| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999657651489216, | |
| "eval_steps": 100, | |
| "global_step": 1460, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 3.424657534246575e-09, | |
| "logits/chosen": -1.796067237854004, | |
| "logits/rejected": -1.6250377893447876, | |
| "logps/chosen": -84.08734130859375, | |
| "logps/rejected": -66.90229797363281, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3.424657534246575e-08, | |
| "logits/chosen": -1.7872660160064697, | |
| "logits/rejected": -1.5217690467834473, | |
| "logps/chosen": -91.57577514648438, | |
| "logps/rejected": -78.510498046875, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4166666567325592, | |
| "rewards/chosen": 0.00252585974521935, | |
| "rewards/margins": 0.003410403151065111, | |
| "rewards/rejected": -0.0008845434640534222, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 6.84931506849315e-08, | |
| "logits/chosen": -1.9197826385498047, | |
| "logits/rejected": -1.6265367269515991, | |
| "logps/chosen": -96.18563079833984, | |
| "logps/rejected": -73.0329818725586, | |
| "loss": 0.6941, | |
| "rewards/accuracies": 0.4333333373069763, | |
| "rewards/chosen": -0.0032726190984249115, | |
| "rewards/margins": -0.004598576575517654, | |
| "rewards/rejected": 0.0013259568950161338, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0273972602739725e-07, | |
| "logits/chosen": -1.8969062566757202, | |
| "logits/rejected": -1.5750768184661865, | |
| "logps/chosen": -96.1051254272461, | |
| "logps/rejected": -74.69762420654297, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5416666865348816, | |
| "rewards/chosen": -0.0011650085216388106, | |
| "rewards/margins": 0.001948213903233409, | |
| "rewards/rejected": -0.0031132223084568977, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.36986301369863e-07, | |
| "logits/chosen": -1.9271103143692017, | |
| "logits/rejected": -1.6277456283569336, | |
| "logps/chosen": -96.0181884765625, | |
| "logps/rejected": -79.47406005859375, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5166666507720947, | |
| "rewards/chosen": -0.001698245876468718, | |
| "rewards/margins": -0.0010797118302434683, | |
| "rewards/rejected": -0.0006185341626405716, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.7123287671232875e-07, | |
| "logits/chosen": -1.9380648136138916, | |
| "logits/rejected": -1.6861069202423096, | |
| "logps/chosen": -93.2973861694336, | |
| "logps/rejected": -76.30517578125, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.34166663885116577, | |
| "rewards/chosen": -0.0030570379458367825, | |
| "rewards/margins": -0.005480821710079908, | |
| "rewards/rejected": 0.002423783764243126, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.054794520547945e-07, | |
| "logits/chosen": -1.832024335861206, | |
| "logits/rejected": -1.511380910873413, | |
| "logps/chosen": -98.84480285644531, | |
| "logps/rejected": -75.86248779296875, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.4500000476837158, | |
| "rewards/chosen": -0.003249814035370946, | |
| "rewards/margins": -0.00364103470928967, | |
| "rewards/rejected": 0.00039122122689150274, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.3972602739726023e-07, | |
| "logits/chosen": -1.9178974628448486, | |
| "logits/rejected": -1.6177875995635986, | |
| "logps/chosen": -92.13301086425781, | |
| "logps/rejected": -75.75408172607422, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5333333611488342, | |
| "rewards/chosen": 0.0013952379813417792, | |
| "rewards/margins": 0.0021971219684928656, | |
| "rewards/rejected": -0.0008018844528123736, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.73972602739726e-07, | |
| "logits/chosen": -1.938677191734314, | |
| "logits/rejected": -1.6598854064941406, | |
| "logps/chosen": -88.68348693847656, | |
| "logps/rejected": -74.39127349853516, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5333333015441895, | |
| "rewards/chosen": 0.0035478367935866117, | |
| "rewards/margins": 0.003417719155550003, | |
| "rewards/rejected": 0.00013011766714043915, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.0821917808219176e-07, | |
| "logits/chosen": -1.832851767539978, | |
| "logits/rejected": -1.5398364067077637, | |
| "logps/chosen": -92.82958984375, | |
| "logps/rejected": -73.15336608886719, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5833333730697632, | |
| "rewards/chosen": 0.00075482705142349, | |
| "rewards/margins": 0.0024536100681871176, | |
| "rewards/rejected": -0.001698783366009593, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.424657534246575e-07, | |
| "logits/chosen": -1.8959871530532837, | |
| "logits/rejected": -1.6301053762435913, | |
| "logps/chosen": -92.27674865722656, | |
| "logps/rejected": -74.60256958007812, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": 0.0006887881900183856, | |
| "rewards/margins": 0.0005900462856516242, | |
| "rewards/rejected": 9.874170791590586e-05, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "eval_logits/chosen": -2.0524938106536865, | |
| "eval_logits/rejected": -1.7860654592514038, | |
| "eval_logps/chosen": -91.99323272705078, | |
| "eval_logps/rejected": -72.10526275634766, | |
| "eval_loss": 0.693277895450592, | |
| "eval_rewards/accuracies": 0.4888888895511627, | |
| "eval_rewards/chosen": -0.0007517762714996934, | |
| "eval_rewards/margins": -0.00027055441751144826, | |
| "eval_rewards/rejected": -0.00048122191219590604, | |
| "eval_runtime": 117.2952, | |
| "eval_samples_per_second": 24.4, | |
| "eval_steps_per_second": 0.767, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.767123287671233e-07, | |
| "logits/chosen": -1.947257399559021, | |
| "logits/rejected": -1.6791489124298096, | |
| "logps/chosen": -93.37996673583984, | |
| "logps/rejected": -72.86904907226562, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.5083333849906921, | |
| "rewards/chosen": -0.002063233172520995, | |
| "rewards/margins": 0.0007919662748463452, | |
| "rewards/rejected": -0.0028551991563290358, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.10958904109589e-07, | |
| "logits/chosen": -1.8895361423492432, | |
| "logits/rejected": -1.587282419204712, | |
| "logps/chosen": -93.67366027832031, | |
| "logps/rejected": -72.37762451171875, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.491666704416275, | |
| "rewards/chosen": 0.0028340499848127365, | |
| "rewards/margins": 0.0019404724007472396, | |
| "rewards/rejected": 0.000893576827365905, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.4520547945205477e-07, | |
| "logits/chosen": -1.9517349004745483, | |
| "logits/rejected": -1.663013219833374, | |
| "logps/chosen": -84.09037780761719, | |
| "logps/rejected": -72.67787170410156, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.46666663885116577, | |
| "rewards/chosen": 0.0004304441681597382, | |
| "rewards/margins": -0.0008509824983775616, | |
| "rewards/rejected": 0.001281426870264113, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.794520547945205e-07, | |
| "logits/chosen": -1.947576880455017, | |
| "logits/rejected": -1.722400426864624, | |
| "logps/chosen": -89.95055389404297, | |
| "logps/rejected": -76.27583312988281, | |
| "loss": 0.6919, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.004270606208592653, | |
| "rewards/margins": 0.009159665554761887, | |
| "rewards/rejected": -0.004889058880507946, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.984779299847793e-07, | |
| "logits/chosen": -1.9852988719940186, | |
| "logits/rejected": -1.6872488260269165, | |
| "logps/chosen": -90.7852783203125, | |
| "logps/rejected": -73.9708251953125, | |
| "loss": 0.6915, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.0020091754850000143, | |
| "rewards/margins": 0.0034632813185453415, | |
| "rewards/rejected": -0.0014541053678840399, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.946727549467275e-07, | |
| "logits/chosen": -1.9156850576400757, | |
| "logits/rejected": -1.5840113162994385, | |
| "logps/chosen": -97.8337173461914, | |
| "logps/rejected": -73.20053100585938, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.0025704619474709034, | |
| "rewards/margins": 0.002448607701808214, | |
| "rewards/rejected": 0.00012185415107524022, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.908675799086758e-07, | |
| "logits/chosen": -1.803034782409668, | |
| "logits/rejected": -1.4945720434188843, | |
| "logps/chosen": -96.1871337890625, | |
| "logps/rejected": -72.80181121826172, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.5333333611488342, | |
| "rewards/chosen": 0.003732017008587718, | |
| "rewards/margins": 0.003292496781796217, | |
| "rewards/rejected": 0.0004395198484417051, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.87062404870624e-07, | |
| "logits/chosen": -1.8866764307022095, | |
| "logits/rejected": -1.5327541828155518, | |
| "logps/chosen": -97.74296569824219, | |
| "logps/rejected": -76.4982681274414, | |
| "loss": 0.6909, | |
| "rewards/accuracies": 0.5916666984558105, | |
| "rewards/chosen": 0.004220059607177973, | |
| "rewards/margins": 0.006730073597282171, | |
| "rewards/rejected": -0.002510013757273555, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.832572298325722e-07, | |
| "logits/chosen": -1.945041298866272, | |
