| { | |
| "best_metric": 14.046432495117188, | |
| "best_model_checkpoint": "./qwen2.5-0.5b/qwen2.5-0.5b-expo-L2EXPO-ES-100/checkpoint-550", | |
| "epoch": 2.9806329711856403, | |
| "eval_steps": 50, | |
| "global_step": 1050, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "dpo_loss": 0.6931471824645996, | |
| "epoch": 0.002834199338686821, | |
| "grad_norm": 36884.87916049903, | |
| "learning_rate": 2.840909090909091e-08, | |
| "logits": -1.359458565711975, | |
| "logps": -84.69721221923828, | |
| "loss": 0.3913, | |
| "objective": 0.3618059456348419, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5833333134651184, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 0.3618059456348419, | |
| "step": 1, | |
| "wo_beta": 14.830931663513184 | |
| }, | |
| { | |
| "dpo_loss": 20.195844650268555, | |
| "epoch": 0.14170996693434104, | |
| "grad_norm": 35665.35173471636, | |
| "learning_rate": 1.4204545454545458e-06, | |
| "logits": -1.4575351476669312, | |
| "logps": -84.27513122558594, | |
| "loss": 43.2587, | |
| "objective": 41.916500091552734, | |
| "ranking_idealized": 0.5221088528633118, | |
| "ranking_idealized_expo": 0.5216836929321289, | |
| "ranking_simple": 0.5216836929321289, | |
| "regularize": 41.916500091552734, | |
| "step": 50, | |
| "wo_beta": 15.635692596435547 | |
| }, | |
| { | |
| "epoch": 0.14170996693434104, | |
| "eval_dpo_loss": 26.447525024414062, | |
| "eval_logits": -1.4447709321975708, | |
| "eval_logps": -90.52921295166016, | |
| "eval_loss": 52.66217041015625, | |
| "eval_objective": 53.697696685791016, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5263975262641907, | |
| "eval_regularize": 53.697696685791016, | |
| "eval_runtime": 309.407, | |
| "eval_samples_per_second": 18.713, | |
| "eval_steps_per_second": 1.561, | |
| "eval_wo_beta": 16.170011520385742, | |
| "step": 50 | |
| }, | |
| { | |
| "dpo_loss": 89.97029876708984, | |
| "epoch": 0.2834199338686821, | |
| "grad_norm": 28235.60144716246, | |
| "learning_rate": 2.8409090909090916e-06, | |
| "logits": -1.398974061012268, | |
| "logps": -82.89569091796875, | |
| "loss": 169.8852, | |
| "objective": 172.36553955078125, | |
| "ranking_idealized": 0.5137500166893005, | |
| "ranking_idealized_expo": 0.5137500166893005, | |
| "ranking_simple": 0.51583331823349, | |
| "regularize": 172.36553955078125, | |
| "step": 100, | |
| "wo_beta": 15.30754566192627 | |
| }, | |
| { | |
| "epoch": 0.2834199338686821, | |
| "eval_dpo_loss": 85.763916015625, | |
| "eval_logits": -1.3620884418487549, | |
| "eval_logps": -85.2786636352539, | |
| "eval_loss": 173.986083984375, | |
| "eval_objective": 172.1890869140625, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5243270993232727, | |
| "eval_regularize": 172.1890869140625, | |
| "eval_runtime": 308.6643, | |
| "eval_samples_per_second": 18.758, | |
| "eval_steps_per_second": 1.565, | |
| "eval_wo_beta": 15.439105033874512, | |
| "step": 100 | |
| }, | |
| { | |
| "dpo_loss": 149.43614196777344, | |
| "epoch": 0.42512990080302315, | |
| "grad_norm": 21173.096618846714, | |
| "learning_rate": 4.2613636363636365e-06, | |
| "logits": -1.213483214378357, | |
| "logps": -78.55652618408203, | |
| "loss": 285.0432, | |
| "objective": 280.7228088378906, | |
| "ranking_idealized": 0.527916669845581, | |
| "ranking_idealized_expo": 0.527916669845581, | |
| "ranking_simple": 0.5229166746139526, | |
| "regularize": 280.7228088378906, | |
| "step": 150, | |
| "wo_beta": 15.080223083496094 | |
| }, | |
| { | |
| "epoch": 0.42512990080302315, | |
| "eval_dpo_loss": 143.02996826171875, | |
| "eval_logits": -1.1693531274795532, | |
| "eval_logps": -83.2181167602539, | |
| "eval_loss": 291.4833679199219, | |
| "eval_objective": 293.4403991699219, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5279502868652344, | |
| "eval_regularize": 293.4403991699219, | |
| "eval_runtime": 311.5234, | |
| "eval_samples_per_second": 18.586, | |
| "eval_steps_per_second": 1.55, | |
| "eval_wo_beta": 15.222454071044922, | |
| "step": 150 | |
| }, | |
| { | |
| "dpo_loss": 181.66571044921875, | |
| "epoch": 0.5668398677373642, | |
| "grad_norm": 20619.52737873687, | |
| "learning_rate": 4.997168347957521e-06, | |
| "logits": -0.9392554759979248, | |
| "logps": -78.61503601074219, | |
| "loss": 355.4066, | |
| "objective": 356.2313232421875, | |
| "ranking_idealized": 0.51541668176651, | |
| "ranking_idealized_expo": 0.51541668176651, | |
| "ranking_simple": 0.5191666483879089, | |
| "regularize": 356.2313232421875, | |
| "step": 200, | |
| "wo_beta": 15.349950790405273 | |
| }, | |
| { | |
| "epoch": 0.5668398677373642, | |
| "eval_dpo_loss": 189.846923828125, | |
| "eval_logits": -0.9273601174354553, | |
| "eval_logps": -84.03199768066406, | |
| "eval_loss": 372.7905578613281, | |
| "eval_objective": 365.21240234375, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5232919454574585, | |
| "eval_regularize": 365.21240234375, | |
| "eval_runtime": 308.8911, | |
| "eval_samples_per_second": 18.744, | |
| "eval_steps_per_second": 1.564, | |
| "eval_wo_beta": 14.868427276611328, | |
| "step": 200 | |
| }, | |
| { | |
| "dpo_loss": 186.671142578125, | |
| "epoch": 0.7085498346717053, | |
| "grad_norm": 17849.874243754268, | |
| "learning_rate": 4.973122855144066e-06, | |
| "logits": -0.822012186050415, | |
| "logps": -76.92431640625, | |
| "loss": 368.9811, | |
| "objective": 366.2552490234375, | |
| "ranking_idealized": 0.5162500143051147, | |
| "ranking_idealized_expo": 0.5162500143051147, | |
| "ranking_simple": 0.5074999928474426, | |
| "regularize": 366.2552490234375, | |
| "step": 250, | |
| "wo_beta": 15.5939359664917 | |
| }, | |
| { | |
| "epoch": 0.7085498346717053, | |
| "eval_dpo_loss": 216.4584197998047, | |
| "eval_logits": -0.7745867967605591, | |
| "eval_logps": -81.50496673583984, | |
| "eval_loss": 446.6966247558594, | |
| "eval_objective": 442.3320617675781, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5258799195289612, | |
| "eval_regularize": 442.3320617675781, | |
| "eval_runtime": 308.996, | |
| "eval_samples_per_second": 18.738, | |
| "eval_steps_per_second": 1.563, | |
| "eval_wo_beta": 14.479047775268555, | |
| "step": 250 | |
| }, | |
| { | |
| "dpo_loss": 182.8459930419922, | |
| "epoch": 0.8502598016060463, | |
| "grad_norm": 16846.286343055544, | |
| "learning_rate": 4.924776641419513e-06, | |
| "logits": -0.6342157125473022, | |
| "logps": -78.78164672851562, | |
| "loss": 360.5868, | |
| "objective": 363.29473876953125, | |
| "ranking_idealized": 0.4950000047683716, | |
| "ranking_idealized_expo": 0.4950000047683716, | |
| "ranking_simple": 0.4970833361148834, | |
| "regularize": 363.29473876953125, | |
| "step": 300, | |
| "wo_beta": 15.358329772949219 | |
| }, | |
| { | |
| "epoch": 0.8502598016060463, | |
| "eval_dpo_loss": 222.88400268554688, | |
| "eval_logits": -0.5983948707580566, | |
| "eval_logps": -82.20111083984375, | |
| "eval_loss": 448.9505920410156, | |
| "eval_objective": 443.9051208496094, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5248447060585022, | |
| "eval_regularize": 443.9051208496094, | |
| "eval_runtime": 309.9292, | |
| "eval_samples_per_second": 18.682, | |
| "eval_steps_per_second": 1.558, | |
| "eval_wo_beta": 14.392961502075195, | |
| "step": 300 | |
| }, | |
| { | |
| "dpo_loss": 171.55615234375, | |
| "epoch": 0.9919697685403873, | |
| "grad_norm": 16864.304890654712, | |
| "learning_rate": 4.8526047530778175e-06, | |
| "logits": -0.6237902641296387, | |
| "logps": -79.1826400756836, | |
| "loss": 338.3987, | |
| "objective": 335.6865234375, | |
| "ranking_idealized": 0.5254166722297668, | |
| "ranking_idealized_expo": 0.5254166722297668, | |
| "ranking_simple": 0.5216666460037231, | |
| "regularize": 335.6865234375, | |
| "step": 350, | |
| "wo_beta": 15.20045280456543 | |
| }, | |
| { | |
| "epoch": 0.9919697685403873, | |
| "eval_dpo_loss": 232.93649291992188, | |
| "eval_logits": -0.78554368019104, | |
| "eval_logps": -84.16381072998047, | |
| "eval_loss": 462.19232177734375, | |
| "eval_objective": 461.2073059082031, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5269151329994202, | |
| "eval_regularize": 461.2073059082031, | |
| "eval_runtime": 308.3435, | |
| "eval_samples_per_second": 18.778, | |
| "eval_steps_per_second": 1.566, | |
| "eval_wo_beta": 14.297853469848633, | |
| "step": 350 | |
| }, | |
| { | |
| "dpo_loss": 158.00930786132812, | |
| "epoch": 1.1336797354747283, | |
| "grad_norm": 16419.225936790586, | |
| "learning_rate": 4.757316345716554e-06, | |
| "logits": -0.689696192741394, | |
| "logps": -80.24182891845703, | |
| "loss": 309.1712, | |
| "objective": 306.0483703613281, | |
| "ranking_idealized": 0.5320833325386047, | |
| "ranking_idealized_expo": 0.5320833325386047, | |
| "ranking_simple": 0.5229166746139526, | |
| "regularize": 306.0483703613281, | |
| "step": 400, | |
| "wo_beta": 15.245408058166504 | |
| }, | |
| { | |
| "epoch": 1.1336797354747283, | |
| "eval_dpo_loss": 248.07177734375, | |
| "eval_logits": -0.6413922905921936, | |
| "eval_logps": -82.49343872070312, | |
| "eval_loss": 480.5964660644531, | |
| "eval_objective": 478.7404479980469, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5253623127937317, | |
| "eval_regularize": 478.7404479980469, | |
| "eval_runtime": 308.667, | |
| "eval_samples_per_second": 18.758, | |
| "eval_steps_per_second": 1.565, | |
| "eval_wo_beta": 14.387246131896973, | |
| "step": 400 | |
| }, | |
| { | |
| "dpo_loss": 142.9551239013672, | |
| "epoch": 1.2753897024090695, | |
| "grad_norm": 16215.238701636586, | |
| "learning_rate": 4.639847716126855e-06, | |
| "logits": -0.5957368612289429, | |
| "logps": -78.66122436523438, | |
| "loss": 298.1424, | |
| "objective": 298.3353271484375, | |
| "ranking_idealized": 0.5191666483879089, | |
| "ranking_idealized_expo": 0.5191666483879089, | |
| "ranking_simple": 0.5204166769981384, | |
| "regularize": 298.3353271484375, | |
| "step": 450, | |
| "wo_beta": 15.736668586730957 | |
| }, | |
| { | |
| "epoch": 1.2753897024090695, | |
| "eval_dpo_loss": 247.8721923828125, | |
| "eval_logits": -0.7014132142066956, | |
| "eval_logps": -82.14649200439453, | |
| "eval_loss": 480.3255615234375, | |
| "eval_objective": 482.1766052246094, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.523809552192688, | |
| "eval_regularize": 482.1766052246094, | |
| "eval_runtime": 309.6985, | |
| "eval_samples_per_second": 18.696, | |
| "eval_steps_per_second": 1.56, | |
| "eval_wo_beta": 14.369455337524414, | |
| "step": 450 | |
| }, | |
| { | |
| "dpo_loss": 135.57347106933594, | |
| "epoch": 1.4170996693434104, | |
| "grad_norm": 15944.03549179383, | |
| "learning_rate": 4.501353102310901e-06, | |
| "logits": -0.5253962278366089, | |
| "logps": -78.11959075927734, | |
| "loss": 282.4504, | |
| "objective": 278.3994445800781, | |
| "ranking_idealized": 0.4970833361148834, | |
| "ranking_idealized_expo": 0.4970833361148834, | |
| "ranking_simple": 0.49958333373069763, | |
| "regularize": 278.3994445800781, | |
| "step": 500, | |
| "wo_beta": 15.295467376708984 | |
| }, | |
| { | |
| "epoch": 1.4170996693434104, | |
| "eval_dpo_loss": 252.20928955078125, | |
| "eval_logits": -0.45775941014289856, | |
| "eval_logps": -83.41010284423828, | |
| "eval_loss": 493.7484436035156, | |
| "eval_objective": 495.763916015625, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5248447060585022, | |
| "eval_regularize": 495.763916015625, | |
| "eval_runtime": 308.7418, | |
| "eval_samples_per_second": 18.754, | |
| "eval_steps_per_second": 1.564, | |
| "eval_wo_beta": 14.174327850341797, | |
| "step": 500 | |
| }, | |
| { | |
| "dpo_loss": 138.2200164794922, | |
| "epoch": 1.5588096362777515, | |
| "grad_norm": 14777.775900953873, | |
| "learning_rate": 4.34319334202531e-06, | |
| "logits": -0.43370625376701355, | |
| "logps": -79.58238220214844, | |
| "loss": 261.1027, | |
| "objective": 261.7706298828125, | |
| "ranking_idealized": 0.5104166865348816, | |
| "ranking_idealized_expo": 0.5104166865348816, | |
| "ranking_simple": 0.5104166865348816, | |
| "regularize": 261.7706298828125, | |
| "step": 550, | |
| "wo_beta": 15.099705696105957 | |
| }, | |
| { | |
| "epoch": 1.5588096362777515, | |
| "eval_dpo_loss": 245.87562561035156, | |
| "eval_logits": -0.5435077548027039, | |
| "eval_logps": -82.8267593383789, | |
| "eval_loss": 486.1625671386719, | |
| "eval_objective": 489.7927551269531, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5253623127937317, | |
| "eval_regularize": 489.7927551269531, | |
| "eval_runtime": 313.2811, | |
| "eval_samples_per_second": 18.482, | |
| "eval_steps_per_second": 1.542, | |
| "eval_wo_beta": 14.046432495117188, | |
| "step": 550 | |
| }, | |
| { | |
| "dpo_loss": 129.54061889648438, | |
| "epoch": 1.7005196032120926, | |
| "grad_norm": 15449.253251441141, | |
| "learning_rate": 4.16692250129073e-06, | |
| "logits": -0.4195112884044647, | |
| "logps": -79.62303924560547, | |
| "loss": 255.9288, | |
| "objective": 256.1207275390625, | |
| "ranking_idealized": 0.5149999856948853, | |
| "ranking_idealized_expo": 0.5149999856948853, | |
| "ranking_simple": 0.5095833539962769, | |
| "regularize": 256.1207275390625, | |
| "step": 600, | |
| "wo_beta": 15.011371612548828 | |
| }, | |
| { | |
| "epoch": 1.7005196032120926, | |
| "eval_dpo_loss": 251.29342651367188, | |
| "eval_logits": -0.5346657633781433, | |
| "eval_logps": -82.17684173583984, | |
| "eval_loss": 500.38006591796875, | |
| "eval_objective": 502.1727294921875, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5269151329994202, | |
| "eval_regularize": 502.1727294921875, | |
| "eval_runtime": 309.4233, | |
| "eval_samples_per_second": 18.712, | |
| "eval_steps_per_second": 1.561, | |
| "eval_wo_beta": 14.243566513061523, | |
| "step": 600 | |
| }, | |
| { | |
| "dpo_loss": 130.9433135986328, | |
| "epoch": 1.8422295701464337, | |
| "grad_norm": 15825.93366010816, | |
| "learning_rate": 3.974272604254906e-06, | |
| "logits": -0.4912276566028595, | |
| "logps": -78.56413269042969, | |
| "loss": 248.6787, | |
| "objective": 253.7882843017578, | |
| "ranking_idealized": 0.527916669845581, | |
| "ranking_idealized_expo": 0.527916669845581, | |
| "ranking_simple": 0.5224999785423279, | |
| "regularize": 253.7882843017578, | |
| "step": 650, | |
| "wo_beta": 15.684560775756836 | |
| }, | |
| { | |
| "epoch": 1.8422295701464337, | |
| "eval_dpo_loss": 254.5959014892578, | |
| "eval_logits": -0.5140498876571655, | |
| "eval_logps": -81.49234771728516, | |
| "eval_loss": 502.3152770996094, | |
| "eval_objective": 504.1581726074219, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5248447060585022, | |
| "eval_regularize": 504.1581726074219, | |
| "eval_runtime": 308.5565, | |
| "eval_samples_per_second": 18.765, | |
| "eval_steps_per_second": 1.565, | |
| "eval_wo_beta": 14.331953048706055, | |
| "step": 650 | |
| }, | |
| { | |
| "dpo_loss": 112.15733337402344, | |
| "epoch": 1.9839395370807746, | |
| "grad_norm": 16229.838704301123, | |
| "learning_rate": 3.767136614452458e-06, | |
| "logits": -0.4758701026439667, | |
| "logps": -78.98329162597656, | |
| "loss": 226.4676, | |
| "objective": 224.5537567138672, | |
| "ranking_idealized": 0.5112500190734863, | |
| "ranking_idealized_expo": 0.5108333230018616, | |
| "ranking_simple": 0.5087500214576721, | |
| "regularize": 224.5537567138672, | |
| "step": 700, | |
| "wo_beta": 15.5169095993042 | |
| }, | |
| { | |
| "epoch": 1.9839395370807746, | |
| "eval_dpo_loss": 264.166015625, | |
| "eval_logits": -0.4816124141216278, | |
| "eval_logps": -83.42156219482422, | |
| "eval_loss": 512.698974609375, | |
| "eval_objective": 516.7103271484375, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5253623127937317, | |
| "eval_regularize": 516.7103271484375, | |
| "eval_runtime": 308.9876, | |
| "eval_samples_per_second": 18.739, | |
| "eval_steps_per_second": 1.563, | |
| "eval_wo_beta": 14.08342456817627, | |
| "step": 700 | |
| }, | |
| { | |
| "dpo_loss": 105.4539566040039, | |
| "epoch": 2.1256495040151155, | |
| "grad_norm": 15281.11762954245, | |
| "learning_rate": 3.547549834686222e-06, | |
| "logits": -0.47171100974082947, | |
| "logps": -79.71944427490234, | |
| "loss": 207.1551, | |
| "objective": 207.78355407714844, | |
| "ranking_idealized": 0.5112500190734863, | |
| "ranking_idealized_expo": 0.5112500190734863, | |
| "ranking_simple": 0.51583331823349, | |
| "regularize": 207.78355407714844, | |
| "step": 750, | |
| "wo_beta": 15.408516883850098 | |
| }, | |
| { | |
| "epoch": 2.1256495040151155, | |
| "eval_dpo_loss": 259.2528381347656, | |
| "eval_logits": -0.5409926772117615, | |
| "eval_logps": -83.45890045166016, | |
| "eval_loss": 506.4237365722656, | |
| "eval_objective": 510.6129150390625, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.523809552192688, | |
| "eval_regularize": 510.6129150390625, | |
| "eval_runtime": 308.6604, | |
| "eval_samples_per_second": 18.758, | |
| "eval_steps_per_second": 1.565, | |
| "eval_wo_beta": 14.129502296447754, | |
| "step": 750 | |
| }, | |
| { | |
| "dpo_loss": 97.64974975585938, | |
| "epoch": 2.2673594709494567, | |
| "grad_norm": 15212.863475959, | |
| "learning_rate": 3.3176699082935546e-06, | |
| "logits": -0.48046550154685974, | |
| "logps": -81.52581787109375, | |
| "loss": 197.3545, | |
| "objective": 198.5575714111328, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 198.5575714111328, | |
| "step": 800, | |
| "wo_beta": 15.000423431396484 | |
| }, | |
| { | |
| "epoch": 2.2673594709494567, | |
| "eval_dpo_loss": 262.3102111816406, | |
| "eval_logits": -0.5658813118934631, | |
| "eval_logps": -84.87467956542969, | |
| "eval_loss": 513.39794921875, | |
| "eval_objective": 514.31201171875, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.522774338722229, | |
| "eval_regularize": 514.31201171875, | |
| "eval_runtime": 310.3348, | |
| "eval_samples_per_second": 18.657, | |
| "eval_steps_per_second": 1.556, | |
| "eval_wo_beta": 14.070388793945312, | |
| "step": 800 | |
| }, | |
| { | |
| "dpo_loss": 94.56294250488281, | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 14099.488226379159, | |
| "learning_rate": 3.0797556183036582e-06, | |
| "logits": -0.5114213824272156, | |
| "logps": -81.04893493652344, | |
| "loss": 182.3796, | |
| "objective": 182.07008361816406, | |
| "ranking_idealized": 0.5145833492279053, | |
| "ranking_idealized_expo": 0.5137500166893005, | |
| "ranking_simple": 0.5129166841506958, | |
| "regularize": 182.07008361816406, | |
| "step": 850, | |
| "wo_beta": 15.241059303283691 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "eval_dpo_loss": 254.12506103515625, | |
| "eval_logits": -0.5509631037712097, | |
| "eval_logps": -82.86239624023438, | |
| "eval_loss": 501.88311767578125, | |
| "eval_objective": 504.852294921875, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5274327397346497, | |
| "eval_regularize": 504.852294921875, | |
| "eval_runtime": 309.6743, | |
| "eval_samples_per_second": 18.697, | |
| "eval_steps_per_second": 1.56, | |
| "eval_wo_beta": 14.17065143585205, | |
| "step": 850 | |
| }, | |
| { | |
| "dpo_loss": 92.21830749511719, | |
| "epoch": 2.555503070382617, | |
| "grad_norm": 14214.15214964791, | |
| "learning_rate": 2.8361446928038298e-06, | |
| "logits": -0.49887704849243164, | |
| "logps": -80.69136810302734, | |
| "loss": 176.042, | |
| "objective": 176.71592712402344, | |
| "ranking_idealized": 0.5183333158493042, | |
| "ranking_idealized_expo": 0.5179166793823242, | |
| "ranking_simple": 0.5129166841506958, | |
| "regularize": 176.71592712402344, | |
| "step": 900, | |
| "wo_beta": 15.599862098693848 | |
| }, | |
| { | |
| "epoch": 2.555503070382617, | |
| "eval_dpo_loss": 263.27996826171875, | |
| "eval_logits": -0.5038771033287048, | |
| "eval_logps": -85.07097625732422, | |
| "eval_loss": 518.1983032226562, | |
| "eval_objective": 519.5007934570312, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.523809552192688, | |
| "eval_regularize": 519.5007934570312, | |
| "eval_runtime": 308.5117, | |
| "eval_samples_per_second": 18.768, | |
| "eval_steps_per_second": 1.566, | |
| "eval_wo_beta": 14.112290382385254, | |
| "step": 900 | |
| }, | |
| { | |
| "dpo_loss": 81.76506805419922, | |
| "epoch": 2.697213037316958, | |
| "grad_norm": 14116.155586290091, | |
| "learning_rate": 2.5892308345974517e-06, | |
| "logits": -0.45734792947769165, | |
| "logps": -81.56855773925781, | |
| "loss": 164.8281, | |
| "objective": 161.75088500976562, | |
| "ranking_idealized": 0.5045833587646484, | |
| "ranking_idealized_expo": 0.5045833587646484, | |
| "ranking_simple": 0.5099999904632568, | |
| "regularize": 161.75088500976562, | |
| "step": 950, | |
| "wo_beta": 15.513051986694336 | |
| }, | |
| { | |
| "epoch": 2.697213037316958, | |
| "eval_dpo_loss": 262.8074035644531, | |
| "eval_logits": -0.5199795961380005, | |
| "eval_logps": -84.58429718017578, | |
| "eval_loss": 512.1844482421875, | |
| "eval_objective": 512.7650756835938, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.523809552192688, | |
| "eval_regularize": 512.7650756835938, | |
| "eval_runtime": 309.6572, | |
| "eval_samples_per_second": 18.698, | |
| "eval_steps_per_second": 1.56, | |
| "eval_wo_beta": 14.164327621459961, | |
| "step": 950 | |
| }, | |
| { | |
| "dpo_loss": 76.9761962890625, | |
| "epoch": 2.838923004251299, | |
| "grad_norm": 13709.192801586882, | |
| "learning_rate": 2.341440200858589e-06, | |
| "logits": -0.46604040265083313, | |
| "logps": -79.8453598022461, | |
| "loss": 150.0401, | |
| "objective": 150.50038146972656, | |
| "ranking_idealized": 0.5112500190734863, | |
| "ranking_idealized_expo": 0.5112500190734863, | |
| "ranking_simple": 0.5099999904632568, | |
| "regularize": 150.50038146972656, | |
| "step": 1000, | |
| "wo_beta": 15.057799339294434 | |
| }, | |
| { | |
| "epoch": 2.838923004251299, | |
| "eval_dpo_loss": 263.6168518066406, | |
| "eval_logits": -0.5218656659126282, | |
| "eval_logps": -83.73433685302734, | |
| "eval_loss": 514.70361328125, | |
| "eval_objective": 516.595947265625, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5258799195289612, | |
| "eval_regularize": 516.595947265625, | |
| "eval_runtime": 308.5762, | |
| "eval_samples_per_second": 18.764, | |
| "eval_steps_per_second": 1.565, | |
| "eval_wo_beta": 14.180042266845703, | |
| "step": 1000 | |
| }, | |
| { | |
| "dpo_loss": 69.53437805175781, | |
| "epoch": 2.9806329711856403, | |
| "grad_norm": 14565.063000853466, | |
| "learning_rate": 2.0952075638923656e-06, | |
| "logits": -0.4686031639575958, | |
| "logps": -80.63922882080078, | |
| "loss": 141.0317, | |
| "objective": 136.92088317871094, | |
| "ranking_idealized": 0.5174999833106995, | |
| "ranking_idealized_expo": 0.5174999833106995, | |
| "ranking_simple": 0.5262500047683716, | |
| "regularize": 136.92088317871094, | |
| "step": 1050, | |
| "wo_beta": 15.512243270874023 | |
| }, | |
| { | |
| "epoch": 2.9806329711856403, | |
| "eval_dpo_loss": 266.94525146484375, | |
| "eval_logits": -0.49528759717941284, | |
| "eval_logps": -84.26761627197266, | |
| "eval_loss": 519.2467041015625, | |
| "eval_objective": 521.8153076171875, | |
| "eval_ranking_idealized": 0.5212215185165405, | |
| "eval_ranking_idealized_expo": 0.5212215185165405, | |
| "eval_ranking_simple": 0.5263975262641907, | |
| "eval_regularize": 521.8153076171875, | |
| "eval_runtime": 308.2205, | |
| "eval_samples_per_second": 18.785, | |
| "eval_steps_per_second": 1.567, | |
| "eval_wo_beta": 14.257741928100586, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.9806329711856403, | |
| "step": 1050, | |
| "total_flos": 0.0, | |
| "train_loss": 38.77721726190476, | |
| "train_runtime": 6898.2426, | |
| "train_samples_per_second": 36.822, | |
| "train_steps_per_second": 0.255 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 1760, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |