| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999345977763244, | |
| "eval_steps": 100, | |
| "global_step": 764, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 6.493506493506494e-09, | |
| "logits/chosen": -2.0615594387054443, | |
| "logits/rejected": -1.9222214221954346, | |
| "logps/chosen": -3380.6083984375, | |
| "logps/rejected": -2521.2978515625, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 6.493506493506492e-08, | |
| "logits/chosen": -1.674426555633545, | |
| "logits/rejected": -1.637134313583374, | |
| "logps/chosen": -2549.3515625, | |
| "logps/rejected": -2319.4013671875, | |
| "loss": 10.0505, | |
| "rewards/accuracies": 0.4333333373069763, | |
| "rewards/chosen": 0.0008169158827513456, | |
| "rewards/margins": 0.0011402772506698966, | |
| "rewards/rejected": -0.00032336192089132965, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.2987012987012984e-07, | |
| "logits/chosen": -1.6043205261230469, | |
| "logits/rejected": -1.5535523891448975, | |
| "logps/chosen": -2340.101318359375, | |
| "logps/rejected": -2224.145263671875, | |
| "loss": 7.4843, | |
| "rewards/accuracies": 0.5200000405311584, | |
| "rewards/chosen": 0.00018432810611557215, | |
| "rewards/margins": 0.0009077669237740338, | |
| "rewards/rejected": -0.0007234388613142073, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.948051948051948e-07, | |
| "logits/chosen": -1.6847426891326904, | |
| "logits/rejected": -1.6577625274658203, | |
| "logps/chosen": -2983.23681640625, | |
| "logps/rejected": -2513.237060546875, | |
| "loss": 9.1379, | |
| "rewards/accuracies": 0.48000001907348633, | |
| "rewards/chosen": 0.010261936113238335, | |
| "rewards/margins": 0.004135974682867527, | |
| "rewards/rejected": 0.006125961430370808, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.597402597402597e-07, | |
| "logits/chosen": -1.859400987625122, | |
| "logits/rejected": -1.8100417852401733, | |
| "logps/chosen": -2879.57470703125, | |
| "logps/rejected": -2273.878173828125, | |
| "loss": 12.271, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.012033696286380291, | |
| "rewards/margins": 0.005555520299822092, | |
| "rewards/rejected": 0.006478174589574337, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.2467532467532465e-07, | |
| "logits/chosen": -1.828608751296997, | |
| "logits/rejected": -1.805625319480896, | |
| "logps/chosen": -2893.784423828125, | |
| "logps/rejected": -2551.77294921875, | |
| "loss": 8.7411, | |
| "rewards/accuracies": 0.5600000023841858, | |
| "rewards/chosen": 0.02166888490319252, | |
| "rewards/margins": 0.007775471545755863, | |
| "rewards/rejected": 0.013893413357436657, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.896103896103896e-07, | |
| "logits/chosen": -1.7459495067596436, | |
| "logits/rejected": -1.6628999710083008, | |
| "logps/chosen": -3231.689697265625, | |
| "logps/rejected": -2554.42919921875, | |
| "loss": 9.758, | |
| "rewards/accuracies": 0.559999942779541, | |
| "rewards/chosen": 0.027519574388861656, | |
| "rewards/margins": 0.008895651437342167, | |
| "rewards/rejected": 0.018623923882842064, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.545454545454545e-07, | |
| "logits/chosen": -1.8072433471679688, | |
| "logits/rejected": -1.7838470935821533, | |
| "logps/chosen": -2829.386474609375, | |
| "logps/rejected": -2542.68701171875, | |
| "loss": 11.0017, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.024034958332777023, | |
| "rewards/margins": 0.006175906863063574, | |
| "rewards/rejected": 0.017859051004052162, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.99976474872689e-07, | |
| "logits/chosen": -1.7730411291122437, | |
| "logits/rejected": -1.7399647235870361, | |
| "logps/chosen": -2769.705322265625, | |
| "logps/rejected": -2476.75634765625, | |
| "loss": 15.623, | |
| "rewards/accuracies": 0.5400000214576721, | |
| "rewards/chosen": 0.008623984642326832, | |
| "rewards/margins": 0.008157819509506226, | |
| "rewards/rejected": 0.0004661638231482357, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.995583735427465e-07, | |
| "logits/chosen": -1.790204644203186, | |
| "logits/rejected": -1.7226215600967407, | |
| "logps/chosen": -2688.0732421875, | |
| "logps/rejected": -2436.649658203125, | |
| "loss": 11.9811, | |
| "rewards/accuracies": 0.6100000143051147, | |
| "rewards/chosen": 0.017978714779019356, | |
| "rewards/margins": 0.017238261178135872, | |
| "rewards/rejected": 0.0007404519128613174, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.986184978516146e-07, | |
| "logits/chosen": -1.7211675643920898, | |
| "logits/rejected": -1.6991230249404907, | |
| "logps/chosen": -2611.177001953125, | |
| "logps/rejected": -2212.4033203125, | |
| "loss": 16.8403, | |
| "rewards/accuracies": 0.5200001001358032, | |
| "rewards/chosen": 0.024822045117616653, | |
| "rewards/margins": 0.00336282467469573, | |