| "logits/rejected": -1.6189870834350586, | |
| "logps/chosen": -94.37440490722656, | |
| "logps/rejected": -73.1822738647461, | |
| "loss": 0.6904, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.004208459984511137, | |
| "rewards/margins": 0.005117190536111593, | |
| "rewards/rejected": -0.0009087308426387608, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.794520547945205e-07, | |
| "logits/chosen": -1.9501270055770874, | |
| "logits/rejected": -1.6498991250991821, | |
| "logps/chosen": -93.06495666503906, | |
| "logps/rejected": -72.0920181274414, | |
| "loss": 0.69, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.0078277587890625, | |
| "rewards/margins": 0.010002164170145988, | |
| "rewards/rejected": -0.0021744065452367067, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "eval_logits/chosen": -2.052419900894165, | |
| "eval_logits/rejected": -1.7859160900115967, | |
| "eval_logps/chosen": -91.95441436767578, | |
| "eval_logps/rejected": -72.11531066894531, | |
| "eval_loss": 0.6901015043258667, | |
| "eval_rewards/accuracies": 0.5611110925674438, | |
| "eval_rewards/chosen": 0.0031298992689698935, | |
| "eval_rewards/margins": 0.004615597892552614, | |
| "eval_rewards/rejected": -0.0014856986235827208, | |
| "eval_runtime": 117.9228, | |
| "eval_samples_per_second": 24.27, | |
| "eval_steps_per_second": 0.763, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.756468797564688e-07, | |
| "logits/chosen": -1.9345300197601318, | |
| "logits/rejected": -1.6839030981063843, | |
| "logps/chosen": -92.93646240234375, | |
| "logps/rejected": -75.16416931152344, | |
| "loss": 0.6903, | |
| "rewards/accuracies": 0.6166666746139526, | |
| "rewards/chosen": 0.0068414295092225075, | |
| "rewards/margins": 0.005408720578998327, | |
| "rewards/rejected": 0.001432707766070962, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.71841704718417e-07, | |
| "logits/chosen": -1.9018728733062744, | |
| "logits/rejected": -1.599656343460083, | |
| "logps/chosen": -94.06185913085938, | |
| "logps/rejected": -72.92243957519531, | |
| "loss": 0.6888, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.00332554685883224, | |
| "rewards/margins": 0.007436770014464855, | |
| "rewards/rejected": -0.004111223388463259, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.680365296803653e-07, | |
| "logits/chosen": -1.879314661026001, | |
| "logits/rejected": -1.5864207744598389, | |
| "logps/chosen": -93.1510238647461, | |
| "logps/rejected": -74.74366760253906, | |
| "loss": 0.6898, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.007647272199392319, | |
| "rewards/margins": 0.00648108497262001, | |
| "rewards/rejected": 0.0011661878088489175, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.642313546423135e-07, | |
| "logits/chosen": -1.975441336631775, | |
| "logits/rejected": -1.7306878566741943, | |
| "logps/chosen": -86.38069915771484, | |
| "logps/rejected": -75.68878173828125, | |
| "loss": 0.6884, | |
| "rewards/accuracies": 0.5749999284744263, | |
| "rewards/chosen": 0.007093862630426884, | |
| "rewards/margins": 0.008020764216780663, | |
| "rewards/rejected": -0.0009269017027691007, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.604261796042618e-07, | |
| "logits/chosen": -1.843636155128479, | |
| "logits/rejected": -1.5651946067810059, | |
| "logps/chosen": -92.92024230957031, | |
| "logps/rejected": -75.78328704833984, | |
| "loss": 0.6864, | |
| "rewards/accuracies": 0.6333333849906921, | |
| "rewards/chosen": 0.007120449095964432, | |
| "rewards/margins": 0.011716886423528194, | |
| "rewards/rejected": -0.004596438258886337, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.5662100456621e-07, | |
| "logits/chosen": -1.8188960552215576, | |
| "logits/rejected": -1.5440260171890259, | |
| "logps/chosen": -91.01820373535156, | |
| "logps/rejected": -72.49656677246094, | |
| "loss": 0.687, | |
| "rewards/accuracies": 0.6166666150093079, | |
| "rewards/chosen": 0.007922597229480743, | |
| "rewards/margins": 0.010534586384892464, | |
| "rewards/rejected": -0.002611987991258502, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.528158295281583e-07, | |
| "logits/chosen": -1.931947946548462, | |
| "logits/rejected": -1.6909589767456055, | |
| "logps/chosen": -88.5571517944336, | |
| "logps/rejected": -72.00991821289062, | |
| "loss": 0.6866, | |
| "rewards/accuracies": 0.5999999642372131, | |
| "rewards/chosen": 0.005753463599830866, | |
| "rewards/margins": 0.010802066884934902, | |
| "rewards/rejected": -0.005048603750765324, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.490106544901065e-07, | |
| "logits/chosen": -1.9274402856826782, | |
| "logits/rejected": -1.6220242977142334, | |
| "logps/chosen": -91.9419174194336, | |
| "logps/rejected": -77.3694076538086, | |
| "loss": 0.6851, | |
| "rewards/accuracies": 0.7083333134651184, | |
| "rewards/chosen": 0.014447471126914024, | |
| "rewards/margins": 0.018269026651978493, | |
| "rewards/rejected": -0.0038215541280806065, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.4520547945205477e-07, | |
| "logits/chosen": -1.8854271173477173, | |
| "logits/rejected": -1.6241188049316406, | |
| "logps/chosen": -92.77326965332031, | |
| "logps/rejected": -77.81119537353516, | |
| "loss": 0.6838, | |
| "rewards/accuracies": 0.64166659116745, | |
| "rewards/chosen": 0.013706192374229431, | |
| "rewards/margins": 0.015467122197151184, | |
| "rewards/rejected": -0.0017609309870749712, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.41400304414003e-07, | |
| "logits/chosen": -1.8017442226409912, | |
| "logits/rejected": -1.5209267139434814, | |
| "logps/chosen": -90.86246490478516, | |
| "logps/rejected": -72.70795440673828, | |
| "loss": 0.6842, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.013066952116787434, | |
| "rewards/margins": 0.018115142360329628, | |
| "rewards/rejected": -0.00504818931221962, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_logits/chosen": -2.051283121109009, | |
| "eval_logits/rejected": -1.7846735715866089, | |
| "eval_logps/chosen": -91.84673309326172, | |
| "eval_logps/rejected": -72.15672302246094, | |
| "eval_loss": 0.6832027435302734, | |
| "eval_rewards/accuracies": 0.6916666626930237, | |
| "eval_rewards/chosen": 0.013898174278438091, | |
| "eval_rewards/margins": 0.01952529139816761, | |
| "eval_rewards/rejected": -0.005627114325761795, | |
| "eval_runtime": 116.9802, | |
| "eval_samples_per_second": 24.466, | |
| "eval_steps_per_second": 0.769, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.375951293759513e-07, | |
| "logits/chosen": -1.8816429376602173, | |
| "logits/rejected": -1.5877922773361206, | |
| "logps/chosen": -92.76510620117188, | |
| "logps/rejected": -77.3439712524414, | |
| "loss": 0.6827, | |
| "rewards/accuracies": 0.7166666388511658, | |
| "rewards/chosen": 0.016739103943109512, | |
| "rewards/margins": 0.020335419103503227, | |
| "rewards/rejected": -0.003596315626055002, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.337899543378995e-07, | |
| "logits/chosen": -1.8938591480255127, | |
| "logits/rejected": -1.6317039728164673, | |
| "logps/chosen": -93.17677307128906, | |
| "logps/rejected": -74.86593627929688, | |
| "loss": 0.6818, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.01637251302599907, | |
| "rewards/margins": 0.019245153293013573, | |
| "rewards/rejected": -0.0028726388700306416, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.2998477929984777e-07, | |
| "logits/chosen": -1.8730627298355103, | |
| "logits/rejected": -1.6184980869293213, | |
| "logps/chosen": -91.06995391845703, | |
| "logps/rejected": -72.5287857055664, | |
| "loss": 0.6808, | |
| "rewards/accuracies": 0.7666667103767395, | |
| "rewards/chosen": 0.01837759278714657, | |
| "rewards/margins": 0.02310130000114441, | |
| "rewards/rejected": -0.004723704420030117, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.26179604261796e-07, | |
| "logits/chosen": -1.8828544616699219, | |
| "logits/rejected": -1.576556921005249, | |
| "logps/chosen": -92.8334732055664, | |
| "logps/rejected": -74.53410339355469, | |
| "loss": 0.6788, | |
| "rewards/accuracies": 0.8166666030883789, | |
| "rewards/chosen": 0.023710301145911217, | |
| "rewards/margins": 0.03184016793966293, | |
| "rewards/rejected": -0.008129866793751717, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.223744292237443e-07, | |
| "logits/chosen": -1.9196786880493164, | |
| "logits/rejected": -1.625704050064087, | |
| "logps/chosen": -90.74723815917969, | |
| "logps/rejected": -76.68948364257812, | |
| "loss": 0.6777, | |
| "rewards/accuracies": 0.8083333969116211, | |
| "rewards/chosen": 0.026256781071424484, | |
| "rewards/margins": 0.03269972652196884, | |
| "rewards/rejected": -0.00644295010715723, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.185692541856925e-07, | |
| "logits/chosen": -1.981180191040039, | |
| "logits/rejected": -1.6961545944213867, | |
| "logps/chosen": -93.02009582519531, | |
| "logps/rejected": -76.1031723022461, | |
| "loss": 0.6762, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.03220153599977493, | |
| "rewards/margins": 0.03597740828990936, | |
| "rewards/rejected": -0.0037758699618279934, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.1476407914764077e-07, | |
| "logits/chosen": -1.8603204488754272, | |
| "logits/rejected": -1.5979934930801392, | |
| "logps/chosen": -94.58716583251953, | |
| "logps/rejected": -75.44288635253906, | |
| "loss": 0.6759, | |
| "rewards/accuracies": 0.841666579246521, | |
| "rewards/chosen": 0.021503183990716934, | |
| "rewards/margins": 0.03375691920518875, | |
| "rewards/rejected": -0.012253734283149242, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.10958904109589e-07, | |
| "logits/chosen": -1.9121630191802979, | |
| "logits/rejected": -1.613445520401001, | |
| "logps/chosen": -93.34215545654297, | |
| "logps/rejected": -74.89527893066406, | |
| "loss": 0.6739, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.02675667405128479, | |
| "rewards/margins": 0.03688011318445206, | |
| "rewards/rejected": -0.010123440995812416, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.071537290715373e-07, | |
| "logits/chosen": -1.9191780090332031, | |
| "logits/rejected": -1.532434105873108, | |
| "logps/chosen": -94.9774169921875, | |
| "logps/rejected": -73.49574279785156, | |
| "loss": 0.6744, | |
| "rewards/accuracies": 0.8666666150093079, | |
| "rewards/chosen": 0.03616553544998169, | |
| "rewards/margins": 0.048189498484134674, | |
| "rewards/rejected": -0.012023964896798134, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.033485540334855e-07, | |
| "logits/chosen": -1.8837106227874756, | |
| "logits/rejected": -1.6179090738296509, | |
| "logps/chosen": -93.18563079833984, | |
| "logps/rejected": -75.48190307617188, | |
| "loss": 0.672, | |
| "rewards/accuracies": 0.8333333730697632, | |
| "rewards/chosen": 0.03295673802495003, | |
| "rewards/margins": 0.044657547026872635, | |
| "rewards/rejected": -0.011700802482664585, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_logits/chosen": -2.0504438877105713, | |
| "eval_logits/rejected": -1.78355073928833, | |
| "eval_logps/chosen": -91.70490264892578, | |
| "eval_logps/rejected": -72.23121643066406, | |
| "eval_loss": 0.671801745891571, | |
| "eval_rewards/accuracies": 0.824999988079071, | |
| "eval_rewards/chosen": 0.028079798445105553, | |
| "eval_rewards/margins": 0.041155941784381866, | |
| "eval_rewards/rejected": -0.013076143339276314, | |
| "eval_runtime": 122.6508, | |
| "eval_samples_per_second": 23.335, | |
| "eval_steps_per_second": 0.734, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 3.9954337899543377e-07, | |
| "logits/chosen": -1.9980239868164062, | |
| "logits/rejected": -1.710172414779663, | |
| "logps/chosen": -90.13986206054688, | |
| "logps/rejected": -74.13956451416016, | |
| "loss": 0.6726, | |
| "rewards/accuracies": 0.7833333015441895, | |
| "rewards/chosen": 0.026057172566652298, | |
| "rewards/margins": 0.03604119271039963, | |
| "rewards/rejected": -0.00998402014374733, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 3.95738203957382e-07, | |
| "logits/chosen": -1.9292805194854736, | |
| "logits/rejected": -1.6599111557006836, | |
| "logps/chosen": -89.1054458618164, | |
| "logps/rejected": -76.2321548461914, | |
| "loss": 0.67, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.03600526601076126, | |
| "rewards/margins": 0.04854360967874527, | |
| "rewards/rejected": -0.012538343667984009, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 3.919330289193303e-07, | |
| "logits/chosen": -1.8889102935791016, | |
| "logits/rejected": -1.5878620147705078, | |
| "logps/chosen": -94.20903015136719, | |
| "logps/rejected": -75.64437103271484, | |
| "loss": 0.6677, | |
| "rewards/accuracies": 0.8833333253860474, | |
| "rewards/chosen": 0.04072072356939316, | |
| "rewards/margins": 0.05353847146034241, | |
| "rewards/rejected": -0.012817745096981525, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.881278538812785e-07, | |
| "logits/chosen": -1.889439344406128, | |
| "logits/rejected": -1.5668418407440186, | |
| "logps/chosen": -92.68305206298828, | |
| "logps/rejected": -71.53545379638672, | |
| "loss": 0.6653, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.042834434658288956, | |
| "rewards/margins": 0.05957134813070297, | |
| "rewards/rejected": -0.016736917197704315, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.8432267884322677e-07, | |
| "logits/chosen": -1.8060448169708252, | |
| "logits/rejected": -1.5509716272354126, | |
| "logps/chosen": -88.83539581298828, | |
| "logps/rejected": -71.70087432861328, | |
| "loss": 0.6645, | |
| "rewards/accuracies": 0.8416666984558105, | |
| "rewards/chosen": 0.04349132627248764, | |
| "rewards/margins": 0.05466403439640999, | |
| "rewards/rejected": -0.011172705329954624, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.80517503805175e-07, | |
| "logits/chosen": -1.8319886922836304, | |
| "logits/rejected": -1.5673277378082275, | |
| "logps/chosen": -88.6297378540039, | |
| "logps/rejected": -74.78864288330078, | |
| "loss": 0.6631, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.044696420431137085, | |
| "rewards/margins": 0.0604400709271431, | |
| "rewards/rejected": -0.015743646770715714, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.767123287671233e-07, | |
| "logits/chosen": -1.9542433023452759, | |
| "logits/rejected": -1.641196608543396, | |
| "logps/chosen": -99.97395324707031, | |
| "logps/rejected": -78.014404296875, | |
| "loss": 0.6626, | |
| "rewards/accuracies": 0.9166666865348816, | |
| "rewards/chosen": 0.047287195920944214, | |
| "rewards/margins": 0.06728993356227875, | |
| "rewards/rejected": -0.020002741366624832, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.729071537290715e-07, | |
| "logits/chosen": -1.916868805885315, | |
| "logits/rejected": -1.6233441829681396, | |
| "logps/chosen": -97.3001937866211, | |
| "logps/rejected": -76.58578491210938, | |
| "loss": 0.6619, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.05124374479055405, | |
| "rewards/margins": 0.07088983058929443, | |
| "rewards/rejected": -0.019646091386675835, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.6910197869101977e-07, | |
| "logits/chosen": -1.9672422409057617, | |
| "logits/rejected": -1.700531244277954, | |
| "logps/chosen": -89.29413604736328, | |
| "logps/rejected": -72.65511322021484, | |
| "loss": 0.66, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.050498634576797485, | |
| "rewards/margins": 0.06749961525201797, | |
| "rewards/rejected": -0.01700098067522049, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.65296803652968e-07, | |
| "logits/chosen": -1.833929419517517, | |
| "logits/rejected": -1.5389481782913208, | |
| "logps/chosen": -85.8098373413086, | |
| "logps/rejected": -72.5254135131836, | |
| "loss": 0.6563, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.05660303682088852, | |
| "rewards/margins": 0.0805855318903923, | |
| "rewards/rejected": -0.023982489481568336, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_logits/chosen": -2.0494003295898438, | |
| "eval_logits/rejected": -1.7820732593536377, | |
| "eval_logps/chosen": -91.48755645751953, | |
| "eval_logps/rejected": -72.31159973144531, | |
| "eval_loss": 0.6574758887290955, | |
| "eval_rewards/accuracies": 0.8861111402511597, | |
| "eval_rewards/chosen": 0.049815867096185684, | |
| "eval_rewards/margins": 0.0709303691983223, | |
| "eval_rewards/rejected": -0.021114489063620567, | |
| "eval_runtime": 126.0719, | |
| "eval_samples_per_second": 22.701, | |
| "eval_steps_per_second": 0.714, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 3.614916286149163e-07, | |
| "logits/chosen": -1.913190484046936, | |
| "logits/rejected": -1.6543632745742798, | |
| "logps/chosen": -89.34779357910156, | |
| "logps/rejected": -71.7219009399414, | |
| "loss": 0.6567, | |
| "rewards/accuracies": 0.8916667103767395, | |
| "rewards/chosen": 0.05136920139193535, | |
| "rewards/margins": 0.07187855988740921, | |
| "rewards/rejected": -0.02050935849547386, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.576864535768645e-07, | |
| "logits/chosen": -1.8918750286102295, | |
| "logits/rejected": -1.621498465538025, | |
| "logps/chosen": -92.99454498291016, | |
| "logps/rejected": -72.34910583496094, | |
| "loss": 0.6569, | |
| "rewards/accuracies": 0.841666579246521, | |
| "rewards/chosen": 0.05106909200549126, | |
| "rewards/margins": 0.0683208703994751, | |
| "rewards/rejected": -0.01725177839398384, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.5388127853881277e-07, | |
| "logits/chosen": -1.8910505771636963, | |
| "logits/rejected": -1.6094516515731812, | |
| "logps/chosen": -92.9715347290039, | |
| "logps/rejected": -75.84952545166016, | |
| "loss": 0.6559, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.05335830897092819, | |
| "rewards/margins": 0.07476408034563065, | |
| "rewards/rejected": -0.021405773237347603, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.50076103500761e-07, | |
| "logits/chosen": -1.8273894786834717, | |
| "logits/rejected": -1.5480557680130005, | |
| "logps/chosen": -93.09770202636719, | |
| "logps/rejected": -77.72835540771484, | |
| "loss": 0.6531, | |
| "rewards/accuracies": 0.8833333253860474, | |
| "rewards/chosen": 0.06002092361450195, | |
| "rewards/margins": 0.08498416841030121, | |
| "rewards/rejected": -0.02496323548257351, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.462709284627093e-07, | |
| "logits/chosen": -1.7952553033828735, | |
| "logits/rejected": -1.5454210042953491, | |
| "logps/chosen": -92.24395751953125, | |
| "logps/rejected": -74.6566390991211, | |
| "loss": 0.6514, | |
| "rewards/accuracies": 0.8666666746139526, | |
| "rewards/chosen": 0.06232692673802376, | |
| "rewards/margins": 0.08038316667079926, | |
| "rewards/rejected": -0.018056249246001244, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.424657534246575e-07, | |
| "logits/chosen": -1.8805965185165405, | |
| "logits/rejected": -1.586458444595337, | |
| "logps/chosen": -96.27812194824219, | |
| "logps/rejected": -75.61478424072266, | |
| "loss": 0.6487, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.06480798870325089, | |
| "rewards/margins": 0.09341150522232056, | |
| "rewards/rejected": -0.02860351838171482, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.3866057838660576e-07, | |
| "logits/chosen": -1.8029924631118774, | |
| "logits/rejected": -1.5792688131332397, | |
| "logps/chosen": -87.8210678100586, | |
| "logps/rejected": -74.4839096069336, | |
| "loss": 0.651, | |
| "rewards/accuracies": 0.8749998807907104, | |
| "rewards/chosen": 0.05512385442852974, | |
| "rewards/margins": 0.07892224937677383, | |
| "rewards/rejected": -0.023798387497663498, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.34855403348554e-07, | |
| "logits/chosen": -1.913487434387207, | |
| "logits/rejected": -1.612151861190796, | |
| "logps/chosen": -94.51315307617188, | |
| "logps/rejected": -74.9591293334961, | |
| "loss": 0.6465, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.0840396136045456, | |
| "rewards/margins": 0.107123002409935, | |
| "rewards/rejected": -0.023083383217453957, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.310502283105023e-07, | |
| "logits/chosen": -1.9018150568008423, | |
| "logits/rejected": -1.6232668161392212, | |
| "logps/chosen": -93.73323822021484, | |
| "logps/rejected": -75.38792419433594, | |
| "loss": 0.6435, | |
| "rewards/accuracies": 0.8666666150093079, | |
| "rewards/chosen": 0.07307516038417816, | |
| "rewards/margins": 0.10251389443874359, | |
| "rewards/rejected": -0.029438745230436325, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.272450532724505e-07, | |
| "logits/chosen": -1.8873398303985596, | |
| "logits/rejected": -1.6442911624908447, | |
| "logps/chosen": -90.5989761352539, | |
| "logps/rejected": -77.07279968261719, | |
| "loss": 0.6437, | |
| "rewards/accuracies": 0.966666579246521, | |
| "rewards/chosen": 0.08184785395860672, | |
| "rewards/margins": 0.11672016233205795, | |
| "rewards/rejected": -0.034872300922870636, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_logits/chosen": -2.048579692840576, | |
| "eval_logits/rejected": -1.780737280845642, | |
| "eval_logps/chosen": -91.28104400634766, | |
| "eval_logps/rejected": -72.44007110595703, | |
| "eval_loss": 0.6415905356407166, | |
| "eval_rewards/accuracies": 0.9111111164093018, | |
| "eval_rewards/chosen": 0.07046664506196976, | |
| "eval_rewards/margins": 0.10442798584699631, | |
| "eval_rewards/rejected": -0.03396133333444595, | |
| "eval_runtime": 117.9826, | |
| "eval_samples_per_second": 24.258, | |
| "eval_steps_per_second": 0.763, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.2343987823439876e-07, | |
| "logits/chosen": -1.9880199432373047, | |
| "logits/rejected": -1.676412582397461, | |
| "logps/chosen": -91.00142669677734, | |
| "logps/rejected": -73.52288818359375, | |
| "loss": 0.6401, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.07217723876237869, | |
| "rewards/margins": 0.1089334487915039, | |
| "rewards/rejected": -0.036756210029125214, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.19634703196347e-07, | |
| "logits/chosen": -1.8773486614227295, | |
| "logits/rejected": -1.573994755744934, | |
| "logps/chosen": -89.79063415527344, | |
| "logps/rejected": -73.29415130615234, | |
| "loss": 0.6391, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.06451607495546341, | |
| "rewards/margins": 0.09732060134410858, | |
| "rewards/rejected": -0.032804541289806366, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.158295281582953e-07, | |
| "logits/chosen": -1.8495439291000366, | |
| "logits/rejected": -1.5227216482162476, | |
| "logps/chosen": -93.12200164794922, | |
| "logps/rejected": -74.12937927246094, | |
| "loss": 0.6371, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.08449102938175201, | |
| "rewards/margins": 0.11922381818294525, | |
| "rewards/rejected": -0.034732796251773834, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.120243531202435e-07, | |
| "logits/chosen": -1.93155837059021, | |
| "logits/rejected": -1.6498839855194092, | |
| "logps/chosen": -94.86695861816406, | |
| "logps/rejected": -76.8067626953125, | |
| "loss": 0.6354, | |
| "rewards/accuracies": 0.8833333253860474, | |
| "rewards/chosen": 0.08870759606361389, | |
| "rewards/margins": 0.12238001823425293, | |
| "rewards/rejected": -0.03367242217063904, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.0821917808219176e-07, | |
| "logits/chosen": -1.8748143911361694, | |
| "logits/rejected": -1.6098239421844482, | |
| "logps/chosen": -91.36283874511719, | |
| "logps/rejected": -73.6336669921875, | |
| "loss": 0.6328, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.0779472142457962, | |
| "rewards/margins": 0.1164277195930481, | |
| "rewards/rejected": -0.038480497896671295, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.0441400304414e-07, | |
| "logits/chosen": -1.8293393850326538, | |
| "logits/rejected": -1.4808156490325928, | |
| "logps/chosen": -94.16930389404297, | |
| "logps/rejected": -72.66043853759766, | |
| "loss": 0.6341, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.09546177834272385, | |
| "rewards/margins": 0.1290528029203415, | |
| "rewards/rejected": -0.033591024577617645, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.006088280060883e-07, | |
| "logits/chosen": -1.8831846714019775, | |
| "logits/rejected": -1.577532172203064, | |
| "logps/chosen": -95.71855926513672, | |
| "logps/rejected": -74.64360046386719, | |
| "loss": 0.6305, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.09222020953893661, | |
| "rewards/margins": 0.1298314929008484, | |
| "rewards/rejected": -0.037611283361911774, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2.968036529680365e-07, | |
| "logits/chosen": -1.884874701499939, | |
| "logits/rejected": -1.56434965133667, | |
| "logps/chosen": -91.71931457519531, | |
| "logps/rejected": -74.07929992675781, | |
| "loss": 0.6298, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.09030507504940033, | |
| "rewards/margins": 0.13949953019618988, | |
| "rewards/rejected": -0.04919447377324104, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2.9299847792998476e-07, | |
| "logits/chosen": -1.939512848854065, | |
| "logits/rejected": -1.6418523788452148, | |
| "logps/chosen": -88.81291198730469, | |
| "logps/rejected": -71.72727966308594, | |
| "loss": 0.6255, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.1102442592382431, | |
| "rewards/margins": 0.14661459624767303, | |
| "rewards/rejected": -0.036370351910591125, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.89193302891933e-07, | |
| "logits/chosen": -1.8088592290878296, | |
| "logits/rejected": -1.554652452468872, | |
| "logps/chosen": -88.36043548583984, | |
| "logps/rejected": -74.16236877441406, | |
| "loss": 0.6261, | |
| "rewards/accuracies": 0.9416666030883789, | |
| "rewards/chosen": 0.09904900938272476, | |
| "rewards/margins": 0.13188037276268005, | |
| "rewards/rejected": -0.032831382006406784, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_logits/chosen": -2.0478439331054688, | |
| "eval_logits/rejected": -1.7795759439468384, | |
| "eval_logps/chosen": -91.1009750366211, | |
| "eval_logps/rejected": -72.53546905517578, | |
| "eval_loss": 0.6277271509170532, | |
| "eval_rewards/accuracies": 0.925000011920929, | |
| "eval_rewards/chosen": 0.08847405016422272, | |
| "eval_rewards/margins": 0.13197554647922516, | |
| "eval_rewards/rejected": -0.04350150376558304, | |
| "eval_runtime": 117.9463, | |
| "eval_samples_per_second": 24.265, | |
| "eval_steps_per_second": 0.763, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.853881278538813e-07, | |
| "logits/chosen": -1.8632869720458984, | |
| "logits/rejected": -1.598181962966919, | |
| "logps/chosen": -89.87626647949219, | |
| "logps/rejected": -74.79205322265625, | |
| "loss": 0.6285, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.09145402163267136, | |
| "rewards/margins": 0.13780589401721954, | |
| "rewards/rejected": -0.04635186120867729, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.815829528158295e-07, | |
| "logits/chosen": -1.8058583736419678, | |
| "logits/rejected": -1.5232179164886475, | |
| "logps/chosen": -87.31756591796875, | |
| "logps/rejected": -72.79600524902344, | |
| "loss": 0.6247, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.09649817645549774, | |
| "rewards/margins": 0.1376573145389557, | |
| "rewards/rejected": -0.041159145534038544, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.7777777777777776e-07, | |
| "logits/chosen": -1.8448421955108643, | |
| "logits/rejected": -1.6317142248153687, | |
| "logps/chosen": -87.55412292480469, | |
| "logps/rejected": -75.91432189941406, | |
| "loss": 0.6245, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.10163182020187378, | |
| "rewards/margins": 0.14317932724952698, | |
| "rewards/rejected": -0.0415474958717823, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.73972602739726e-07, | |
| "logits/chosen": -1.800663709640503, | |
| "logits/rejected": -1.5165865421295166, | |
| "logps/chosen": -92.58036041259766, | |
| "logps/rejected": -73.22102355957031, | |
| "loss": 0.6216, | |
| "rewards/accuracies": 0.908333420753479, | |
| "rewards/chosen": 0.09606580436229706, | |
| "rewards/margins": 0.14485225081443787, | |
| "rewards/rejected": -0.04878643900156021, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.701674277016743e-07, | |
| "logits/chosen": -1.884385347366333, | |
| "logits/rejected": -1.6005618572235107, | |
| "logps/chosen": -94.69025421142578, | |
| "logps/rejected": -74.93373107910156, | |
| "loss": 0.6225, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.11928985267877579, | |
| "rewards/margins": 0.16091035306453705, | |
| "rewards/rejected": -0.04162050783634186, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.663622526636225e-07, | |
| "logits/chosen": -1.8744373321533203, | |
| "logits/rejected": -1.5992999076843262, | |
| "logps/chosen": -93.0952377319336, | |
| "logps/rejected": -76.86521911621094, | |
| "loss": 0.6187, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.12268207967281342, | |
| "rewards/margins": 0.15957853198051453, | |
| "rewards/rejected": -0.036896444857120514, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6255707762557076e-07, | |
| "logits/chosen": -1.9210466146469116, | |
| "logits/rejected": -1.6140285730361938, | |
| "logps/chosen": -94.59263610839844, | |
| "logps/rejected": -75.74002838134766, | |
| "loss": 0.6174, | |
| "rewards/accuracies": 0.908333420753479, | |
| "rewards/chosen": 0.09473783522844315, | |
| "rewards/margins": 0.14541365206241608, | |
| "rewards/rejected": -0.050675809383392334, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.58751902587519e-07, | |
| "logits/chosen": -1.9041986465454102, | |
| "logits/rejected": -1.643701195716858, | |
| "logps/chosen": -92.38731384277344, | |
| "logps/rejected": -76.4394302368164, | |
| "loss": 0.6147, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.09374178946018219, | |
| "rewards/margins": 0.14521454274654388, | |
| "rewards/rejected": -0.05147276073694229, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.549467275494673e-07, | |
| "logits/chosen": -1.8866844177246094, | |
| "logits/rejected": -1.5806655883789062, | |
| "logps/chosen": -89.06953430175781, | |
| "logps/rejected": -75.25362396240234, | |
| "loss": 0.6127, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.1129077672958374, | |
| "rewards/margins": 0.1759023219347, | |
| "rewards/rejected": -0.06299454718828201, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.511415525114155e-07, | |
| "logits/chosen": -1.9603450298309326, | |
| "logits/rejected": -1.658216118812561, | |
| "logps/chosen": -91.58163452148438, | |
| "logps/rejected": -74.63312530517578, | |
| "loss": 0.6117, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.1321893036365509, | |
| "rewards/margins": 0.18407562375068665, | |
| "rewards/rejected": -0.051886312663555145, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_logits/chosen": -2.047370672225952, | |
| "eval_logits/rejected": -1.7785520553588867, | |
| "eval_logps/chosen": -90.88909149169922, | |
| "eval_logps/rejected": -72.66747283935547, | |
| "eval_loss": 0.6126503348350525, | |
| "eval_rewards/accuracies": 0.9222221970558167, | |
| "eval_rewards/chosen": 0.10966197401285172, | |
| "eval_rewards/margins": 0.16636402904987335, | |
| "eval_rewards/rejected": -0.05670207738876343, | |
| "eval_runtime": 117.9996, | |
| "eval_samples_per_second": 24.254, | |
| "eval_steps_per_second": 0.763, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4733637747336376e-07, | |
| "logits/chosen": -1.9528766870498657, | |
| "logits/rejected": -1.594366192817688, | |
| "logps/chosen": -97.9262466430664, | |
| "logps/rejected": -73.47969055175781, | |
| "loss": 0.6145, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.12691722810268402, | |
| "rewards/margins": 0.1831960827112198, | |
| "rewards/rejected": -0.056278862059116364, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.43531202435312e-07, | |
| "logits/chosen": -1.9507348537445068, | |
| "logits/rejected": -1.6255791187286377, | |
| "logps/chosen": -93.60541534423828, | |
| "logps/rejected": -74.82536315917969, | |
| "loss": 0.6094, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.11346453428268433, | |
| "rewards/margins": 0.17843127250671387, | |
| "rewards/rejected": -0.06496672332286835, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3972602739726023e-07, | |
| "logits/chosen": -1.7865006923675537, | |
| "logits/rejected": -1.5033773183822632, | |
| "logps/chosen": -89.35474395751953, | |
| "logps/rejected": -70.51929473876953, | |
| "loss": 0.6081, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.1159995049238205, | |
| "rewards/margins": 0.1914626657962799, | |
| "rewards/rejected": -0.07546313852071762, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.359208523592085e-07, | |
| "logits/chosen": -1.9572938680648804, | |
| "logits/rejected": -1.6931850910186768, | |
| "logps/chosen": -90.90480041503906, | |
| "logps/rejected": -77.85359191894531, | |
| "loss": 0.6088, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.10320776700973511, | |
| "rewards/margins": 0.16814057528972626, | |
| "rewards/rejected": -0.06493280827999115, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.3211567732115676e-07, | |
| "logits/chosen": -2.0167603492736816, | |
| "logits/rejected": -1.715150237083435, | |
| "logps/chosen": -90.37574768066406, | |
| "logps/rejected": -77.0097427368164, | |
| "loss": 0.6039, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.1519448608160019, | |
| "rewards/margins": 0.21030330657958984, | |
| "rewards/rejected": -0.058358438313007355, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.28310502283105e-07, | |
| "logits/chosen": -1.8651390075683594, | |
| "logits/rejected": -1.5632587671279907, | |
| "logps/chosen": -94.64379119873047, | |
| "logps/rejected": -75.53591918945312, | |
| "loss": 0.6078, | |
| "rewards/accuracies": 0.9416667222976685, | |
| "rewards/chosen": 0.13561630249023438, | |
| "rewards/margins": 0.19979830086231232, | |
| "rewards/rejected": -0.06418199837207794, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.2450532724505325e-07, | |
| "logits/chosen": -1.9391330480575562, | |
| "logits/rejected": -1.663351058959961, | |
| "logps/chosen": -86.51952362060547, | |
| "logps/rejected": -73.44007873535156, | |
| "loss": 0.6057, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.1343904286623001, | |
| "rewards/margins": 0.1892845332622528, | |
| "rewards/rejected": -0.054894138127565384, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.207001522070015e-07, | |
| "logits/chosen": -1.842283010482788, | |
| "logits/rejected": -1.5807665586471558, | |
| "logps/chosen": -90.40550231933594, | |
| "logps/rejected": -77.0767593383789, | |
| "loss": 0.6045, | |
| "rewards/accuracies": 0.9249998927116394, | |
| "rewards/chosen": 0.13356170058250427, | |
| "rewards/margins": 0.1978417932987213, | |
| "rewards/rejected": -0.06428009271621704, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.1689497716894975e-07, | |
| "logits/chosen": -1.913578748703003, | |
| "logits/rejected": -1.648938536643982, | |
| "logps/chosen": -87.49850463867188, | |
| "logps/rejected": -74.8587646484375, | |
| "loss": 0.6026, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.13597458600997925, | |
| "rewards/margins": 0.20577768981456757, | |
| "rewards/rejected": -0.06980310380458832, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2.13089802130898e-07, | |
| "logits/chosen": -1.8712265491485596, | |
| "logits/rejected": -1.5505828857421875, | |
| "logps/chosen": -92.89669036865234, | |
| "logps/rejected": -77.97479248046875, | |
| "loss": 0.6002, | |
| "rewards/accuracies": 0.9416667222976685, | |
| "rewards/chosen": 0.1514441967010498, | |
| "rewards/margins": 0.2083124816417694, | |
| "rewards/rejected": -0.056868284940719604, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "eval_logits/chosen": -2.046788454055786, | |
| "eval_logits/rejected": -1.7776765823364258, | |
| "eval_logps/chosen": -90.75981140136719, | |
| "eval_logps/rejected": -72.78363037109375, | |
| "eval_loss": 0.6019285321235657, | |
| "eval_rewards/accuracies": 0.9277777671813965, | |
| "eval_rewards/chosen": 0.12258908152580261, | |
| "eval_rewards/margins": 0.1909066140651703, | |
| "eval_rewards/rejected": -0.06831753998994827, | |
| "eval_runtime": 117.8634, | |
| "eval_samples_per_second": 24.282, | |
| "eval_steps_per_second": 0.764, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2.0928462709284625e-07, | |
| "logits/chosen": -1.7578074932098389, | |
| "logits/rejected": -1.445682406425476, | |
| "logps/chosen": -92.38932800292969, | |
| "logps/rejected": -74.73234558105469, | |
| "loss": 0.5977, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.14447703957557678, | |
| "rewards/margins": 0.2118067443370819, | |
| "rewards/rejected": -0.06732969731092453, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2.054794520547945e-07, | |
| "logits/chosen": -1.9356153011322021, | |
| "logits/rejected": -1.6309950351715088, | |
| "logps/chosen": -90.90232849121094, | |
| "logps/rejected": -75.2989730834961, | |
| "loss": 0.5962, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.14640390872955322, | |
| "rewards/margins": 0.21756890416145325, | |
| "rewards/rejected": -0.07116499543190002, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2.0167427701674275e-07, | |
| "logits/chosen": -1.903969407081604, | |
| "logits/rejected": -1.6307262182235718, | |
| "logps/chosen": -91.58943939208984, | |
| "logps/rejected": -75.18601989746094, | |
| "loss": 0.6003, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.13933506608009338, | |
| "rewards/margins": 0.20189881324768066, | |
| "rewards/rejected": -0.06256375461816788, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.97869101978691e-07, | |
| "logits/chosen": -1.866371512413025, | |
| "logits/rejected": -1.6297776699066162, | |
| "logps/chosen": -91.3348617553711, | |
| "logps/rejected": -75.30091857910156, | |
| "loss": 0.5968, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.16422715783119202, | |
| "rewards/margins": 0.22424063086509705, | |
| "rewards/rejected": -0.06001347303390503, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.9406392694063925e-07, | |
| "logits/chosen": -1.860414743423462, | |
| "logits/rejected": -1.5813719034194946, | |
| "logps/chosen": -94.010498046875, | |
| "logps/rejected": -73.02679443359375, | |
| "loss": 0.595, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.14462696015834808, | |
| "rewards/margins": 0.22170260548591614, | |
| "rewards/rejected": -0.07707564532756805, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.902587519025875e-07, | |
| "logits/chosen": -1.755253791809082, | |
| "logits/rejected": -1.5063341856002808, | |
| "logps/chosen": -89.60089111328125, | |
| "logps/rejected": -75.72676086425781, | |
| "loss": 0.5942, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.12570711970329285, | |
| "rewards/margins": 0.19667670130729675, | |
| "rewards/rejected": -0.0709695890545845, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.8645357686453575e-07, | |
| "logits/chosen": -1.9189532995224, | |
| "logits/rejected": -1.60482919216156, | |
| "logps/chosen": -91.0718765258789, | |
| "logps/rejected": -76.86122131347656, | |
| "loss": 0.5919, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.14830803871154785, | |
| "rewards/margins": 0.22014153003692627, | |
| "rewards/rejected": -0.07183349877595901, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.82648401826484e-07, | |
| "logits/chosen": -1.9582935571670532, | |
| "logits/rejected": -1.6904990673065186, | |
| "logps/chosen": -92.6358871459961, | |
| "logps/rejected": -75.13966369628906, | |
| "loss": 0.593, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.12001453340053558, | |
| "rewards/margins": 0.20278160274028778, | |
| "rewards/rejected": -0.0827670693397522, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.7884322678843225e-07, | |
| "logits/chosen": -1.7855488061904907, | |
| "logits/rejected": -1.498254418373108, | |
| "logps/chosen": -89.93516540527344, | |
| "logps/rejected": -71.06876373291016, | |
| "loss": 0.5923, | |
| "rewards/accuracies": 0.9666666984558105, | |
| "rewards/chosen": 0.13533183932304382, | |
| "rewards/margins": 0.2196374386548996, | |
| "rewards/rejected": -0.08430557698011398, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.750380517503805e-07, | |
| "logits/chosen": -1.9469058513641357, | |
| "logits/rejected": -1.655491828918457, | |
| "logps/chosen": -92.86201477050781, | |
| "logps/rejected": -77.54034423828125, | |
| "loss": 0.5912, | |
| "rewards/accuracies": 0.9499999284744263, | |
| "rewards/chosen": 0.14615695178508759, | |
| "rewards/margins": 0.22164049744606018, | |
| "rewards/rejected": -0.0754835233092308, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_logits/chosen": -2.046569585800171, | |
| "eval_logits/rejected": -1.7769988775253296, | |
| "eval_logps/chosen": -90.64215850830078, | |
| "eval_logps/rejected": -72.90531921386719, | |
| "eval_loss": 0.5911818742752075, | |
| "eval_rewards/accuracies": 0.9333333373069763, | |
| "eval_rewards/chosen": 0.13435469567775726, | |
| "eval_rewards/margins": 0.21484099328517914, | |
| "eval_rewards/rejected": -0.08048629015684128, | |
| "eval_runtime": 117.9339, | |
| "eval_samples_per_second": 24.268, | |
| "eval_steps_per_second": 0.763, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.7123287671232875e-07, | |
| "logits/chosen": -1.940800666809082, | |
| "logits/rejected": -1.6664402484893799, | |
| "logps/chosen": -86.94990539550781, | |
| "logps/rejected": -73.69987487792969, | |
| "loss": 0.5931, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.09665495157241821, | |
| "rewards/margins": 0.18213674426078796, | |
| "rewards/rejected": -0.08548180013895035, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.67427701674277e-07, | |
| "logits/chosen": -1.9208215475082397, | |
| "logits/rejected": -1.6697795391082764, | |
| "logps/chosen": -90.12159729003906, | |
| "logps/rejected": -75.43165588378906, | |
| "loss": 0.5936, | |
| "rewards/accuracies": 0.9083331823348999, | |
| "rewards/chosen": 0.1299724578857422, | |
| "rewards/margins": 0.21237368881702423, | |
| "rewards/rejected": -0.08240120857954025, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.6362252663622525e-07, | |
| "logits/chosen": -1.8031762838363647, | |
| "logits/rejected": -1.552947759628296, | |
| "logps/chosen": -93.24359130859375, | |
| "logps/rejected": -73.16694641113281, | |
| "loss": 0.595, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.14134086668491364, | |
| "rewards/margins": 0.22129371762275696, | |
| "rewards/rejected": -0.07995286583900452, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.598173515981735e-07, | |
| "logits/chosen": -1.9094417095184326, | |
| "logits/rejected": -1.6032779216766357, | |
| "logps/chosen": -95.291015625, | |
| "logps/rejected": -76.30043029785156, | |
| "loss": 0.588, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.17578403651714325, | |
| "rewards/margins": 0.24385884404182434, | |
| "rewards/rejected": -0.0680748000741005, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.5601217656012175e-07, | |
| "logits/chosen": -1.8715407848358154, | |
| "logits/rejected": -1.5754361152648926, | |
| "logps/chosen": -93.08331298828125, | |
| "logps/rejected": -71.55984497070312, | |
| "loss": 0.5885, | |
| "rewards/accuracies": 0.908333420753479, | |
| "rewards/chosen": 0.16208195686340332, | |
| "rewards/margins": 0.24299950897693634, | |
| "rewards/rejected": -0.08091756701469421, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.5220700152207e-07, | |
| "logits/chosen": -1.9729465246200562, | |
| "logits/rejected": -1.6472351551055908, | |
| "logps/chosen": -96.95377349853516, | |
| "logps/rejected": -74.73013305664062, | |
| "loss": 0.5829, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.17102012038230896, | |
| "rewards/margins": 0.26194968819618225, | |
| "rewards/rejected": -0.09092956781387329, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.4840182648401825e-07, | |
| "logits/chosen": -1.87423574924469, | |
| "logits/rejected": -1.5873512029647827, | |
| "logps/chosen": -92.32403564453125, | |
| "logps/rejected": -77.21125793457031, | |
| "loss": 0.5842, | |
| "rewards/accuracies": 0.966666579246521, | |
| "rewards/chosen": 0.1644188016653061, | |
| "rewards/margins": 0.27040743827819824, | |
| "rewards/rejected": -0.10598863661289215, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.445966514459665e-07, | |
| "logits/chosen": -1.8510887622833252, | |
| "logits/rejected": -1.5857640504837036, | |
| "logps/chosen": -92.05542755126953, | |
| "logps/rejected": -77.68433380126953, | |
| "loss": 0.5877, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.146043062210083, | |
| "rewards/margins": 0.22200524806976318, | |
| "rewards/rejected": -0.07596220076084137, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.4079147640791475e-07, | |
| "logits/chosen": -1.8213762044906616, | |
| "logits/rejected": -1.5474189519882202, | |
| "logps/chosen": -91.6738052368164, | |
| "logps/rejected": -74.36346435546875, | |
| "loss": 0.5796, | |
| "rewards/accuracies": 0.9833332896232605, | |
| "rewards/chosen": 0.1721878945827484, | |
| "rewards/margins": 0.26567989587783813, | |
| "rewards/rejected": -0.09349202364683151, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.36986301369863e-07, | |
| "logits/chosen": -1.8960554599761963, | |
| "logits/rejected": -1.5947376489639282, | |
| "logps/chosen": -95.58143615722656, | |
| "logps/rejected": -75.37010192871094, | |
| "loss": 0.5822, | |
| "rewards/accuracies": 0.9166666865348816, | |
| "rewards/chosen": 0.18196699023246765, | |
| "rewards/margins": 0.25899866223335266, | |
| "rewards/rejected": -0.0770316869020462, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_logits/chosen": -2.0461771488189697, | |
| "eval_logits/rejected": -1.7763375043869019, | |
| "eval_logps/chosen": -90.54474639892578, | |
| "eval_logps/rejected": -73.00917053222656, | |
| "eval_loss": 0.5822051763534546, | |
| "eval_rewards/accuracies": 0.9472222328186035, | |
| "eval_rewards/chosen": 0.14409679174423218, | |
| "eval_rewards/margins": 0.23496907949447632, | |
| "eval_rewards/rejected": -0.09087225794792175, | |
| "eval_runtime": 117.8174, | |
| "eval_samples_per_second": 24.292, | |
| "eval_steps_per_second": 0.764, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.3318112633181125e-07, | |
| "logits/chosen": -1.893864393234253, | |
| "logits/rejected": -1.5917404890060425, | |
| "logps/chosen": -90.5267105102539, | |
| "logps/rejected": -73.85464477539062, | |
| "loss": 0.5809, | |
| "rewards/accuracies": 0.9666666984558105, | |
| "rewards/chosen": 0.18791857361793518, | |
| "rewards/margins": 0.26672154664993286, | |
| "rewards/rejected": -0.07880295813083649, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.293759512937595e-07, | |
| "logits/chosen": -1.851485013961792, | |
| "logits/rejected": -1.5561844110488892, | |
| "logps/chosen": -91.46730041503906, | |
| "logps/rejected": -75.5124740600586, | |
| "loss": 0.5803, | |
| "rewards/accuracies": 0.9083333015441895, | |
| "rewards/chosen": 0.15885117650032043, | |
| "rewards/margins": 0.24810326099395752, | |
| "rewards/rejected": -0.08925210684537888, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.2557077625570775e-07, | |
| "logits/chosen": -1.9113174676895142, | |
| "logits/rejected": -1.6106466054916382, | |
| "logps/chosen": -88.4524917602539, | |
| "logps/rejected": -71.1933364868164, | |
| "loss": 0.5805, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.1571049988269806, | |
| "rewards/margins": 0.23198041319847107, | |
| "rewards/rejected": -0.07487543672323227, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 1.21765601217656e-07, | |
| "logits/chosen": -1.9750347137451172, | |
| "logits/rejected": -1.6996724605560303, | |
| "logps/chosen": -91.44354248046875, | |
| "logps/rejected": -77.03483581542969, | |
| "loss": 0.5768, | |
| "rewards/accuracies": 0.9416666030883789, | |
| "rewards/chosen": 0.14498132467269897, | |
| "rewards/margins": 0.23422233760356903, | |
| "rewards/rejected": -0.08924100548028946, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.1796042617960425e-07, | |
| "logits/chosen": -1.9331929683685303, | |
| "logits/rejected": -1.6460412740707397, | |
| "logps/chosen": -84.95231628417969, | |
| "logps/rejected": -73.00121307373047, | |
| "loss": 0.5759, | |
| "rewards/accuracies": 0.9416667222976685, | |
| "rewards/chosen": 0.17358574271202087, | |
| "rewards/margins": 0.2643177807331085, | |
| "rewards/rejected": -0.09073203802108765, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.141552511415525e-07, | |
| "logits/chosen": -1.9368797540664673, | |
| "logits/rejected": -1.6109968423843384, | |
| "logps/chosen": -95.30493927001953, | |
| "logps/rejected": -79.38502502441406, | |
| "loss": 0.5725, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.16715973615646362, | |
| "rewards/margins": 0.26029014587402344, | |
| "rewards/rejected": -0.09313040971755981, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.1035007610350075e-07, | |
| "logits/chosen": -1.9934200048446655, | |
| "logits/rejected": -1.7067596912384033, | |
| "logps/chosen": -91.30809020996094, | |
| "logps/rejected": -75.1029052734375, | |
| "loss": 0.577, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.1644877940416336, | |
| "rewards/margins": 0.2513524889945984, | |
| "rewards/rejected": -0.08686470985412598, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.06544901065449e-07, | |
| "logits/chosen": -1.8585717678070068, | |
| "logits/rejected": -1.593147873878479, | |
| "logps/chosen": -94.06452941894531, | |
| "logps/rejected": -78.18238830566406, | |
| "loss": 0.5798, | |
| "rewards/accuracies": 0.9166666865348816, | |
| "rewards/chosen": 0.1385236382484436, | |
| "rewards/margins": 0.2384444922208786, | |
| "rewards/rejected": -0.09992088377475739, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.0273972602739725e-07, | |
| "logits/chosen": -1.921449065208435, | |
| "logits/rejected": -1.6535272598266602, | |
| "logps/chosen": -87.8677749633789, | |
| "logps/rejected": -78.9840087890625, | |
| "loss": 0.5767, | |
| "rewards/accuracies": 0.9666666984558105, | |
| "rewards/chosen": 0.14824248850345612, | |
| "rewards/margins": 0.25046244263648987, | |
| "rewards/rejected": -0.10221991688013077, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 9.89345509893455e-08, | |
| "logits/chosen": -1.8294260501861572, | |
| "logits/rejected": -1.5607296228408813, | |
| "logps/chosen": -94.44920349121094, | |
| "logps/rejected": -77.64144134521484, | |
| "loss": 0.5789, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.16983038187026978, | |
| "rewards/margins": 0.2730123996734619, | |
| "rewards/rejected": -0.10318204015493393, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_logits/chosen": -2.0464720726013184, | |
| "eval_logits/rejected": -1.7763383388519287, | |
| "eval_logps/chosen": -90.46904754638672, | |
| "eval_logps/rejected": -73.09234619140625, | |
| "eval_loss": 0.5758996605873108, | |
| "eval_rewards/accuracies": 0.9333333373069763, | |
| "eval_rewards/chosen": 0.1516665369272232, | |
| "eval_rewards/margins": 0.2508557140827179, | |
| "eval_rewards/rejected": -0.0991891473531723, | |
| "eval_runtime": 117.7184, | |
| "eval_samples_per_second": 24.312, | |
| "eval_steps_per_second": 0.765, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 9.512937595129374e-08, | |
| "logits/chosen": -1.8505995273590088, | |
| "logits/rejected": -1.570237398147583, | |
| "logps/chosen": -95.08009338378906, | |
| "logps/rejected": -75.39155578613281, | |
| "loss": 0.5739, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.14622533321380615, | |
| "rewards/margins": 0.23580579459667206, | |
| "rewards/rejected": -0.08958044648170471, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 9.1324200913242e-08, | |
| "logits/chosen": -1.8380448818206787, | |
| "logits/rejected": -1.5534820556640625, | |
| "logps/chosen": -89.47401428222656, | |
| "logps/rejected": -73.60565948486328, | |
| "loss": 0.5772, | |
| "rewards/accuracies": 0.9416667222976685, | |
| "rewards/chosen": 0.1780690848827362, | |
| "rewards/margins": 0.2615908980369568, | |
| "rewards/rejected": -0.08352181315422058, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.751902587519024e-08, | |
| "logits/chosen": -1.975015640258789, | |
| "logits/rejected": -1.6878995895385742, | |
| "logps/chosen": -93.83460998535156, | |
| "logps/rejected": -76.86270904541016, | |
| "loss": 0.576, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.17377497255802155, | |
| "rewards/margins": 0.25830045342445374, | |
| "rewards/rejected": -0.08452550321817398, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.37138508371385e-08, | |
| "logits/chosen": -1.8430579900741577, | |
| "logits/rejected": -1.5395987033843994, | |
| "logps/chosen": -92.52012634277344, | |
| "logps/rejected": -79.96121215820312, | |
| "loss": 0.5745, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.15905624628067017, | |
| "rewards/margins": 0.25326135754585266, | |
| "rewards/rejected": -0.0942051112651825, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 7.990867579908676e-08, | |
| "logits/chosen": -1.8872146606445312, | |
| "logits/rejected": -1.5909473896026611, | |
| "logps/chosen": -89.97090148925781, | |
| "logps/rejected": -77.28529357910156, | |
| "loss": 0.5727, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.16229455173015594, | |
| "rewards/margins": 0.2598329186439514, | |
| "rewards/rejected": -0.09753839671611786, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 7.6103500761035e-08, | |
| "logits/chosen": -1.8499984741210938, | |
| "logits/rejected": -1.5525275468826294, | |
| "logps/chosen": -93.28520202636719, | |
| "logps/rejected": -74.24462890625, | |
| "loss": 0.5731, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.14929968118667603, | |
| "rewards/margins": 0.2529754042625427, | |
| "rewards/rejected": -0.10367570072412491, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 7.229832572298326e-08, | |
| "logits/chosen": -2.009666681289673, | |
| "logits/rejected": -1.7177025079727173, | |
| "logps/chosen": -89.4050521850586, | |
| "logps/rejected": -79.06332397460938, | |
| "loss": 0.5712, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.1704416573047638, | |
| "rewards/margins": 0.27492305636405945, | |
| "rewards/rejected": -0.10448137670755386, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.84931506849315e-08, | |
| "logits/chosen": -1.81943678855896, | |
| "logits/rejected": -1.5563119649887085, | |
| "logps/chosen": -88.48675537109375, | |
| "logps/rejected": -72.3918685913086, | |
| "loss": 0.5729, | |
| "rewards/accuracies": 0.8916667103767395, | |
| "rewards/chosen": 0.13842932879924774, | |
| "rewards/margins": 0.24020667374134064, | |
| "rewards/rejected": -0.10177735984325409, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.468797564687976e-08, | |
| "logits/chosen": -1.8918097019195557, | |
| "logits/rejected": -1.6081184148788452, | |
| "logps/chosen": -87.69499969482422, | |
| "logps/rejected": -71.82918548583984, | |
| "loss": 0.5745, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.17439430952072144, | |
| "rewards/margins": 0.2683844268321991, | |
| "rewards/rejected": -0.09399012476205826, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 6.0882800608828e-08, | |
| "logits/chosen": -1.7783477306365967, | |
| "logits/rejected": -1.4948413372039795, | |
| "logps/chosen": -89.221435546875, | |
| "logps/rejected": -72.36725616455078, | |
| "loss": 0.5689, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.16254201531410217, | |
| "rewards/margins": 0.27369171380996704, | |
| "rewards/rejected": -0.11114968359470367, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_logits/chosen": -2.0464510917663574, | |
| "eval_logits/rejected": -1.7761657238006592, | |
| "eval_logps/chosen": -90.43045043945312, | |
| "eval_logps/rejected": -73.13316345214844, | |
| "eval_loss": 0.5722280740737915, | |
| "eval_rewards/accuracies": 0.949999988079071, | |
| "eval_rewards/chosen": 0.15552671253681183, | |
| "eval_rewards/margins": 0.2587975263595581, | |
| "eval_rewards/rejected": -0.10327085852622986, | |
| "eval_runtime": 118.3212, | |
| "eval_samples_per_second": 24.188, | |
| "eval_steps_per_second": 0.761, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5.707762557077625e-08, | |
| "logits/chosen": -1.7794984579086304, | |
| "logits/rejected": -1.5128498077392578, | |
| "logps/chosen": -94.00843048095703, | |
| "logps/rejected": -75.231201171875, | |
| "loss": 0.5719, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.18278029561042786, | |
| "rewards/margins": 0.2786504328250885, | |
| "rewards/rejected": -0.09587012976408005, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5.32724505327245e-08, | |
| "logits/chosen": -1.9731611013412476, | |
| "logits/rejected": -1.6866945028305054, | |
| "logps/chosen": -93.30955505371094, | |
| "logps/rejected": -78.57191467285156, | |
| "loss": 0.571, | |
| "rewards/accuracies": 0.9583333134651184, | |
| "rewards/chosen": 0.15462180972099304, | |
| "rewards/margins": 0.26251837611198425, | |
| "rewards/rejected": -0.10789655148983002, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.946727549467275e-08, | |
| "logits/chosen": -1.9133844375610352, | |
| "logits/rejected": -1.636694312095642, | |
| "logps/chosen": -88.56840515136719, | |
| "logps/rejected": -76.69215393066406, | |
| "loss": 0.5716, | |
| "rewards/accuracies": 0.966666579246521, | |
| "rewards/chosen": 0.15411558747291565, | |
| "rewards/margins": 0.26415151357650757, | |
| "rewards/rejected": -0.11003589630126953, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.5662100456621e-08, | |
| "logits/chosen": -1.8889293670654297, | |
| "logits/rejected": -1.5668977499008179, | |
| "logps/chosen": -93.04912567138672, | |
| "logps/rejected": -74.54277038574219, | |
| "loss": 0.5671, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.1801871955394745, | |
| "rewards/margins": 0.2698620557785034, | |
| "rewards/rejected": -0.08967487514019012, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.185692541856925e-08, | |
| "logits/chosen": -1.853014588356018, | |
| "logits/rejected": -1.6049798727035522, | |
| "logps/chosen": -92.84537506103516, | |
| "logps/rejected": -76.91346740722656, | |
| "loss": 0.5734, | |
| "rewards/accuracies": 0.9333332777023315, | |
| "rewards/chosen": 0.1830371469259262, | |
| "rewards/margins": 0.27097687125205994, | |
| "rewards/rejected": -0.08793972432613373, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 3.80517503805175e-08, | |
| "logits/chosen": -1.8321574926376343, | |
| "logits/rejected": -1.559768795967102, | |
| "logps/chosen": -90.3908462524414, | |
| "logps/rejected": -77.5791015625, | |
| "loss": 0.5728, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.15050294995307922, | |
| "rewards/margins": 0.2644408941268921, | |
| "rewards/rejected": -0.11393795162439346, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.424657534246575e-08, | |
| "logits/chosen": -1.8819458484649658, | |
| "logits/rejected": -1.5830302238464355, | |
| "logps/chosen": -96.20283508300781, | |
| "logps/rejected": -79.02750396728516, | |
| "loss": 0.57, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.1495673507452011, | |
| "rewards/margins": 0.26126623153686523, | |
| "rewards/rejected": -0.11169885098934174, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.0441400304414e-08, | |
| "logits/chosen": -1.8712704181671143, | |
| "logits/rejected": -1.5640711784362793, | |
| "logps/chosen": -87.56044006347656, | |
| "logps/rejected": -73.2048110961914, | |
| "loss": 0.5701, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.15671199560165405, | |
| "rewards/margins": 0.2683314383029938, | |
| "rewards/rejected": -0.11161943525075912, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.663622526636225e-08, | |
| "logits/chosen": -1.9705963134765625, | |
| "logits/rejected": -1.6459972858428955, | |
| "logps/chosen": -93.333251953125, | |
| "logps/rejected": -75.71430969238281, | |
| "loss": 0.5669, | |
| "rewards/accuracies": 0.9833332896232605, | |
| "rewards/chosen": 0.22007617354393005, | |
| "rewards/margins": 0.3152759373188019, | |
| "rewards/rejected": -0.09519973397254944, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.28310502283105e-08, | |
| "logits/chosen": -1.9400713443756104, | |
| "logits/rejected": -1.6670045852661133, | |
| "logps/chosen": -90.7359619140625, | |
| "logps/rejected": -72.49601745605469, | |
| "loss": 0.5694, | |
| "rewards/accuracies": 0.9416667222976685, | |
| "rewards/chosen": 0.18735943734645844, | |
| "rewards/margins": 0.2668268084526062, | |
| "rewards/rejected": -0.07946738600730896, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_logits/chosen": -2.0464956760406494, | |
| "eval_logits/rejected": -1.776126742362976, | |
| "eval_logps/chosen": -90.40695190429688, | |
| "eval_logps/rejected": -73.16618347167969, | |
| "eval_loss": 0.5701765418052673, | |
| "eval_rewards/accuracies": 0.9416666626930237, | |
| "eval_rewards/chosen": 0.15787601470947266, | |
| "eval_rewards/margins": 0.26444879174232483, | |
| "eval_rewards/rejected": -0.10657278448343277, | |
| "eval_runtime": 118.3051, | |
| "eval_samples_per_second": 24.192, | |
| "eval_steps_per_second": 0.761, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.902587519025875e-08, | |
| "logits/chosen": -1.9084300994873047, | |
| "logits/rejected": -1.5941137075424194, | |
| "logps/chosen": -91.3935546875, | |
| "logps/rejected": -76.21849060058594, | |
| "loss": 0.5703, | |
| "rewards/accuracies": 0.9416666030883789, | |
| "rewards/chosen": 0.18224991858005524, | |
| "rewards/margins": 0.2809165418148041, | |
| "rewards/rejected": -0.09866663068532944, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5220700152207e-08, | |
| "logits/chosen": -1.8584785461425781, | |
| "logits/rejected": -1.5252989530563354, | |
| "logps/chosen": -94.4938735961914, | |
| "logps/rejected": -76.2500228881836, | |
| "loss": 0.5705, | |
| "rewards/accuracies": 0.9166666269302368, | |
| "rewards/chosen": 0.17414768040180206, | |
| "rewards/margins": 0.2775163948535919, | |
| "rewards/rejected": -0.10336872190237045, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.141552511415525e-08, | |
| "logits/chosen": -1.8703104257583618, | |
| "logits/rejected": -1.5645638704299927, | |
| "logps/chosen": -92.91007232666016, | |
| "logps/rejected": -72.98766326904297, | |
| "loss": 0.5703, | |
| "rewards/accuracies": 0.9416666030883789, | |
| "rewards/chosen": 0.1589893251657486, | |
| "rewards/margins": 0.26451554894447327, | |
| "rewards/rejected": -0.10552623122930527, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 7.6103500761035e-09, | |
| "logits/chosen": -1.9101779460906982, | |
| "logits/rejected": -1.6234633922576904, | |
| "logps/chosen": -88.99760437011719, | |
| "logps/rejected": -73.46092224121094, | |
| "loss": 0.5694, | |
| "rewards/accuracies": 0.9333333969116211, | |
| "rewards/chosen": 0.17230169475078583, | |
| "rewards/margins": 0.25403517484664917, | |
| "rewards/rejected": -0.08173345029354095, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.80517503805175e-09, | |
| "logits/chosen": -1.837728500366211, | |
| "logits/rejected": -1.6049699783325195, | |
| "logps/chosen": -92.38600158691406, | |
| "logps/rejected": -78.56689453125, | |
| "loss": 0.571, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.18492794036865234, | |
| "rewards/margins": 0.28263044357299805, | |
| "rewards/rejected": -0.0977025032043457, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -1.8631808757781982, | |
| "logits/rejected": -1.5711183547973633, | |
| "logps/chosen": -90.29685974121094, | |
| "logps/rejected": -76.57972717285156, | |
| "loss": 0.5684, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.18871954083442688, | |
| "rewards/margins": 0.3021387457847595, | |
| "rewards/rejected": -0.11341919749975204, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1460, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6280729855576607, | |
| "train_runtime": 9689.6427, | |
| "train_samples_per_second": 14.469, | |
| "train_steps_per_second": 0.151 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1460, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |