{ "best_metric": 3.6610631942749023, "best_model_checkpoint": "multilingual-e5-small-aligned-readability/checkpoint-40644", "epoch": 3.0, "eval_steps": 500, "global_step": 40644, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03690581635665781, "grad_norm": 48.44145965576172, "learning_rate": 4.938490306072237e-05, "loss": 25.4936, "step": 500 }, { "epoch": 0.07381163271331562, "grad_norm": 56.23558807373047, "learning_rate": 4.876980612144474e-05, "loss": 5.5206, "step": 1000 }, { "epoch": 0.11071744906997343, "grad_norm": 22.196273803710938, "learning_rate": 4.815470918216711e-05, "loss": 4.8622, "step": 1500 }, { "epoch": 0.14762326542663123, "grad_norm": 49.71085739135742, "learning_rate": 4.7539612242889484e-05, "loss": 4.7329, "step": 2000 }, { "epoch": 0.18452908178328906, "grad_norm": 33.54679489135742, "learning_rate": 4.692451530361185e-05, "loss": 4.4192, "step": 2500 }, { "epoch": 0.22143489813994685, "grad_norm": 33.129676818847656, "learning_rate": 4.6309418364334224e-05, "loss": 4.4362, "step": 3000 }, { "epoch": 0.2583407144966047, "grad_norm": 32.080562591552734, "learning_rate": 4.5694321425056594e-05, "loss": 4.2238, "step": 3500 }, { "epoch": 0.29524653085326247, "grad_norm": 49.822025299072266, "learning_rate": 4.507922448577896e-05, "loss": 4.207, "step": 4000 }, { "epoch": 0.33215234720992026, "grad_norm": 38.24595642089844, "learning_rate": 4.4464127546501335e-05, "loss": 4.1541, "step": 4500 }, { "epoch": 0.3690581635665781, "grad_norm": 38.85857391357422, "learning_rate": 4.38490306072237e-05, "loss": 4.1168, "step": 5000 }, { "epoch": 0.4059639799232359, "grad_norm": 26.093944549560547, "learning_rate": 4.323393366794607e-05, "loss": 4.035, "step": 5500 }, { "epoch": 0.4428697962798937, "grad_norm": 38.26423263549805, "learning_rate": 4.261883672866844e-05, "loss": 4.0707, "step": 6000 }, { "epoch": 0.4797756126365515, "grad_norm": 30.129150390625, "learning_rate": 4.200373978939081e-05, "loss": 4.0642, "step": 6500 }, { "epoch": 0.5166814289932093, "grad_norm": 28.189598083496094, "learning_rate": 4.138864285011318e-05, "loss": 4.0913, "step": 7000 }, { "epoch": 0.5535872453498671, "grad_norm": 25.67119789123535, "learning_rate": 4.077354591083555e-05, "loss": 3.8925, "step": 7500 }, { "epoch": 0.5904930617065249, "grad_norm": 40.868289947509766, "learning_rate": 4.015844897155792e-05, "loss": 4.0213, "step": 8000 }, { "epoch": 0.6273988780631827, "grad_norm": 26.44133186340332, "learning_rate": 3.954335203228029e-05, "loss": 4.0, "step": 8500 }, { "epoch": 0.6643046944198405, "grad_norm": 25.9945068359375, "learning_rate": 3.892825509300266e-05, "loss": 3.9923, "step": 9000 }, { "epoch": 0.7012105107764984, "grad_norm": 44.50141143798828, "learning_rate": 3.8313158153725024e-05, "loss": 3.8293, "step": 9500 }, { "epoch": 0.7381163271331562, "grad_norm": 43.48475646972656, "learning_rate": 3.76980612144474e-05, "loss": 3.8908, "step": 10000 }, { "epoch": 0.775022143489814, "grad_norm": 34.09233093261719, "learning_rate": 3.708296427516977e-05, "loss": 3.8289, "step": 10500 }, { "epoch": 0.8119279598464718, "grad_norm": 19.0194034576416, "learning_rate": 3.6467867335892135e-05, "loss": 3.9371, "step": 11000 }, { "epoch": 0.8488337762031296, "grad_norm": 23.901264190673828, "learning_rate": 3.585277039661451e-05, "loss": 3.9056, "step": 11500 }, { "epoch": 0.8857395925597874, "grad_norm": 17.51069450378418, "learning_rate": 3.5237673457336876e-05, "loss": 3.8314, "step": 12000 }, { "epoch": 0.9226454089164452, "grad_norm": 41.49451446533203, "learning_rate": 3.4622576518059246e-05, "loss": 3.8817, "step": 12500 }, { "epoch": 0.959551225273103, "grad_norm": 54.885257720947266, "learning_rate": 3.400747957878162e-05, "loss": 3.7882, "step": 13000 }, { "epoch": 0.9964570416297609, "grad_norm": 23.55073356628418, "learning_rate": 3.3392382639503986e-05, "loss": 3.8356, "step": 13500 }, { "epoch": 1.0, "eval_loss": 3.8022446632385254, "eval_mse": 3.80224462961076, "eval_runtime": 54.6351, "eval_samples_per_second": 1763.299, "eval_steps_per_second": 220.426, "step": 13548 }, { "epoch": 1.0333628579864187, "grad_norm": 19.580780029296875, "learning_rate": 3.277728570022636e-05, "loss": 3.3048, "step": 14000 }, { "epoch": 1.0702686743430765, "grad_norm": 18.51485824584961, "learning_rate": 3.216218876094873e-05, "loss": 3.1909, "step": 14500 }, { "epoch": 1.1071744906997343, "grad_norm": 24.523353576660156, "learning_rate": 3.15470918216711e-05, "loss": 3.134, "step": 15000 }, { "epoch": 1.144080307056392, "grad_norm": 23.871919631958008, "learning_rate": 3.093199488239347e-05, "loss": 3.182, "step": 15500 }, { "epoch": 1.1809861234130499, "grad_norm": 27.006973266601562, "learning_rate": 3.0316897943115834e-05, "loss": 3.3002, "step": 16000 }, { "epoch": 1.2178919397697077, "grad_norm": 46.61240005493164, "learning_rate": 2.9701801003838208e-05, "loss": 3.1962, "step": 16500 }, { "epoch": 1.2547977561263655, "grad_norm": 23.656471252441406, "learning_rate": 2.9086704064560578e-05, "loss": 3.2212, "step": 17000 }, { "epoch": 1.2917035724830233, "grad_norm": 38.43106460571289, "learning_rate": 2.8471607125282945e-05, "loss": 3.2474, "step": 17500 }, { "epoch": 1.328609388839681, "grad_norm": 19.142786026000977, "learning_rate": 2.7856510186005312e-05, "loss": 3.1591, "step": 18000 }, { "epoch": 1.3655152051963388, "grad_norm": 30.900583267211914, "learning_rate": 2.7241413246727686e-05, "loss": 3.1354, "step": 18500 }, { "epoch": 1.4024210215529966, "grad_norm": 25.125627517700195, "learning_rate": 2.6626316307450056e-05, "loss": 3.1656, "step": 19000 }, { "epoch": 1.4393268379096544, "grad_norm": 28.680898666381836, "learning_rate": 2.6011219368172423e-05, "loss": 3.1617, "step": 19500 }, { "epoch": 1.4762326542663124, "grad_norm": 28.307249069213867, "learning_rate": 2.5396122428894797e-05, "loss": 3.1673, "step": 20000 }, { "epoch": 1.51313847062297, "grad_norm": 27.13144302368164, "learning_rate": 2.4781025489617167e-05, "loss": 3.1657, "step": 20500 }, { "epoch": 1.550044286979628, "grad_norm": 17.530426025390625, "learning_rate": 2.4165928550339534e-05, "loss": 3.1608, "step": 21000 }, { "epoch": 1.5869501033362858, "grad_norm": 51.65472412109375, "learning_rate": 2.3550831611061904e-05, "loss": 3.1323, "step": 21500 }, { "epoch": 1.6238559196929436, "grad_norm": 23.71245574951172, "learning_rate": 2.2935734671784274e-05, "loss": 3.2361, "step": 22000 }, { "epoch": 1.6607617360496014, "grad_norm": 17.434144973754883, "learning_rate": 2.2320637732506645e-05, "loss": 3.1969, "step": 22500 }, { "epoch": 1.6976675524062592, "grad_norm": 22.121740341186523, "learning_rate": 2.1705540793229015e-05, "loss": 3.1437, "step": 23000 }, { "epoch": 1.734573368762917, "grad_norm": 19.105876922607422, "learning_rate": 2.1090443853951382e-05, "loss": 3.1227, "step": 23500 }, { "epoch": 1.7714791851195748, "grad_norm": 27.000789642333984, "learning_rate": 2.0475346914673755e-05, "loss": 3.066, "step": 24000 }, { "epoch": 1.8083850014762326, "grad_norm": 24.040668487548828, "learning_rate": 1.9860249975396122e-05, "loss": 3.1173, "step": 24500 }, { "epoch": 1.8452908178328906, "grad_norm": 29.920883178710938, "learning_rate": 1.9245153036118493e-05, "loss": 3.0847, "step": 25000 }, { "epoch": 1.8821966341895484, "grad_norm": 44.850303649902344, "learning_rate": 1.8630056096840863e-05, "loss": 3.0993, "step": 25500 }, { "epoch": 1.9191024505462062, "grad_norm": 66.01287078857422, "learning_rate": 1.8014959157563233e-05, "loss": 3.0649, "step": 26000 }, { "epoch": 1.956008266902864, "grad_norm": 57.569190979003906, "learning_rate": 1.7399862218285603e-05, "loss": 3.1672, "step": 26500 }, { "epoch": 1.9929140832595218, "grad_norm": 16.41364860534668, "learning_rate": 1.678476527900797e-05, "loss": 3.1173, "step": 27000 }, { "epoch": 2.0, "eval_loss": 3.7406532764434814, "eval_mse": 3.7406532746764647, "eval_runtime": 53.1552, "eval_samples_per_second": 1812.39, "eval_steps_per_second": 226.563, "step": 27096 }, { "epoch": 2.0298198996161796, "grad_norm": 16.53102684020996, "learning_rate": 1.6169668339730344e-05, "loss": 2.6256, "step": 27500 }, { "epoch": 2.0667257159728374, "grad_norm": 23.657825469970703, "learning_rate": 1.555457140045271e-05, "loss": 2.6492, "step": 28000 }, { "epoch": 2.103631532329495, "grad_norm": 20.9498233795166, "learning_rate": 1.4939474461175081e-05, "loss": 2.6146, "step": 28500 }, { "epoch": 2.140537348686153, "grad_norm": 16.935983657836914, "learning_rate": 1.4324377521897453e-05, "loss": 2.6738, "step": 29000 }, { "epoch": 2.1774431650428108, "grad_norm": 35.98014450073242, "learning_rate": 1.3709280582619822e-05, "loss": 2.5908, "step": 29500 }, { "epoch": 2.2143489813994686, "grad_norm": 30.703969955444336, "learning_rate": 1.3094183643342192e-05, "loss": 2.6565, "step": 30000 }, { "epoch": 2.2512547977561264, "grad_norm": 40.74715805053711, "learning_rate": 1.2479086704064562e-05, "loss": 2.6106, "step": 30500 }, { "epoch": 2.288160614112784, "grad_norm": 18.857763290405273, "learning_rate": 1.186398976478693e-05, "loss": 2.6383, "step": 31000 }, { "epoch": 2.325066430469442, "grad_norm": 68.10604095458984, "learning_rate": 1.1248892825509301e-05, "loss": 2.5259, "step": 31500 }, { "epoch": 2.3619722468260997, "grad_norm": 37.21859359741211, "learning_rate": 1.0633795886231671e-05, "loss": 2.5619, "step": 32000 }, { "epoch": 2.3988780631827575, "grad_norm": 58.106285095214844, "learning_rate": 1.001869894695404e-05, "loss": 2.6748, "step": 32500 }, { "epoch": 2.4357838795394153, "grad_norm": 47.5319938659668, "learning_rate": 9.40360200767641e-06, "loss": 2.6488, "step": 33000 }, { "epoch": 2.472689695896073, "grad_norm": 20.28361701965332, "learning_rate": 8.78850506839878e-06, "loss": 2.6175, "step": 33500 }, { "epoch": 2.509595512252731, "grad_norm": 19.851940155029297, "learning_rate": 8.17340812912115e-06, "loss": 2.6142, "step": 34000 }, { "epoch": 2.5465013286093887, "grad_norm": 32.45155334472656, "learning_rate": 7.55831118984352e-06, "loss": 2.5369, "step": 34500 }, { "epoch": 2.5834071449660465, "grad_norm": 60.2480354309082, "learning_rate": 6.94321425056589e-06, "loss": 2.5508, "step": 35000 }, { "epoch": 2.6203129613227043, "grad_norm": 16.603755950927734, "learning_rate": 6.328117311288259e-06, "loss": 2.6182, "step": 35500 }, { "epoch": 2.657218777679362, "grad_norm": 20.604549407958984, "learning_rate": 5.713020372010629e-06, "loss": 2.6126, "step": 36000 }, { "epoch": 2.69412459403602, "grad_norm": 18.057249069213867, "learning_rate": 5.097923432732999e-06, "loss": 2.6001, "step": 36500 }, { "epoch": 2.7310304103926777, "grad_norm": 22.73322296142578, "learning_rate": 4.482826493455368e-06, "loss": 2.4779, "step": 37000 }, { "epoch": 2.7679362267493355, "grad_norm": 101.31739807128906, "learning_rate": 3.8677295541777385e-06, "loss": 2.4957, "step": 37500 }, { "epoch": 2.8048420431059933, "grad_norm": 25.522369384765625, "learning_rate": 3.2526326149001084e-06, "loss": 2.5047, "step": 38000 }, { "epoch": 2.841747859462651, "grad_norm": 14.60265827178955, "learning_rate": 2.6375356756224782e-06, "loss": 2.5191, "step": 38500 }, { "epoch": 2.878653675819309, "grad_norm": 39.825443267822266, "learning_rate": 2.022438736344848e-06, "loss": 2.5414, "step": 39000 }, { "epoch": 2.9155594921759667, "grad_norm": 19.467710494995117, "learning_rate": 1.4073417970672177e-06, "loss": 2.564, "step": 39500 }, { "epoch": 2.952465308532625, "grad_norm": 57.59708023071289, "learning_rate": 7.922448577895876e-07, "loss": 2.5478, "step": 40000 }, { "epoch": 2.9893711248892827, "grad_norm": 15.459783554077148, "learning_rate": 1.771479185119575e-07, "loss": 2.6344, "step": 40500 }, { "epoch": 3.0, "eval_loss": 3.6610631942749023, "eval_mse": 3.6610634619705245, "eval_runtime": 53.3029, "eval_samples_per_second": 1807.368, "eval_steps_per_second": 225.935, "step": 40644 }, { "epoch": 3.0, "step": 40644, "total_flos": 4.283504864539085e+16, "train_loss": 3.5571271605219925, "train_runtime": 3169.0778, "train_samples_per_second": 820.783, "train_steps_per_second": 12.825 } ], "logging_steps": 500, "max_steps": 40644, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4.283504864539085e+16, "train_batch_size": 64, "trial_name": null, "trial_params": null }