| "rewards/rejected": 0.021459218114614487, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_logits/chosen": -1.6729556322097778, | |
| "eval_logits/rejected": -1.6068017482757568, | |
| "eval_logps/chosen": -2806.55517578125, | |
| "eval_logps/rejected": -2491.901123046875, | |
| "eval_loss": 19.51178741455078, | |
| "eval_rewards/accuracies": 0.52734375, | |
| "eval_rewards/chosen": 0.025559017434716225, | |
| "eval_rewards/margins": 0.008243386633694172, | |
| "eval_rewards/rejected": 0.017315629869699478, | |
| "eval_runtime": 115.2508, | |
| "eval_samples_per_second": 17.353, | |
| "eval_steps_per_second": 0.278, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.971588128827782e-07, | |
| "logits/chosen": -1.7473026514053345, | |
| "logits/rejected": -1.6806236505508423, | |
| "logps/chosen": -3125.757080078125, | |
| "logps/rejected": -2645.337158203125, | |
| "loss": 26.9149, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.016931097954511642, | |
| "rewards/margins": 0.002864243695512414, | |
| "rewards/rejected": 0.014066850766539574, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.951823705321981e-07, | |
| "logits/chosen": -1.7069530487060547, | |
| "logits/rejected": -1.6579583883285522, | |
| "logps/chosen": -2828.78662109375, | |
| "logps/rejected": -2442.76416015625, | |
| "loss": 33.872, | |
| "rewards/accuracies": 0.5600000023841858, | |
| "rewards/chosen": 0.013961514458060265, | |
| "rewards/margins": 0.00896529946476221, | |
| "rewards/rejected": 0.004996216390281916, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.926933031274343e-07, | |
| "logits/chosen": -1.7224699258804321, | |
| "logits/rejected": -1.6934731006622314, | |
| "logps/chosen": -2923.9306640625, | |
| "logps/rejected": -2566.210693359375, | |
| "loss": 39.0757, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.038237668573856354, | |
| "rewards/margins": 0.006029448006302118, | |
| "rewards/rejected": 0.03220822289586067, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.896968147878145e-07, | |
| "logits/chosen": -1.7280409336090088, | |
| "logits/rejected": -1.7070726156234741, | |
| "logps/chosen": -2737.75927734375, | |
| "logps/rejected": -2486.45751953125, | |
| "loss": 18.5231, | |
| "rewards/accuracies": 0.6299999952316284, | |
| "rewards/chosen": 0.031596291810274124, | |
| "rewards/margins": 0.04216960817575455, | |
| "rewards/rejected": -0.010573318228125572, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.861991705437081e-07, | |
| "logits/chosen": -1.7859830856323242, | |
| "logits/rejected": -1.7191492319107056, | |
| "logps/chosen": -2743.43310546875, | |
| "logps/rejected": -2297.162109375, | |
| "loss": 20.7835, | |
| "rewards/accuracies": 0.5800000429153442, | |
| "rewards/chosen": 0.0336376316845417, | |
| "rewards/margins": 0.011832155287265778, | |
| "rewards/rejected": 0.021805476397275925, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.822076832376586e-07, | |
| "logits/chosen": -1.8132251501083374, | |
| "logits/rejected": -1.7665789127349854, | |
| "logps/chosen": -2841.165771484375, | |
| "logps/rejected": -2748.486572265625, | |
| "loss": 57.9401, | |
| "rewards/accuracies": 0.5099999904632568, | |
| "rewards/chosen": 0.006118610501289368, | |
| "rewards/margins": 0.0022155127953737974, | |
| "rewards/rejected": 0.0039030970074236393, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.777306982347594e-07, | |
| "logits/chosen": -1.6557657718658447, | |
| "logits/rejected": -1.5996118783950806, | |
| "logps/chosen": -3055.95361328125, | |
| "logps/rejected": -2603.83642578125, | |
| "loss": 23.1296, | |
| "rewards/accuracies": 0.6200000047683716, | |
| "rewards/chosen": 0.028251701965928078, | |
| "rewards/margins": 0.020935241132974625, | |
| "rewards/rejected": 0.007316464092582464, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.7277757597424075e-07, | |
| "logits/chosen": -1.8335905075073242, | |
| "logits/rejected": -1.7595329284667969, | |
| "logps/chosen": -2963.73779296875, | |
| "logps/rejected": -2540.163818359375, | |
| "loss": 40.5046, | |
| "rewards/accuracies": 0.5400000810623169, | |
| "rewards/chosen": 0.018591446802020073, | |
| "rewards/margins": -0.0024230503477156162, | |
| "rewards/rejected": 0.02101449854671955, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.6735867239874904e-07, | |
| "logits/chosen": -1.8637840747833252, | |
| "logits/rejected": -1.7640159130096436, | |
| "logps/chosen": -3237.434814453125, | |
| "logps/rejected": -2429.197998046875, | |
| "loss": 36.3042, | |
| "rewards/accuracies": 0.6200000047683716, | |
| "rewards/chosen": 0.04794805496931076, | |
| "rewards/margins": 0.019117821007966995, | |
| "rewards/rejected": 0.028830235823988914, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.6148531730223733e-07, | |
| "logits/chosen": -1.6909841299057007, | |
| "logits/rejected": -1.6915366649627686, | |
| "logps/chosen": -2649.89404296875, | |
| "logps/rejected": -2436.87353515625, | |
| "loss": 28.1241, | |
| "rewards/accuracies": 0.5300000309944153, | |
| "rewards/chosen": 0.007661645300686359, | |
| "rewards/margins": 0.0055509163066744804, | |
| "rewards/rejected": 0.0021107294596731663, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_logits/chosen": -1.681164264678955, | |
| "eval_logits/rejected": -1.618328332901001, | |
| "eval_logps/chosen": -2808.258056640625, | |
| "eval_logps/rejected": -2494.01953125, | |
| "eval_loss": 32.517486572265625, | |
| "eval_rewards/accuracies": 0.5234375, | |
| "eval_rewards/chosen": 0.008527392521500587, | |
| "eval_rewards/margins": 0.012391308322548866, | |
| "eval_rewards/rejected": -0.0038639232516288757, | |
| "eval_runtime": 113.682, | |
| "eval_samples_per_second": 17.593, | |
| "eval_steps_per_second": 0.281, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5516979064173524e-07, | |
| "logits/chosen": -1.749903917312622, | |
| "logits/rejected": -1.7615283727645874, | |
| "logps/chosen": -2285.7451171875, | |
| "logps/rejected": -2269.229736328125, | |
| "loss": 25.9535, | |
| "rewards/accuracies": 0.6100000143051147, | |
| "rewards/chosen": 0.011981850489974022, | |
| "rewards/margins": 0.014764687046408653, | |
| "rewards/rejected": -0.0027828349266201258, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.484252968625277e-07, | |
| "logits/chosen": -1.716509222984314, | |
| "logits/rejected": -1.6396989822387695, | |
| "logps/chosen": -2435.95556640625, | |
| "logps/rejected": -1922.770751953125, | |
| "loss": 28.3739, | |
| "rewards/accuracies": 0.6200000047683716, | |
| "rewards/chosen": 0.004359879065304995, | |
| "rewards/margins": 0.007711753249168396, | |
| "rewards/rejected": -0.0033518739510327578, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.4126593729042193e-07, | |
| "logits/chosen": -1.799469232559204, | |
| "logits/rejected": -1.757004737854004, | |
| "logps/chosen": -3254.6396484375, | |
| "logps/rejected": -2515.59619140625, | |
| "loss": 39.4707, | |
| "rewards/accuracies": 0.5900000333786011, | |
| "rewards/chosen": 0.03561704605817795, | |
| "rewards/margins": 0.019830647855997086, | |
| "rewards/rejected": 0.015786398202180862, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.3370668064882397e-07, | |
| "logits/chosen": -1.7325947284698486, | |
| "logits/rejected": -1.7474550008773804, | |
| "logps/chosen": -2579.47412109375, | |
| "logps/rejected": -2328.500732421875, | |
| "loss": 44.2727, | |
| "rewards/accuracies": 0.5100000500679016, | |
| "rewards/chosen": 0.04269097000360489, | |
| "rewards/margins": 0.02060030959546566, | |
| "rewards/rejected": 0.02209065482020378, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.2576333176226944e-07, | |
| "logits/chosen": -1.7366650104522705, | |
| "logits/rejected": -1.706789255142212, | |
| "logps/chosen": -2479.5576171875, | |
| "logps/rejected": -2277.726318359375, | |
| "loss": 29.5758, | |
| "rewards/accuracies": 0.5300000309944153, | |
| "rewards/chosen": 0.1058274507522583, | |
| "rewards/margins": 0.013660475611686707, | |
| "rewards/rejected": 0.0921669602394104, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.17452498511841e-07, | |
| "logits/chosen": -1.7807962894439697, | |
| "logits/rejected": -1.7134149074554443, | |
| "logps/chosen": -2989.12841796875, | |
| "logps/rejected": -2354.25830078125, | |
| "loss": 38.7316, | |
| "rewards/accuracies": 0.5200000405311584, | |
| "rewards/chosen": 0.023859605193138123, | |
| "rewards/margins": 0.005521018523722887, | |
| "rewards/rejected": 0.018338587135076523, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.087915571115629e-07, | |
| "logits/chosen": -1.8165556192398071, | |
| "logits/rejected": -1.7687098979949951, | |
| "logps/chosen": -2833.55859375, | |
| "logps/rejected": -2183.32470703125, | |
| "loss": 330.4642, | |
| "rewards/accuracies": 0.5699999928474426, | |
| "rewards/chosen": 0.031318746507167816, | |
| "rewards/margins": 0.029996121302247047, | |
| "rewards/rejected": 0.0013226259034126997, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.997986157783715e-07, | |
| "logits/chosen": -1.6980018615722656, | |
| "logits/rejected": -1.589050531387329, | |
| "logps/chosen": -3510.792236328125, | |
| "logps/rejected": -2689.208984375, | |
| "loss": 58.1646, | |
| "rewards/accuracies": 0.5200000405311584, | |
| "rewards/chosen": 0.014776378870010376, | |
| "rewards/margins": 0.011294273659586906, | |
| "rewards/rejected": 0.0034821047447621822, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.9049247687162155e-07, | |
| "logits/chosen": -1.7791646718978882, | |
| "logits/rejected": -1.7399044036865234, | |
| "logps/chosen": -2478.590576171875, | |
| "logps/rejected": -2269.01416015625, | |
| "loss": 31.6725, | |
| "rewards/accuracies": 0.5699999928474426, | |
| "rewards/chosen": 0.04884537309408188, | |
| "rewards/margins": 0.0339895561337471, | |
| "rewards/rejected": 0.014855814166367054, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8089259758128543e-07, | |
| "logits/chosen": -1.670789361000061, | |
| "logits/rejected": -1.6030629873275757, | |
| "logps/chosen": -2726.465576171875, | |
| "logps/rejected": -2119.26123046875, | |
| "loss": 84.7591, | |
| "rewards/accuracies": 0.5699999928474426, | |
| "rewards/chosen": 0.021672677248716354, | |
| "rewards/margins": -0.010785548016428947, | |
| "rewards/rejected": 0.03245822712779045, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_logits/chosen": -1.660080909729004, | |
| "eval_logits/rejected": -1.596778154373169, | |
| "eval_logps/chosen": -2806.140625, | |
| "eval_logps/rejected": -2492.270263671875, | |
| "eval_loss": 47.80431365966797, | |
| "eval_rewards/accuracies": 0.5390625, | |
| "eval_rewards/chosen": 0.029702020809054375, | |
| "eval_rewards/margins": 0.01607733778655529, | |
| "eval_rewards/rejected": 0.013624681159853935, | |
| "eval_runtime": 116.3019, | |
| "eval_samples_per_second": 17.197, | |
| "eval_steps_per_second": 0.275, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.710190492470386e-07, | |
| "logits/chosen": -1.6620228290557861, | |
| "logits/rejected": -1.7311099767684937, | |
| "logps/chosen": -2315.977294921875, | |
| "logps/rejected": -2199.08251953125, | |
| "loss": 43.6013, | |
| "rewards/accuracies": 0.5400000214576721, | |
| "rewards/chosen": 0.032384876161813736, | |
| "rewards/margins": 0.008921505883336067, | |
| "rewards/rejected": 0.02346337027847767, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.6089247539328616e-07, | |
| "logits/chosen": -1.7675050497055054, | |
| "logits/rejected": -1.7156997919082642, | |
| "logps/chosen": -2859.810791015625, | |
| "logps/rejected": -2569.75537109375, | |
| "loss": 38.8904, | |
| "rewards/accuracies": 0.559999942779541, | |
| "rewards/chosen": 0.020630866289138794, | |
| "rewards/margins": 0.0018306337296962738, | |
| "rewards/rejected": 0.01880022883415222, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5053404856787166e-07, | |
| "logits/chosen": -1.6446609497070312, | |
| "logits/rejected": -1.5918724536895752, | |
| "logps/chosen": -3104.72802734375, | |
| "logps/rejected": -2430.239013671875, | |
| "loss": 84.9753, | |
| "rewards/accuracies": 0.47999995946884155, | |
| "rewards/chosen": 0.053660690784454346, | |
| "rewards/margins": -0.005831834394484758, | |
| "rewards/rejected": 0.05949252098798752, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.399654260747078e-07, | |
| "logits/chosen": -1.699196219444275, | |
| "logits/rejected": -1.7045748233795166, | |
| "logps/chosen": -2584.699462890625, | |
| "logps/rejected": -2263.678466796875, | |
| "loss": 38.1532, | |
| "rewards/accuracies": 0.5300000309944153, | |
| "rewards/chosen": 0.02709970250725746, | |
| "rewards/margins": 0.01412280835211277, | |
| "rewards/rejected": 0.012976895086467266, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2920870469288373e-07, | |
| "logits/chosen": -1.7267532348632812, | |
| "logits/rejected": -1.6659395694732666, | |
| "logps/chosen": -2935.341796875, | |
| "logps/rejected": -2503.583984375, | |
| "loss": 47.1836, | |
| "rewards/accuracies": 0.5199999809265137, | |
| "rewards/chosen": 0.031215447932481766, | |
| "rewards/margins": 0.022346725687384605, | |
| "rewards/rejected": 0.008868719451129436, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.182863744769218e-07, | |
| "logits/chosen": -1.7288787364959717, | |
| "logits/rejected": -1.6928844451904297, | |
| "logps/chosen": -2811.489501953125, | |
| "logps/rejected": -2596.68310546875, | |
| "loss": 36.8176, | |
| "rewards/accuracies": 0.5099999904632568, | |
| "rewards/chosen": 0.1375296413898468, | |
| "rewards/margins": 0.0820910781621933, | |
| "rewards/rejected": 0.05543852597475052, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.072212717347776e-07, | |
| "logits/chosen": -1.7680120468139648, | |
| "logits/rejected": -1.6781940460205078, | |
| "logps/chosen": -3101.98583984375, | |
| "logps/rejected": -2426.4716796875, | |
| "loss": 36.7837, | |
| "rewards/accuracies": 0.5199999809265137, | |
| "rewards/chosen": 0.022122934460639954, | |
| "rewards/margins": 0.011781491339206696, | |
| "rewards/rejected": 0.010341441258788109, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9603653128189665e-07, | |
| "logits/chosen": -1.6812299489974976, | |
| "logits/rejected": -1.7215496301651, | |
| "logps/chosen": -2823.8291015625, | |
| "logps/rejected": -2762.53076171875, | |
| "loss": 42.732, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.028385426849126816, | |
| "rewards/margins": -0.006515379063785076, | |
| "rewards/rejected": 0.03490080684423447, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8475553807115387e-07, | |
| "logits/chosen": -1.8070951700210571, | |
| "logits/rejected": -1.7426990270614624, | |
| "logps/chosen": -2697.833251953125, | |
| "logps/rejected": -2263.9990234375, | |
| "loss": 55.8683, | |
| "rewards/accuracies": 0.5099999904632568, | |
| "rewards/chosen": 0.012206131592392921, | |
| "rewards/margins": 0.011461116373538971, | |
| "rewards/rejected": 0.0007450145785696805, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7340187829980883e-07, | |
| "logits/chosen": -1.8249183893203735, | |
| "logits/rejected": -1.7130759954452515, | |
| "logps/chosen": -2940.11181640625, | |
| "logps/rejected": -2463.068359375, | |
| "loss": 40.7835, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.0059137181378901005, | |
| "rewards/margins": 0.01795141212642193, | |
| "rewards/rejected": -0.012037692591547966, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_logits/chosen": -1.6917269229888916, | |
| "eval_logits/rejected": -1.628839373588562, | |
| "eval_logps/chosen": -2807.42626953125, | |
| "eval_logps/rejected": -2493.92041015625, | |
| "eval_loss": 30.672218322753906, | |
| "eval_rewards/accuracies": 0.5546875, | |
| "eval_rewards/chosen": 0.016848012804985046, | |
| "eval_rewards/margins": 0.019721925258636475, | |
| "eval_rewards/rejected": -0.0028739143162965775, | |
| "eval_runtime": 110.0303, | |
| "eval_samples_per_second": 18.177, | |
| "eval_steps_per_second": 0.291, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6199929009569996e-07, | |
| "logits/chosen": -1.7034717798233032, | |
| "logits/rejected": -1.707564353942871, | |
| "logps/chosen": -2599.38330078125, | |
| "logps/rejected": -2273.864990234375, | |
| "loss": 43.9981, | |
| "rewards/accuracies": 0.5600000023841858, | |
| "rewards/chosen": 0.02160579524934292, | |
| "rewards/margins": 0.0038177832029759884, | |
| "rewards/rejected": 0.017788011580705643, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5057161388578505e-07, | |
| "logits/chosen": -1.7964134216308594, | |
| "logits/rejected": -1.730661392211914, | |
| "logps/chosen": -3038.08740234375, | |
| "logps/rejected": -2405.333740234375, | |
| "loss": 31.4477, | |
| "rewards/accuracies": 0.5600000619888306, | |
| "rewards/chosen": 0.05013390630483627, | |
| "rewards/margins": 0.029845798388123512, | |
| "rewards/rejected": 0.02028810977935791, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.391427425507943e-07, | |
| "logits/chosen": -1.6959331035614014, | |
| "logits/rejected": -1.6784296035766602, | |
| "logps/chosen": -2696.2236328125, | |
| "logps/rejected": -2173.48583984375, | |
| "loss": 32.2174, | |
| "rewards/accuracies": 0.5600000619888306, | |
| "rewards/chosen": 0.01727980561554432, | |
| "rewards/margins": 0.013602805323898792, | |
| "rewards/rejected": 0.0036770000588148832, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2773657147021465e-07, | |
| "logits/chosen": -1.8469693660736084, | |
| "logits/rejected": -1.7459551095962524, | |
| "logps/chosen": -3117.762451171875, | |
| "logps/rejected": -2390.564208984375, | |
| "loss": 37.6526, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.01079155970364809, | |
| "rewards/margins": 0.011436818167567253, | |
| "rewards/rejected": -0.00064525764901191, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1637694856204885e-07, | |
| "logits/chosen": -1.751587152481079, | |
| "logits/rejected": -1.6395552158355713, | |
| "logps/chosen": -2887.770751953125, | |
| "logps/rejected": -2129.771728515625, | |
| "loss": 53.6906, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.004514098167419434, | |
| "rewards/margins": 0.00042482782737351954, | |
| "rewards/rejected": 0.004089272115379572, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0508762442180743e-07, | |
| "logits/chosen": -1.8443762063980103, | |
| "logits/rejected": -1.792295217514038, | |
| "logps/chosen": -2964.06494140625, | |
| "logps/rejected": -2577.50244140625, | |
| "loss": 62.4137, | |
| "rewards/accuracies": 0.5600000023841858, | |
| "rewards/chosen": 0.04290894791483879, | |
| "rewards/margins": 0.01085699163377285, | |
| "rewards/rejected": 0.03205195814371109, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.93892202664981e-07, | |
| "logits/chosen": -1.6403262615203857, | |
| "logits/rejected": -1.712969183921814, | |
| "logps/chosen": -2689.706787109375, | |
| "logps/rejected": -2513.2998046875, | |
| "loss": 31.7885, | |
| "rewards/accuracies": 0.5100000500679016, | |
| "rewards/chosen": 0.01223880797624588, | |
| "rewards/margins": 0.011182873509824276, | |
| "rewards/rejected": 0.0010559323709458113, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8281409057681686e-07, | |
| "logits/chosen": -1.651449203491211, | |
| "logits/rejected": -1.5920675992965698, | |
| "logps/chosen": -3211.50341796875, | |
| "logps/rejected": -2753.0322265625, | |
| "loss": 103.2519, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.0340498685836792, | |
| "rewards/margins": 0.005449384916573763, | |
| "rewards/rejected": 0.028600484132766724, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7187645017258195e-07, | |
| "logits/chosen": -1.823428750038147, | |
| "logits/rejected": -1.7740917205810547, | |
| "logps/chosen": -2745.991455078125, | |
| "logps/rejected": -2407.27978515625, | |
| "loss": 48.2582, | |
| "rewards/accuracies": 0.5300000309944153, | |
| "rewards/chosen": 0.03057839907705784, | |
| "rewards/margins": 0.003420495195314288, | |
| "rewards/rejected": 0.027157902717590332, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6110214977063343e-07, | |
| "logits/chosen": -1.7967636585235596, | |
| "logits/rejected": -1.7410199642181396, | |
| "logps/chosen": -2905.251708984375, | |
| "logps/rejected": -2435.401611328125, | |
| "loss": 36.2204, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.013903191313147545, | |
| "rewards/margins": 0.00013892585411667824, | |
| "rewards/rejected": 0.013764267787337303, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_logits/chosen": -1.6842743158340454, | |
| "eval_logits/rejected": -1.6236169338226318, | |
| "eval_logps/chosen": -2806.076171875, | |
| "eval_logps/rejected": -2491.544677734375, | |
| "eval_loss": 31.220157623291016, | |
| "eval_rewards/accuracies": 0.53515625, | |
| "eval_rewards/chosen": 0.030346479266881943, | |
| "eval_rewards/margins": 0.009465347044169903, | |
| "eval_rewards/rejected": 0.020881133154034615, | |
| "eval_runtime": 112.3374, | |
| "eval_samples_per_second": 17.804, | |
| "eval_steps_per_second": 0.285, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5051371617954777e-07, | |
| "logits/chosen": -1.6810442209243774, | |
| "logits/rejected": -1.6596931219100952, | |
| "logps/chosen": -2559.396728515625, | |
| "logps/rejected": -2228.120361328125, | |
| "loss": 44.2046, | |
| "rewards/accuracies": 0.5399999618530273, | |
| "rewards/chosen": 0.016220757737755775, | |
| "rewards/margins": 0.007862111553549767, | |
| "rewards/rejected": 0.00835864432156086, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4013328759927622e-07, | |
| "logits/chosen": -1.6315361261367798, | |
| "logits/rejected": -1.6191142797470093, | |
| "logps/chosen": -2893.280029296875, | |
| "logps/rejected": -2805.344970703125, | |
| "loss": 31.2508, | |
| "rewards/accuracies": 0.6299999952316284, | |
| "rewards/chosen": 0.02805119752883911, | |
| "rewards/margins": 0.01223050244152546, | |
| "rewards/rejected": 0.0158206969499588, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.2998256733479896e-07, | |
| "logits/chosen": -1.810739278793335, | |
| "logits/rejected": -1.8173195123672485, | |
| "logps/chosen": -2332.383056640625, | |
| "logps/rejected": -1922.5443115234375, | |
| "loss": 226.0483, | |
| "rewards/accuracies": 0.5600000023841858, | |
| "rewards/chosen": 0.020228227600455284, | |
| "rewards/margins": 0.009395391680300236, | |
| "rewards/rejected": 0.010832836851477623, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.200827784190537e-07, | |
| "logits/chosen": -1.6795597076416016, | |
| "logits/rejected": -1.6883628368377686, | |
| "logps/chosen": -3027.91796875, | |
| "logps/rejected": -2619.313232421875, | |
| "loss": 29.3654, | |
| "rewards/accuracies": 0.5600000619888306, | |
| "rewards/chosen": 0.01968817412853241, | |
| "rewards/margins": 0.008831174112856388, | |
| "rewards/rejected": 0.010857000946998596, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1045461924001323e-07, | |
| "logits/chosen": -1.791738748550415, | |
| "logits/rejected": -1.8031442165374756, | |
| "logps/chosen": -2852.6904296875, | |
| "logps/rejected": -2462.853271484375, | |
| "loss": 45.3966, | |
| "rewards/accuracies": 0.46000003814697266, | |
| "rewards/chosen": 0.010838394984602928, | |
| "rewards/margins": 0.003685446921736002, | |
| "rewards/rejected": 0.007152946200221777, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0111822026468514e-07, | |
| "logits/chosen": -1.7872514724731445, | |
| "logits/rejected": -1.658860445022583, | |
| "logps/chosen": -2903.530029296875, | |
| "logps/rejected": -2319.64599609375, | |
| "loss": 67.4473, | |
| "rewards/accuracies": 0.5700000524520874, | |
| "rewards/chosen": 0.009608490392565727, | |
| "rewards/margins": 0.004570655524730682, | |
| "rewards/rejected": 0.0050378344021737576, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.209310195051581e-08, | |
| "logits/chosen": -1.8252109289169312, | |
| "logits/rejected": -1.6855742931365967, | |
| "logps/chosen": -2538.860107421875, | |
| "logps/rejected": -1955.6048583984375, | |
| "loss": 63.0174, | |
| "rewards/accuracies": 0.6299999952316284, | |
| "rewards/chosen": 0.04829864576458931, | |
| "rewards/margins": 0.022980675101280212, | |
| "rewards/rejected": 0.025317972525954247, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.339813393219713e-08, | |
| "logits/chosen": -1.739793062210083, | |
| "logits/rejected": -1.641005516052246, | |
| "logps/chosen": -2791.561767578125, | |
| "logps/rejected": -2475.72998046875, | |
| "loss": 59.8369, | |
| "rewards/accuracies": 0.5699999928474426, | |
| "rewards/chosen": 0.05231914669275284, | |
| "rewards/margins": 0.021858692169189453, | |
| "rewards/rejected": 0.030460450798273087, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.505149556920698e-08, | |
| "logits/chosen": -1.8431494235992432, | |
| "logits/rejected": -1.7774893045425415, | |
| "logps/chosen": -2542.13427734375, | |
| "logps/rejected": -2193.825927734375, | |
| "loss": 29.3999, | |
| "rewards/accuracies": 0.5800000429153442, | |
| "rewards/chosen": 0.04752471297979355, | |
| "rewards/margins": 0.017444033175706863, | |
| "rewards/rejected": 0.030080681666731834, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.707063793657064e-08, | |
| "logits/chosen": -1.7773969173431396, | |
| "logits/rejected": -1.6891686916351318, | |
| "logps/chosen": -2942.21240234375, | |
| "logps/rejected": -2429.352294921875, | |
| "loss": 99.7738, | |
| "rewards/accuracies": 0.6200000047683716, | |
| "rewards/chosen": 0.03306427597999573, | |
| "rewards/margins": 0.01405587512999773, | |
| "rewards/rejected": 0.019008399918675423, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_logits/chosen": -1.6827195882797241, | |
| "eval_logits/rejected": -1.6222153902053833, | |
| "eval_logps/chosen": -2804.348388671875, | |
| "eval_logps/rejected": -2489.908935546875, | |
| "eval_loss": 33.74028778076172, | |
| "eval_rewards/accuracies": 0.5390625, | |
| "eval_rewards/chosen": 0.04762275516986847, | |
| "eval_rewards/margins": 0.010385587811470032, | |
| "eval_rewards/rejected": 0.037237171083688736, | |
| "eval_runtime": 106.3716, | |
| "eval_samples_per_second": 18.802, | |
| "eval_steps_per_second": 0.301, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.947224733831363e-08, | |
| "logits/chosen": -1.759399175643921, | |
| "logits/rejected": -1.7431520223617554, | |
| "logps/chosen": -2756.701416015625, | |
| "logps/rejected": -2470.905029296875, | |
| "loss": 51.5387, | |
| "rewards/accuracies": 0.5199999809265137, | |
| "rewards/chosen": 0.011415710672736168, | |
| "rewards/margins": 0.009652274660766125, | |
| "rewards/rejected": 0.0017634350806474686, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.227221041988955e-08, | |
| "logits/chosen": -1.7857062816619873, | |
| "logits/rejected": -1.725630760192871, | |
| "logps/chosen": -2520.410400390625, | |
| "logps/rejected": -2319.78564453125, | |
| "loss": 28.3912, | |
| "rewards/accuracies": 0.5900000333786011, | |
| "rewards/chosen": 0.01811736635863781, | |
| "rewards/margins": 0.01297797542065382, | |
| "rewards/rejected": 0.005139390472322702, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.548558095252758e-08, | |
| "logits/chosen": -1.6374757289886475, | |
| "logits/rejected": -1.673044204711914, | |
| "logps/chosen": -2845.2119140625, | |
| "logps/rejected": -2698.294677734375, | |
| "loss": 42.0619, | |
| "rewards/accuracies": 0.48000001907348633, | |
| "rewards/chosen": 0.030720695853233337, | |
| "rewards/margins": 0.020541973412036896, | |
| "rewards/rejected": 0.010178723372519016, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9126548358945635e-08, | |
| "logits/chosen": -1.7063062191009521, | |
| "logits/rejected": -1.6988853216171265, | |
| "logps/chosen": -3136.520263671875, | |
| "logps/rejected": -2731.25927734375, | |
| "loss": 46.0608, | |
| "rewards/accuracies": 0.5600000619888306, | |
| "rewards/chosen": 0.029350021854043007, | |
| "rewards/margins": 0.010425332933664322, | |
| "rewards/rejected": 0.018924688920378685, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.3208408046234896e-08, | |
| "logits/chosen": -1.8164535760879517, | |
| "logits/rejected": -1.7656440734863281, | |
| "logps/chosen": -2538.8046875, | |
| "logps/rejected": -2061.81591796875, | |
| "loss": 40.0838, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.01678399369120598, | |
| "rewards/margins": 0.009666666388511658, | |
| "rewards/rejected": 0.007117328234016895, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.774353360794493e-08, | |
| "logits/chosen": -1.7155154943466187, | |
| "logits/rejected": -1.7442939281463623, | |
| "logps/chosen": -2761.740966796875, | |
| "logps/rejected": -2534.80419921875, | |
| "loss": 36.8374, | |
| "rewards/accuracies": 0.64000004529953, | |
| "rewards/chosen": 0.03588343411684036, | |
| "rewards/margins": 0.037469957023859024, | |
| "rewards/rejected": -0.0015865217428654432, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.2743350953487422e-08, | |
| "logits/chosen": -1.6992709636688232, | |
| "logits/rejected": -1.7416222095489502, | |
| "logps/chosen": -2850.97705078125, | |
| "logps/rejected": -2569.76611328125, | |
| "loss": 86.4259, | |
| "rewards/accuracies": 0.5300000905990601, | |
| "rewards/chosen": 0.018378589302301407, | |
| "rewards/margins": 0.004103804472833872, | |
| "rewards/rejected": 0.014274786226451397, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8218314418949387e-08, | |
| "logits/chosen": -1.718764305114746, | |
| "logits/rejected": -1.6741740703582764, | |
| "logps/chosen": -2353.093017578125, | |
| "logps/rejected": -2190.77001953125, | |
| "loss": 46.1473, | |
| "rewards/accuracies": 0.5699999928474426, | |
| "rewards/chosen": 0.0031641994137316942, | |
| "rewards/margins": 0.0004996396601200104, | |
| "rewards/rejected": 0.002664559753611684, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.4177884909263277e-08, | |
| "logits/chosen": -1.6867101192474365, | |
| "logits/rejected": -1.652515172958374, | |
| "logps/chosen": -2937.97412109375, | |
| "logps/rejected": -2552.07177734375, | |
| "loss": 37.0029, | |
| "rewards/accuracies": 0.5199999809265137, | |
| "rewards/chosen": 0.007288885302841663, | |
| "rewards/margins": -0.0001541988895041868, | |
| "rewards/rejected": 0.007443083915859461, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.063051011743335e-08, | |
| "logits/chosen": -1.7554800510406494, | |
| "logits/rejected": -1.7419729232788086, | |
| "logps/chosen": -2755.109619140625, | |
| "logps/rejected": -2368.246337890625, | |
| "loss": 41.8506, | |
| "rewards/accuracies": 0.46000003814697266, | |
| "rewards/chosen": 0.006798497401177883, | |
| "rewards/margins": 0.010020612739026546, | |
| "rewards/rejected": -0.0032221146393567324, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_logits/chosen": -1.682308554649353, | |
| "eval_logits/rejected": -1.6210675239562988, | |
| "eval_logps/chosen": -2806.1005859375, | |
| "eval_logps/rejected": -2491.68505859375, | |
| "eval_loss": 32.91334915161133, | |
| "eval_rewards/accuracies": 0.5546875, | |
| "eval_rewards/chosen": 0.030103469267487526, | |
| "eval_rewards/margins": 0.01062812004238367, | |
| "eval_rewards/rejected": 0.01947534643113613, | |
| "eval_runtime": 110.9725, | |
| "eval_samples_per_second": 18.022, | |
| "eval_steps_per_second": 0.288, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.58360686217671e-09, | |
| "logits/chosen": -1.7843902111053467, | |
| "logits/rejected": -1.6857761144638062, | |
| "logps/chosen": -2821.310546875, | |
| "logps/rejected": -2445.586669921875, | |
| "loss": 44.0616, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.025277357548475266, | |
| "rewards/margins": 0.014521745964884758, | |
| "rewards/rejected": 0.010755611583590508, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.043545580906694e-09, | |
| "logits/chosen": -1.7102206945419312, | |
| "logits/rejected": -1.6009165048599243, | |
| "logps/chosen": -2682.336669921875, | |
| "logps/rejected": -2234.36279296875, | |
| "loss": 44.9558, | |
| "rewards/accuracies": 0.6299999952316284, | |
| "rewards/chosen": 0.018401915207505226, | |
| "rewards/margins": 0.030992329120635986, | |
| "rewards/rejected": -0.012590417638421059, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.015637010480576e-09, | |
| "logits/chosen": -1.7701480388641357, | |
| "logits/rejected": -1.7436597347259521, | |
| "logps/chosen": -3042.88330078125, | |
| "logps/rejected": -2501.55615234375, | |
| "loss": 35.3819, | |
| "rewards/accuracies": 0.5199999809265137, | |
| "rewards/chosen": 0.012798592448234558, | |
| "rewards/margins": -0.0020437492057681084, | |
| "rewards/rejected": 0.014842341654002666, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5041210835596285e-09, | |
| "logits/chosen": -1.703537940979004, | |
| "logits/rejected": -1.6897165775299072, | |
| "logps/chosen": -2817.1484375, | |
| "logps/rejected": -2390.23095703125, | |
| "loss": 55.7217, | |
| "rewards/accuracies": 0.5800000429153442, | |
| "rewards/chosen": 0.022753870114684105, | |
| "rewards/margins": 0.017634030431509018, | |
| "rewards/rejected": 0.005119838751852512, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5.121580637968137e-10, | |
| "logits/chosen": -1.7322509288787842, | |
| "logits/rejected": -1.6345192193984985, | |
| "logps/chosen": -2836.63623046875, | |
| "logps/rejected": -2363.93505859375, | |
| "loss": 67.2692, | |
| "rewards/accuracies": 0.6300000548362732, | |
| "rewards/chosen": 0.02889620140194893, | |
| "rewards/margins": 0.014323192648589611, | |
| "rewards/rejected": 0.014573007822036743, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.1821938386477075e-11, | |
| "logits/chosen": -1.7962977886199951, | |
| "logits/rejected": -1.7127879858016968, | |
| "logps/chosen": -2801.806640625, | |
| "logps/rejected": -2341.844970703125, | |
| "loss": 46.0674, | |
| "rewards/accuracies": 0.5700000524520874, | |
| "rewards/chosen": 0.009056088514626026, | |
| "rewards/margins": 0.013271180912852287, | |
| "rewards/rejected": -0.004215092398226261, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 764, | |
| "total_flos": 0.0, | |
| "train_loss": 46.613421885577296, | |
| "train_runtime": 4597.6924, | |
| "train_samples_per_second": 13.297, | |
| "train_steps_per_second": 0.166 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 764, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 5, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |