824295044, | |
| "logps/chosen": -1.1860034465789795, | |
| "logps/rejected": -1.7242858409881592, | |
| "loss": 2.0744, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.1860034465789795, | |
| "rewards/margins": 0.5382825136184692, | |
| "rewards/rejected": -1.7242858409881592, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 59.5883485681765, | |
| "learning_rate": 5.742420566891749e-07, | |
| "logits/chosen": -0.4103716015815735, | |
| "logits/rejected": -0.3968586027622223, | |
| "logps/chosen": -1.076053261756897, | |
| "logps/rejected": -1.4917399883270264, | |
| "loss": 2.0443, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.076053261756897, | |
| "rewards/margins": 0.4156867563724518, | |
| "rewards/rejected": -1.4917399883270264, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 80.31752384431189, | |
| "learning_rate": 5.696061036755478e-07, | |
| "logits/chosen": -0.4146521985530853, | |
| "logits/rejected": -0.38292473554611206, | |
| "logps/chosen": -1.0837781429290771, | |
| "logps/rejected": -1.5503333806991577, | |
| "loss": 2.0282, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0837781429290771, | |
| "rewards/margins": 0.4665554165840149, | |
| "rewards/rejected": -1.5503333806991577, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 98.56244903302768, | |
| "learning_rate": 5.64608735499618e-07, | |
| "logits/chosen": -0.28679025173187256, | |
| "logits/rejected": -0.2377845048904419, | |
| "logps/chosen": -1.111647367477417, | |
| "logps/rejected": -1.687975287437439, | |
| "loss": 2.0185, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.111647367477417, | |
| "rewards/margins": 0.576327919960022, | |
| "rewards/rejected": -1.687975287437439, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 20.27186258736798, | |
| "learning_rate": 5.592566512850545e-07, | |
| "logits/chosen": -0.27318406105041504, | |
| "logits/rejected": -0.2487379014492035, | |
| "logps/chosen": -1.2702689170837402, | |
| "logps/rejected": -1.556516408920288, | |
| "loss": 2.0975, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2702689170837402, | |
| "rewards/margins": 0.2862473428249359, | |
| "rewards/rejected": -1.556516408920288, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 30.996972753737214, | |
| "learning_rate": 5.535570256631384e-07, | |
| "logits/chosen": -0.32224705815315247, | |
| "logits/rejected": -0.2949572503566742, | |
| "logps/chosen": -1.0966984033584595, | |
| "logps/rejected": -1.425978660583496, | |
| "loss": 2.0513, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0966984033584595, | |
| "rewards/margins": 0.3292803168296814, | |
| "rewards/rejected": -1.425978660583496, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 106.95100145667757, | |
| "learning_rate": 5.475174991549528e-07, | |
| "logits/chosen": -0.2838582396507263, | |
| "logits/rejected": -0.27573472261428833, | |
| "logps/chosen": -1.065918207168579, | |
| "logps/rejected": -1.2746469974517822, | |
| "loss": 2.0692, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.065918207168579, | |
| "rewards/margins": 0.2087288349866867, | |
| "rewards/rejected": -1.2746469974517822, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 18.450255655703543, | |
| "learning_rate": 5.411461679290317e-07, | |
| "logits/chosen": -0.33663275837898254, | |
| "logits/rejected": -0.24262118339538574, | |
| "logps/chosen": -1.1360746622085571, | |
| "logps/rejected": -1.793914556503296, | |
| "loss": 2.0112, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.1360746622085571, | |
| "rewards/margins": 0.6578398942947388, | |
| "rewards/rejected": -1.793914556503296, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 20.28586481649801, | |
| "learning_rate": 5.34451572948201e-07, | |
| "logits/chosen": -0.25463438034057617, | |
| "logits/rejected": -0.18874910473823547, | |
| "logps/chosen": -1.2043225765228271, | |
| "logps/rejected": -1.7759593725204468, | |
| "loss": 1.9956, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2043225765228271, | |
| "rewards/margins": 0.5716367959976196, | |
| "rewards/rejected": -1.7759593725204468, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 16.12998962894494, | |
| "learning_rate": 5.274426885201582e-07, | |
| "logits/chosen": -0.3120715320110321, | |
| "logits/rejected": -0.28251615166664124, | |
| "logps/chosen": -1.1517010927200317, | |
| "logps/rejected": -1.5577958822250366, | |
| "loss": 2.04, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1517010927200317, | |
| "rewards/margins": 0.4060949683189392, | |
| "rewards/rejected": -1.5577958822250366, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 18.374398699221082, | |
| "learning_rate": 5.201289102671411e-07, | |
| "logits/chosen": -0.29038459062576294, | |
| "logits/rejected": -0.25689178705215454, | |
| "logps/chosen": -1.0535448789596558, | |
| "logps/rejected": -1.4643501043319702, | |
| "loss": 2.0025, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0535448789596558, | |
| "rewards/margins": 0.41080522537231445, | |
| "rewards/rejected": -1.4643501043319702, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 21.42424041369703, | |
| "learning_rate": 5.12520042530811e-07, | |
| "logits/chosen": -0.33090248703956604, | |
| "logits/rejected": -0.26949256658554077, | |
| "logps/chosen": -1.1385178565979004, | |
| "logps/rejected": -1.5117579698562622, | |
| "loss": 2.0025, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1385178565979004, | |
| "rewards/margins": 0.3732401728630066, | |
| "rewards/rejected": -1.5117579698562622, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 17.090434114730737, | |
| "learning_rate": 5.046262852292346e-07, | |
| "logits/chosen": -0.2471882402896881, | |
| "logits/rejected": -0.1910603940486908, | |
| "logps/chosen": -1.179958701133728, | |
| "logps/rejected": -1.6289422512054443, | |
| "loss": 2.0344, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.179958701133728, | |
| "rewards/margins": 0.4489835202693939, | |
| "rewards/rejected": -1.6289422512054443, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 13.121179286172973, | |
| "learning_rate": 4.964582201835856e-07, | |
| "logits/chosen": -0.2789207696914673, | |
| "logits/rejected": -0.2165801227092743, | |
| "logps/chosen": -1.1143217086791992, | |
| "logps/rejected": -1.6779086589813232, | |
| "loss": 2.012, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1143217086791992, | |
| "rewards/margins": 0.5635868906974792, | |
| "rewards/rejected": -1.6779086589813232, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 13.822111204791623, | |
| "learning_rate": 4.880267969328908e-07, | |
| "logits/chosen": -0.26305219531059265, | |
| "logits/rejected": -0.16298075020313263, | |
| "logps/chosen": -1.2098379135131836, | |
| "logps/rejected": -1.6500240564346313, | |
| "loss": 2.0205, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2098379135131836, | |
| "rewards/margins": 0.44018617272377014, | |
| "rewards/rejected": -1.6500240564346313, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 18.589343338653535, | |
| "learning_rate": 4.793433180558423e-07, | |
| "logits/chosen": -0.26549848914146423, | |
| "logits/rejected": -0.13549579679965973, | |
| "logps/chosen": -1.1815907955169678, | |
| "logps/rejected": -1.6547002792358398, | |
| "loss": 2.0137, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1815907955169678, | |
| "rewards/margins": 0.4731093943119049, | |
| "rewards/rejected": -1.6547002792358398, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 24.192793785270847, | |
| "learning_rate": 4.704194240193467e-07, | |
| "logits/chosen": -0.20712879300117493, | |
| "logits/rejected": -0.14872625470161438, | |
| "logps/chosen": -1.2047398090362549, | |
| "logps/rejected": -1.6949182748794556, | |
| "loss": 2.0473, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2047398090362549, | |
| "rewards/margins": 0.49017828702926636, | |
| "rewards/rejected": -1.6949182748794556, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 21.72801779202531, | |
| "learning_rate": 4.6126707757412686e-07, | |
| "logits/chosen": -0.19427147507667542, | |
| "logits/rejected": -0.08655323088169098, | |
| "logps/chosen": -1.2102793455123901, | |
| "logps/rejected": -1.956017255783081, | |
| "loss": 1.9488, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2102793455123901, | |
| "rewards/margins": 0.745737612247467, | |
| "rewards/rejected": -1.956017255783081, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 52.99611536180731, | |
| "learning_rate": 4.5189854771829086e-07, | |
| "logits/chosen": -0.27349403500556946, | |
| "logits/rejected": -0.19519653916358948, | |
| "logps/chosen": -1.2331929206848145, | |
| "logps/rejected": -1.693256139755249, | |
| "loss": 2.0896, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.2331929206848145, | |
| "rewards/margins": 0.4600633680820465, | |
| "rewards/rejected": -1.693256139755249, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 23.313047834378704, | |
| "learning_rate": 4.4232639325036807e-07, | |
| "logits/chosen": -0.2565682530403137, | |
| "logits/rejected": -0.20012107491493225, | |
| "logps/chosen": -1.2263553142547607, | |
| "logps/rejected": -1.6081535816192627, | |
| "loss": 2.0274, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2263553142547607, | |
| "rewards/margins": 0.38179832696914673, | |
| "rewards/rejected": -1.6081535816192627, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 28.77315959004056, | |
| "learning_rate": 4.32563445933859e-07, | |
| "logits/chosen": -0.2804745137691498, | |
| "logits/rejected": -0.2464865893125534, | |
| "logps/chosen": -1.2270160913467407, | |
| "logps/rejected": -1.6264985799789429, | |
| "loss": 2.046, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2270160913467407, | |
| "rewards/margins": 0.3994825482368469, | |
| "rewards/rejected": -1.6264985799789429, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 24.629126590195398, | |
| "learning_rate": 4.226227932958664e-07, | |
| "logits/chosen": -0.21598652005195618, | |
| "logits/rejected": -0.16539430618286133, | |
| "logps/chosen": -1.0299084186553955, | |
| "logps/rejected": -1.6446430683135986, | |
| "loss": 1.9694, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0299084186553955, | |
| "rewards/margins": 0.6147347688674927, | |
| "rewards/rejected": -1.6446430683135986, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 18.99935065324488, | |
| "learning_rate": 4.1251776108286854e-07, | |
| "logits/chosen": -0.24660630524158478, | |
| "logits/rejected": -0.19936877489089966, | |
| "logps/chosen": -1.2271109819412231, | |
| "logps/rejected": -1.5406509637832642, | |
| "loss": 2.0403, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.2271109819412231, | |
| "rewards/margins": 0.31353995203971863, | |
| "rewards/rejected": -1.5406509637832642, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 28.14089359211311, | |
| "learning_rate": 4.022618953971514e-07, | |
| "logits/chosen": -0.28811579942703247, | |
| "logits/rejected": -0.23570355772972107, | |
| "logps/chosen": -1.1424353122711182, | |
| "logps/rejected": -1.6885287761688232, | |
| "loss": 2.0049, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.1424353122711182, | |
| "rewards/margins": 0.5460935831069946, | |
| "rewards/rejected": -1.6885287761688232, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 32.89974992330637, | |
| "learning_rate": 3.918689445378477e-07, | |
| "logits/chosen": -0.30860984325408936, | |
| "logits/rejected": -0.18307650089263916, | |
| "logps/chosen": -1.2112575769424438, | |
| "logps/rejected": -1.7577863931655884, | |
| "loss": 2.0035, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.2112575769424438, | |
| "rewards/margins": 0.5465287566184998, | |
| "rewards/rejected": -1.7577863931655884, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 20.40577743783828, | |
| "learning_rate": 3.813528405709251e-07, | |
| "logits/chosen": -0.29336491227149963, | |
| "logits/rejected": -0.19095419347286224, | |
| "logps/chosen": -1.1172130107879639, | |
| "logps/rejected": -1.7578521966934204, | |
| "loss": 1.949, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1172130107879639, | |
| "rewards/margins": 0.6406393051147461, | |
| "rewards/rejected": -1.7578521966934204, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 22.351884239960583, | |
| "learning_rate": 3.707276806528282e-07, | |
| "logits/chosen": -0.3431912362575531, | |
| "logits/rejected": -0.21362292766571045, | |
| "logps/chosen": -1.175959587097168, | |
| "logps/rejected": -1.914841890335083, | |
| "loss": 1.9597, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.175959587097168, | |
| "rewards/margins": 0.7388821840286255, | |
| "rewards/rejected": -1.914841890335083, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 32.41861896385678, | |
| "learning_rate": 3.6000770813281334e-07, | |
| "logits/chosen": -0.281482458114624, | |
| "logits/rejected": -0.2136625498533249, | |
| "logps/chosen": -1.1437963247299194, | |
| "logps/rejected": -1.6638180017471313, | |
| "loss": 1.9987, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1437963247299194, | |
| "rewards/margins": 0.5200216770172119, | |
| "rewards/rejected": -1.6638180017471313, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 25.062414755869742, | |
| "learning_rate": 3.4920729345930654e-07, | |
| "logits/chosen": -0.31132057309150696, | |
| "logits/rejected": -0.2606331408023834, | |
| "logps/chosen": -1.1404446363449097, | |
| "logps/rejected": -1.7029993534088135, | |
| "loss": 2.0234, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1404446363449097, | |
| "rewards/margins": 0.5625545978546143, | |
| "rewards/rejected": -1.7029993534088135, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 23.400197287310235, | |
| "learning_rate": 3.383409149158814e-07, | |
| "logits/chosen": -0.34879469871520996, | |
| "logits/rejected": -0.27785512804985046, | |
| "logps/chosen": -1.2463206052780151, | |
| "logps/rejected": -1.6453937292099, | |
| "loss": 2.0244, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2463206052780151, | |
| "rewards/margins": 0.39907318353652954, | |
| "rewards/rejected": -1.6453937292099, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 30.06789343033768, | |
| "learning_rate": 3.2742313921268035e-07, | |
| "logits/chosen": -0.2991330623626709, | |
| "logits/rejected": -0.24600133299827576, | |
| "logps/chosen": -1.1448876857757568, | |
| "logps/rejected": -1.7555782794952393, | |
| "loss": 2.0331, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1448876857757568, | |
| "rewards/margins": 0.6106906533241272, | |
| "rewards/rejected": -1.7555782794952393, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 26.26439332360777, | |
| "learning_rate": 3.1646860195929825e-07, | |
| "logits/chosen": -0.26393812894821167, | |
| "logits/rejected": -0.15189418196678162, | |
| "logps/chosen": -1.2170337438583374, | |
| "logps/rejected": -1.7952607870101929, | |
| "loss": 1.9685, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2170337438583374, | |
| "rewards/margins": 0.5782270431518555, | |
| "rewards/rejected": -1.7952607870101929, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 30.83419650992516, | |
| "learning_rate": 3.054919880453032e-07, | |
| "logits/chosen": -0.24465498328208923, | |
| "logits/rejected": -0.1657981127500534, | |
| "logps/chosen": -1.1195638179779053, | |
| "logps/rejected": -1.8283071517944336, | |
| "loss": 2.0058, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1195638179779053, | |
| "rewards/margins": 0.7087433934211731, | |
| "rewards/rejected": -1.8283071517944336, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 26.538461186209517, | |
| "learning_rate": 2.9450801195469686e-07, | |
| "logits/chosen": -0.2653834819793701, | |
| "logits/rejected": -0.21019785106182098, | |
| "logps/chosen": -1.2122917175292969, | |
| "logps/rejected": -1.5932838916778564, | |
| "loss": 1.9747, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2122917175292969, | |
| "rewards/margins": 0.38099202513694763, | |
| "rewards/rejected": -1.5932838916778564, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 30.571225220792357, | |
| "learning_rate": 2.835313980407017e-07, | |
| "logits/chosen": -0.2485169917345047, | |
| "logits/rejected": -0.1672702431678772, | |
| "logps/chosen": -1.2699711322784424, | |
| "logps/rejected": -1.6770769357681274, | |
| "loss": 2.0095, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2699711322784424, | |
| "rewards/margins": 0.40710583329200745, | |
| "rewards/rejected": -1.6770769357681274, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 15.142492975343366, | |
| "learning_rate": 2.7257686078731973e-07, | |
| "logits/chosen": -0.2676723599433899, | |
| "logits/rejected": -0.14488348364830017, | |
| "logps/chosen": -1.1701332330703735, | |
| "logps/rejected": -1.882002830505371, | |
| "loss": 1.9973, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1701332330703735, | |
| "rewards/margins": 0.7118695974349976, | |
| "rewards/rejected": -1.882002830505371, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 17.861104389604897, | |
| "learning_rate": 2.6165908508411857e-07, | |
| "logits/chosen": -0.27734139561653137, | |
| "logits/rejected": -0.1873548924922943, | |
| "logps/chosen": -1.096847653388977, | |
| "logps/rejected": -1.556873083114624, | |
| "loss": 1.979, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.096847653388977, | |
| "rewards/margins": 0.46002545952796936, | |
| "rewards/rejected": -1.556873083114624, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 29.049210332966997, | |
| "learning_rate": 2.5079270654069354e-07, | |
| "logits/chosen": -0.22700035572052002, | |
| "logits/rejected": -0.20169806480407715, | |
| "logps/chosen": -1.1871209144592285, | |
| "logps/rejected": -1.729832410812378, | |
| "loss": 1.9858, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1871209144592285, | |
| "rewards/margins": 0.5427114367485046, | |
| "rewards/rejected": -1.729832410812378, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 19.177779527300032, | |
| "learning_rate": 2.399922918671867e-07, | |
| "logits/chosen": -0.27257853746414185, | |
| "logits/rejected": -0.20175373554229736, | |
| "logps/chosen": -1.1817193031311035, | |
| "logps/rejected": -1.81912362575531, | |
| "loss": 1.9728, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1817193031311035, | |
| "rewards/margins": 0.6374045610427856, | |
| "rewards/rejected": -1.81912362575531, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 39.27420791348495, | |
| "learning_rate": 2.2927231934717176e-07, | |
| "logits/chosen": -0.2650902271270752, | |
| "logits/rejected": -0.204110786318779, | |
| "logps/chosen": -1.1773895025253296, | |
| "logps/rejected": -1.8753960132598877, | |
| "loss": 2.0113, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.1773895025253296, | |
| "rewards/margins": 0.6980065107345581, | |
| "rewards/rejected": -1.8753960132598877, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 19.055812699519027, | |
| "learning_rate": 2.1864715942907487e-07, | |
| "logits/chosen": -0.31268611550331116, | |
| "logits/rejected": -0.26396140456199646, | |
| "logps/chosen": -1.2095041275024414, | |
| "logps/rejected": -1.6991554498672485, | |
| "loss": 1.9925, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2095041275024414, | |
| "rewards/margins": 0.4896513819694519, | |
| "rewards/rejected": -1.6991554498672485, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 24.000715730668365, | |
| "learning_rate": 2.081310554621522e-07, | |
| "logits/chosen": -0.2226269692182541, | |
| "logits/rejected": -0.17259590327739716, | |
| "logps/chosen": -1.2206140756607056, | |
| "logps/rejected": -1.8647050857543945, | |
| "loss": 1.9507, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2206140756607056, | |
| "rewards/margins": 0.6440912485122681, | |
| "rewards/rejected": -1.8647050857543945, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 24.680587241234605, | |
| "learning_rate": 1.9773810460284862e-07, | |
| "logits/chosen": -0.21720829606056213, | |
| "logits/rejected": -0.22596517205238342, | |
| "logps/chosen": -1.121246576309204, | |
| "logps/rejected": -1.6572654247283936, | |
| "loss": 1.9548, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.121246576309204, | |
| "rewards/margins": 0.536018967628479, | |
| "rewards/rejected": -1.6572654247283936, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 191.6202253290827, | |
| "learning_rate": 1.874822389171314e-07, | |
| "logits/chosen": -0.23975515365600586, | |
| "logits/rejected": -0.13572129607200623, | |
| "logps/chosen": -1.1023863554000854, | |
| "logps/rejected": -1.9153015613555908, | |
| "loss": 1.9671, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1023863554000854, | |
| "rewards/margins": 0.8129149675369263, | |
| "rewards/rejected": -1.9153015613555908, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 29.251518787297723, | |
| "learning_rate": 1.7737720670413356e-07, | |
| "logits/chosen": -0.19940456748008728, | |
| "logits/rejected": -0.15788142383098602, | |
| "logps/chosen": -1.2180979251861572, | |
| "logps/rejected": -1.7693755626678467, | |
| "loss": 1.9257, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2180979251861572, | |
| "rewards/margins": 0.5512775182723999, | |
| "rewards/rejected": -1.7693755626678467, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 38.573296633637, | |
| "learning_rate": 1.6743655406614095e-07, | |
| "logits/chosen": -0.2212747037410736, | |
| "logits/rejected": -0.13944557309150696, | |
| "logps/chosen": -1.165976881980896, | |
| "logps/rejected": -1.8154420852661133, | |
| "loss": 1.9466, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.165976881980896, | |
| "rewards/margins": 0.6494652032852173, | |
| "rewards/rejected": -1.8154420852661133, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 29.264456688369886, | |
| "learning_rate": 1.5767360674963198e-07, | |
| "logits/chosen": -0.21240024268627167, | |
| "logits/rejected": -0.13640090823173523, | |
| "logps/chosen": -1.1329911947250366, | |
| "logps/rejected": -1.5841938257217407, | |
| "loss": 1.9788, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1329911947250366, | |
| "rewards/margins": 0.45120254158973694, | |
| "rewards/rejected": -1.5841938257217407, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 21.84656926034516, | |
| "learning_rate": 1.4810145228170922e-07, | |
| "logits/chosen": -0.2895652651786804, | |
| "logits/rejected": -0.20374973118305206, | |
| "logps/chosen": -1.0948199033737183, | |
| "logps/rejected": -1.5535070896148682, | |
| "loss": 1.9921, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0948199033737183, | |
| "rewards/margins": 0.4586872160434723, | |
| "rewards/rejected": -1.5535070896148682, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 60.16510624521895, | |
| "learning_rate": 1.3873292242587306e-07, | |
| "logits/chosen": -0.2519373893737793, | |
| "logits/rejected": -0.1798809915781021, | |
| "logps/chosen": -1.3221559524536133, | |
| "logps/rejected": -1.7844518423080444, | |
| "loss": 2.0219, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3221559524536133, | |
| "rewards/margins": 0.4622960686683655, | |
| "rewards/rejected": -1.7844518423080444, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 31.10556234832626, | |
| "learning_rate": 1.295805759806533e-07, | |
| "logits/chosen": -0.28459057211875916, | |
| "logits/rejected": -0.1859065145254135, | |
| "logps/chosen": -1.1906864643096924, | |
| "logps/rejected": -1.7581002712249756, | |
| "loss": 1.9961, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.1906864643096924, | |
| "rewards/margins": 0.567413866519928, | |
| "rewards/rejected": -1.7581002712249756, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 21.145366603131734, | |
| "learning_rate": 1.2065668194415777e-07, | |
| "logits/chosen": -0.19020649790763855, | |
| "logits/rejected": -0.15534143149852753, | |
| "logps/chosen": -1.1794466972351074, | |
| "logps/rejected": -1.6421940326690674, | |
| "loss": 2.0254, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1794466972351074, | |
| "rewards/margins": 0.4627472758293152, | |
| "rewards/rejected": -1.6421940326690674, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 29.29098473580897, | |
| "learning_rate": 1.1197320306710923e-07, | |
| "logits/chosen": -0.19632667303085327, | |
| "logits/rejected": -0.12513799965381622, | |
| "logps/chosen": -1.0603824853897095, | |
| "logps/rejected": -1.7353988885879517, | |
| "loss": 1.9288, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0603824853897095, | |
| "rewards/margins": 0.6750164031982422, | |
| "rewards/rejected": -1.7353988885879517, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 24.609476201697003, | |
| "learning_rate": 1.035417798164145e-07, | |
| "logits/chosen": -0.2547205984592438, | |
| "logits/rejected": -0.17141126096248627, | |
| "logps/chosen": -1.0777822732925415, | |
| "logps/rejected": -1.6378345489501953, | |
| "loss": 1.9002, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0777822732925415, | |
| "rewards/margins": 0.5600521564483643, | |
| "rewards/rejected": -1.6378345489501953, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 44.888955546131726, | |
| "learning_rate": 9.537371477076535e-08, | |
| "logits/chosen": -0.24833258986473083, | |
| "logits/rejected": -0.14280755817890167, | |
| "logps/chosen": -1.2661911249160767, | |
| "logps/rejected": -1.8641719818115234, | |
| "loss": 2.0054, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2661911249160767, | |
| "rewards/margins": 0.5979806184768677, | |
| "rewards/rejected": -1.8641719818115234, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 29.759713813955717, | |
| "learning_rate": 8.747995746918898e-08, | |
| "logits/chosen": -0.18603017926216125, | |
| "logits/rejected": -0.11623908579349518, | |
| "logps/chosen": -1.2385270595550537, | |
| "logps/rejected": -1.8711185455322266, | |
| "loss": 1.9635, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2385270595550537, | |
| "rewards/margins": 0.6325916051864624, | |
| "rewards/rejected": -1.8711185455322266, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 18.386697317487712, | |
| "learning_rate": 7.987108973285888e-08, | |
| "logits/chosen": -0.2141023874282837, | |
| "logits/rejected": -0.24412047863006592, | |
| "logps/chosen": -1.2150145769119263, | |
| "logps/rejected": -1.7247231006622314, | |
| "loss": 1.9899, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2150145769119263, | |
| "rewards/margins": 0.5097082853317261, | |
| "rewards/rejected": -1.7247231006622314, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 24.333148504874714, | |
| "learning_rate": 7.255731147984174e-08, | |
| "logits/chosen": -0.24650990962982178, | |
| "logits/rejected": -0.18783050775527954, | |
| "logps/chosen": -1.2390451431274414, | |
| "logps/rejected": -1.6897704601287842, | |
| "loss": 1.9425, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2390451431274414, | |
| "rewards/margins": 0.45072537660598755, | |
| "rewards/rejected": -1.6897704601287842, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 22.373563868926805, | |
| "learning_rate": 6.554842705179898e-08, | |
| "logits/chosen": -0.2532094120979309, | |
| "logits/rejected": -0.2002202570438385, | |
| "logps/chosen": -1.157409429550171, | |
| "logps/rejected": -1.7103229761123657, | |
| "loss": 1.9743, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.157409429550171, | |
| "rewards/margins": 0.55291348695755, | |
| "rewards/rejected": -1.7103229761123657, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 27.22781403012628, | |
| "learning_rate": 5.885383207096832e-08, | |
| "logits/chosen": -0.25118163228034973, | |
| "logits/rejected": -0.17533348500728607, | |
| "logps/chosen": -1.158809781074524, | |
| "logps/rejected": -1.7991712093353271, | |
| "loss": 1.9567, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.158809781074524, | |
| "rewards/margins": 0.6403613090515137, | |
| "rewards/rejected": -1.7991712093353271, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 41.14703828256896, | |
| "learning_rate": 5.2482500845047165e-08, | |
| "logits/chosen": -0.2559366524219513, | |
| "logits/rejected": -0.13099372386932373, | |
| "logps/chosen": -1.1643617153167725, | |
| "logps/rejected": -1.7700135707855225, | |
| "loss": 1.9559, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.1643617153167725, | |
| "rewards/margins": 0.60565185546875, | |
| "rewards/rejected": -1.7700135707855225, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 24.3693128346703, | |
| "learning_rate": 4.644297433686162e-08, | |
| "logits/chosen": -0.19337531924247742, | |
| "logits/rejected": -0.14422497153282166, | |
| "logps/chosen": -1.0990240573883057, | |
| "logps/rejected": -1.7358309030532837, | |
| "loss": 1.9382, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0990240573883057, | |
| "rewards/margins": 0.6368069648742676, | |
| "rewards/rejected": -1.7358309030532837, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 27.686221570305076, | |
| "learning_rate": 4.074334871494558e-08, | |
| "logits/chosen": -0.27724021673202515, | |
| "logits/rejected": -0.21434080600738525, | |
| "logps/chosen": -1.2541942596435547, | |
| "logps/rejected": -1.889288306236267, | |
| "loss": 1.981, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2541942596435547, | |
| "rewards/margins": 0.6350940465927124, | |
| "rewards/rejected": -1.889288306236267, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 27.73558369495071, | |
| "learning_rate": 3.5391264500382e-08, | |
| "logits/chosen": -0.22901353240013123, | |
| "logits/rejected": -0.1688699871301651, | |
| "logps/chosen": -1.0957512855529785, | |
| "logps/rejected": -1.5964380502700806, | |
| "loss": 1.9786, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0957512855529785, | |
| "rewards/margins": 0.500686526298523, | |
| "rewards/rejected": -1.5964380502700806, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 25.24757762967279, | |
| "learning_rate": 3.0393896324452226e-08, | |
| "logits/chosen": -0.23829559981822968, | |
| "logits/rejected": -0.14140795171260834, | |
| "logps/chosen": -1.1398870944976807, | |
| "logps/rejected": -1.7605117559432983, | |
| "loss": 1.9365, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1398870944976807, | |
| "rewards/margins": 0.6206245422363281, | |
| "rewards/rejected": -1.7605117559432983, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 24.11409806888563, | |
| "learning_rate": 2.5757943310825026e-08, | |
| "logits/chosen": -0.21624989807605743, | |
| "logits/rejected": -0.15116100013256073, | |
| "logps/chosen": -1.1420520544052124, | |
| "logps/rejected": -1.7557262182235718, | |
| "loss": 1.9847, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1420520544052124, | |
| "rewards/margins": 0.6136741042137146, | |
| "rewards/rejected": -1.7557262182235718, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 31.071203059945226, | |
| "learning_rate": 2.148962009517823e-08, | |
| "logits/chosen": -0.1776510775089264, | |
| "logits/rejected": -0.11379513889551163, | |
| "logps/chosen": -1.2811336517333984, | |
| "logps/rejected": -1.7914276123046875, | |
| "loss": 1.9474, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2811336517333984, | |
| "rewards/margins": 0.5102939605712891, | |
| "rewards/rejected": -1.7914276123046875, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 22.593650984666684, | |
| "learning_rate": 1.759464849429082e-08, | |
| "logits/chosen": -0.2011154592037201, | |
| "logits/rejected": -0.14906981587409973, | |
| "logps/chosen": -1.2140544652938843, | |
| "logps/rejected": -1.8150146007537842, | |
| "loss": 1.9428, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2140544652938843, | |
| "rewards/margins": 0.6009601354598999, | |
| "rewards/rejected": -1.8150146007537842, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 20.886785791055413, | |
| "learning_rate": 1.4078249835774169e-08, | |
| "logits/chosen": -0.2394082248210907, | |
| "logits/rejected": -0.1875392496585846, | |
| "logps/chosen": -1.15377676486969, | |
| "logps/rejected": -1.962689995765686, | |
| "loss": 1.931, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.15377676486969, | |
| "rewards/margins": 0.8089130520820618, | |
| "rewards/rejected": -1.962689995765686, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 24.47138715404228, | |
| "learning_rate": 1.0945137958723705e-08, | |
| "logits/chosen": -0.0983675867319107, | |
| "logits/rejected": -0.08297935873270035, | |
| "logps/chosen": -1.2065281867980957, | |
| "logps/rejected": -1.7344862222671509, | |
| "loss": 2.0038, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2065281867980957, | |
| "rewards/margins": 0.5279580354690552, | |
| "rewards/rejected": -1.7344862222671509, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 24.937539236981156, | |
| "learning_rate": 8.19951289467482e-09, | |
| "logits/chosen": -0.2168574333190918, | |
| "logits/rejected": -0.1608891487121582, | |
| "logps/chosen": -1.1793944835662842, | |
| "logps/rejected": -1.7278966903686523, | |
| "loss": 2.0001, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1793944835662842, | |
| "rewards/margins": 0.5485021471977234, | |
| "rewards/rejected": -1.7278966903686523, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 25.949713098879222, | |
| "learning_rate": 5.84505523733293e-09, | |
| "logits/chosen": -0.13714662194252014, | |
| "logits/rejected": -0.09888915717601776, | |
| "logps/chosen": -1.2185251712799072, | |
| "logps/rejected": -1.736971139907837, | |
| "loss": 1.9405, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2185251712799072, | |
| "rewards/margins": 0.5184457302093506, | |
| "rewards/rejected": -1.736971139907837, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 48.46134541581324, | |
| "learning_rate": 3.8849212086261466e-09, | |
| "logits/chosen": -0.17583271861076355, | |
| "logits/rejected": -0.12143261730670929, | |
| "logps/chosen": -1.3378379344940186, | |
| "logps/rejected": -1.6454353332519531, | |
| "loss": 1.997, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3378379344940186, | |
| "rewards/margins": 0.3075973391532898, | |
| "rewards/rejected": -1.6454353332519531, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 17.18309794816949, | |
| "learning_rate": 2.3217384276938756e-09, | |
| "logits/chosen": -0.15463075041770935, | |
| "logits/rejected": -0.10197849571704865, | |
| "logps/chosen": -1.0989463329315186, | |
| "logps/rejected": -1.7458966970443726, | |
| "loss": 1.932, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0989463329315186, | |
| "rewards/margins": 0.6469505429267883, | |
| "rewards/rejected": -1.7458966970443726, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 23.454717675856944, | |
| "learning_rate": 1.1576023884836472e-09, | |
| "logits/chosen": -0.25512656569480896, | |
| "logits/rejected": -0.15672564506530762, | |
| "logps/chosen": -1.2135329246520996, | |
| "logps/rejected": -1.7796437740325928, | |
| "loss": 1.9646, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.2135329246520996, | |
| "rewards/margins": 0.5661108493804932, | |
| "rewards/rejected": -1.7796437740325928, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 38.4586695917519, | |
| "learning_rate": 3.940736506780395e-10, | |
| "logits/chosen": -0.2280760258436203, | |
| "logits/rejected": -0.14886632561683655, | |
| "logps/chosen": -1.1674778461456299, | |
| "logps/rejected": -1.6563600301742554, | |
| "loss": 2.0158, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1674778461456299, | |
| "rewards/margins": 0.4888822138309479, | |
| "rewards/rejected": -1.6563600301742554, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 20.371210573575656, | |
| "learning_rate": 3.2175747716822744e-11, | |
| "logits/chosen": -0.268063485622406, | |
| "logits/rejected": -0.13501767814159393, | |
| "logps/chosen": -1.2155284881591797, | |
| "logps/rejected": -1.7273550033569336, | |
| "loss": 1.9722, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.2155284881591797, | |
| "rewards/margins": 0.5118265151977539, | |
| "rewards/rejected": -1.7273550033569336, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 2.0149041866606385, | |
| "train_runtime": 17499.3556, | |
| "train_samples_per_second": 3.494, | |
| "train_steps_per_second": 0.027 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |