diff --git a/slots/0/checkpoint-8009/config.json b/slots/0/checkpoint-8009/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/0/checkpoint-8009/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/0/checkpoint-8009/trainer_state.json b/slots/0/checkpoint-8009/trainer_state.json deleted file mode 100644 index 6f925cc05dec8c70e2f2fd8d7ddb847176019495..0000000000000000000000000000000000000000 --- a/slots/0/checkpoint-8009/trainer_state.json +++ /dev/null @@ -1,2841 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.5147158405056085, - "eval_steps": 500, - "global_step": 8009, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 31.780010223388672, - "learning_rate": 4.999985708624277e-05, - "loss": 47.03160400390625, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 22.647113800048828, - "learning_rate": 4.999851618030373e-05, - "loss": 39.12403564453125, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 24.094038009643555, - "learning_rate": 4.9995763869538124e-05, - "loss": 32.462225341796874, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 25.049236297607422, - "learning_rate": 4.999160030934015e-05, - "loss": 25.110198974609375, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 20.87278938293457, - "learning_rate": 4.9986025734782515e-05, - "loss": 24.496664428710936, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 19.164220809936523, - "learning_rate": 4.997904046060316e-05, - "loss": 21.625892639160156, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 17.22857666015625, - "learning_rate": 4.997064488118748e-05, - "loss": 18.14871826171875, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 17.680524826049805, - "learning_rate": 4.996083947054607e-05, - "loss": 19.17234649658203, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 17.1239070892334, - "learning_rate": 4.9949624782287944e-05, - "loss": 17.536575317382812, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 15.556852340698242, - "learning_rate": 4.993700144958927e-05, - "loss": 14.654452514648437, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.166357040405273, - "learning_rate": 4.992297018515767e-05, - "loss": 12.504849243164063, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 15.202767372131348, - "learning_rate": 4.990753178119193e-05, - "loss": 14.847886657714843, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 14.017868995666504, - "learning_rate": 4.989068710933732e-05, - "loss": 10.980046844482422, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 16.90302848815918, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.107038879394532, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 26.98078155517578, - "learning_rate": 4.985278284547495e-05, - "loss": 10.064899444580078, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 19.16476058959961, - "learning_rate": 4.983172539352465e-05, - "loss": 9.747372436523438, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 14.147124290466309, - "learning_rate": 4.980926595367956e-05, - "loss": 7.905096435546875, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 16.14385986328125, - "learning_rate": 4.978540579398941e-05, - "loss": 8.444129943847656, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 13.877010345458984, - "learning_rate": 4.9760146261587936e-05, - "loss": 6.83173828125, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 13.799625396728516, - "learning_rate": 4.973348878261681e-05, - "loss": 7.934703826904297, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 15.579751014709473, - "learning_rate": 4.970543486214516e-05, - "loss": 6.793711853027344, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 13.521928787231445, - "learning_rate": 4.967598608408454e-05, - "loss": 7.2837577819824215, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 15.898303031921387, - "learning_rate": 4.9645144111099515e-05, - "loss": 6.6693672180175785, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 24.22979164123535, - "learning_rate": 4.961291068451385e-05, - "loss": 5.961820983886719, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 11.246175765991211, - "learning_rate": 4.9579287624212093e-05, - "loss": 4.978904724121094, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 24.801748275756836, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.248088073730469, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.347773551940918, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.6643627166748045, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 12.957473754882812, - "learning_rate": 4.94701000160798e-05, - "loss": 4.121147918701172, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.496082305908203, - "learning_rate": 4.943093818728689e-05, - "loss": 3.8258975982666015, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 12.250146865844727, - "learning_rate": 4.939039699886213e-05, - "loss": 3.8255241394042967, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 12.506306648254395, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.5919628143310547, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 16.467334747314453, - "learning_rate": 4.930518577661388e-05, - "loss": 3.715103530883789, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 8.267026901245117, - "learning_rate": 4.926052055377716e-05, - "loss": 3.194218063354492, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 9.150392532348633, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9999900817871095, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 11.638959884643555, - "learning_rate": 4.916708349342674e-05, - "loss": 2.837971496582031, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 10.091157913208008, - "learning_rate": 4.911831693132686e-05, - "loss": 2.7495031356811523, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 9.66382884979248, - "learning_rate": 4.906818866004871e-05, - "loss": 2.6875839233398438, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 8.437216758728027, - "learning_rate": 4.901670150981164e-05, - "loss": 2.5731342315673826, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 8.248567581176758, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.859779930114746, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 11.652517318725586, - "learning_rate": 4.890966227678239e-05, - "loss": 2.2908452987670898, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 14.507343292236328, - "learning_rate": 4.885411623737651e-05, - "loss": 2.302378463745117, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 9.44881820678711, - "learning_rate": 4.87972234054431e-05, - "loss": 2.3078046798706056, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 9.367117881774902, - "learning_rate": 4.87389869931255e-05, - "loss": 2.1523136138916015, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 8.667911529541016, - "learning_rate": 4.8679410288425e-05, - "loss": 2.2135799407958983, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 16.578901290893555, - "learning_rate": 4.86184966550152e-05, - "loss": 2.1074777603149415, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.302729606628418, - "learning_rate": 4.8556249532052045e-05, - "loss": 2.0359333038330076, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 20.793943405151367, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.2183906555175783, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.577836990356445, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9698509216308593, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 5.63059139251709, - "learning_rate": 4.836154274553069e-05, - "loss": 1.652799415588379, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 6.654581069946289, - "learning_rate": 4.829399755867636e-05, - "loss": 1.740532112121582, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 11.970749855041504, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.6816909790039063, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 7.981494426727295, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.61510009765625, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 6.4518561363220215, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.6320558547973634, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 11.362939834594727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.474619197845459, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 4.846298694610596, - "learning_rate": 4.793662296282393e-05, - "loss": 1.4175973892211915, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 7.837019920349121, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.5572842597961425, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 5.593219757080078, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.439277744293213, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 5.2232160568237305, - "learning_rate": 4.770662587550643e-05, - "loss": 1.3270212173461915, - "step": 7460 - }, - { - "epoch": 2.3485872989391248, - "grad_norm": 6.035870552062988, - "learning_rate": 4.762739032415868e-05, - "loss": 1.3490946769714356, - "step": 7480 - }, - { - "epoch": 2.354868151172262, - "grad_norm": 6.241731643676758, - "learning_rate": 4.754687724066917e-05, - "loss": 1.2901057243347167, - "step": 7500 - }, - { - "epoch": 2.3611490034053997, - "grad_norm": 11.480992317199707, - "learning_rate": 4.746509117076991e-05, - "loss": 1.252013874053955, - "step": 7520 - }, - { - "epoch": 2.367429855638537, - "grad_norm": 3.9125211238861084, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.154312515258789, - "step": 7540 - }, - { - "epoch": 2.373710707871674, - "grad_norm": 5.499811172485352, - "learning_rate": 4.72977186137705e-05, - "loss": 1.0750234603881836, - "step": 7560 - }, - { - "epoch": 2.379991560104812, - "grad_norm": 5.505170822143555, - "learning_rate": 4.721214157644861e-05, - "loss": 1.490649127960205, - "step": 7580 - }, - { - "epoch": 2.386272412337949, - "grad_norm": 5.891989231109619, - "learning_rate": 4.712531045173996e-05, - "loss": 1.1875076293945312, - "step": 7600 - }, - { - "epoch": 2.3925532645710863, - "grad_norm": 5.132697582244873, - "learning_rate": 4.703723014209025e-05, - "loss": 1.0753458976745605, - "step": 7620 - }, - { - "epoch": 2.398834116804224, - "grad_norm": 5.394096851348877, - "learning_rate": 4.694790562047361e-05, - "loss": 1.0024099349975586, - "step": 7640 - }, - { - "epoch": 2.405114969037361, - "grad_norm": 4.041260719299316, - "learning_rate": 4.685734193011182e-05, - "loss": 1.2125065803527832, - "step": 7660 - }, - { - "epoch": 2.4113958212704985, - "grad_norm": 4.972134113311768, - "learning_rate": 4.676554418418953e-05, - "loss": 0.9521866798400879, - "step": 7680 - }, - { - "epoch": 2.417676673503636, - "grad_norm": 10.228482246398926, - "learning_rate": 4.667251756556563e-05, - "loss": 1.0725263595581054, - "step": 7700 - }, - { - "epoch": 2.4239575257367734, - "grad_norm": 3.5744264125823975, - "learning_rate": 4.657826732648061e-05, - "loss": 0.9225356101989746, - "step": 7720 - }, - { - "epoch": 2.4302383779699106, - "grad_norm": 4.8591203689575195, - "learning_rate": 4.648279878825998e-05, - "loss": 0.8432273864746094, - "step": 7740 - }, - { - "epoch": 2.4365192302030483, - "grad_norm": 3.154543161392212, - "learning_rate": 4.63861173410139e-05, - "loss": 0.9964156150817871, - "step": 7760 - }, - { - "epoch": 2.4428000824361855, - "grad_norm": 3.979494571685791, - "learning_rate": 4.628822844333278e-05, - "loss": 0.8577518463134766, - "step": 7780 - }, - { - "epoch": 2.4490809346693228, - "grad_norm": 5.186657428741455, - "learning_rate": 4.618913762197918e-05, - "loss": 0.8347867965698242, - "step": 7800 - }, - { - "epoch": 2.4553617869024604, - "grad_norm": 2.6562411785125732, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.8252105712890625, - "step": 7820 - }, - { - "epoch": 2.4616426391355977, - "grad_norm": 3.5569987297058105, - "learning_rate": 4.59873726542891e-05, - "loss": 1.0673627853393555, - "step": 7840 - }, - { - "epoch": 2.467923491368735, - "grad_norm": 4.606166839599609, - "learning_rate": 4.58847098995107e-05, - "loss": 0.9000311851501465, - "step": 7860 - }, - { - "epoch": 2.4742043436018726, - "grad_norm": 5.916482925415039, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8248819351196289, - "step": 7880 - }, - { - "epoch": 2.48048519583501, - "grad_norm": 2.7883312702178955, - "learning_rate": 4.56758528292216e-05, - "loss": 0.7192736148834229, - "step": 7900 - }, - { - "epoch": 2.486766048068147, - "grad_norm": 7.564154148101807, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.7242763042449951, - "step": 7920 - }, - { - "epoch": 2.4930469003012847, - "grad_norm": 4.136919975280762, - "learning_rate": 4.546232642794249e-05, - "loss": 0.7495751857757569, - "step": 7940 - }, - { - "epoch": 2.499327752534422, - "grad_norm": 5.9343743324279785, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.6677919387817383, - "step": 7960 - }, - { - "epoch": 2.505608604767559, - "grad_norm": 5.331826210021973, - "learning_rate": 4.52441789174058e-05, - "loss": 0.611272144317627, - "step": 7980 - }, - { - "epoch": 2.511889457000697, - "grad_norm": 2.7531578540802, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.7472806930541992, - "step": 8000 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.8789489359467776e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/0/latest.json b/slots/0/latest.json deleted file mode 100644 index 5074bed96e96c42809f792e1cc267638a134978a..0000000000000000000000000000000000000000 --- a/slots/0/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:0", "checkpoint": "slots/0/checkpoint-8009", "step": 8009, "updated_at": 1776417347} diff --git a/slots/1/checkpoint-7862/config.json b/slots/1/checkpoint-7862/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/1/checkpoint-7862/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/1/checkpoint-7862/trainer_state.json b/slots/1/checkpoint-7862/trainer_state.json deleted file mode 100644 index d9a3636b9a21a7693194201b4b01cf8c67338fb5..0000000000000000000000000000000000000000 --- a/slots/1/checkpoint-7862/trainer_state.json +++ /dev/null @@ -1,2792 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.468551576592049, - "eval_steps": 500, - "global_step": 7862, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 35.15928649902344, - "learning_rate": 4.999985708624277e-05, - "loss": 42.4168701171875, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 24.441553115844727, - "learning_rate": 4.999851618030373e-05, - "loss": 34.59369812011719, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 27.142696380615234, - "learning_rate": 4.9995763869538124e-05, - "loss": 29.539852905273438, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 20.90863037109375, - "learning_rate": 4.999160030934015e-05, - "loss": 26.295675659179686, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 19.69276237487793, - "learning_rate": 4.9986025734782515e-05, - "loss": 26.864825439453124, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 20.81488037109375, - "learning_rate": 4.997904046060316e-05, - "loss": 21.519651794433592, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 19.697011947631836, - "learning_rate": 4.997064488118748e-05, - "loss": 20.201951599121095, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 19.01127815246582, - "learning_rate": 4.996083947054607e-05, - "loss": 18.92204895019531, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 21.61073112487793, - "learning_rate": 4.9949624782287944e-05, - "loss": 18.937400817871094, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 23.090789794921875, - "learning_rate": 4.993700144958927e-05, - "loss": 15.520936584472656, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.096433639526367, - "learning_rate": 4.992297018515767e-05, - "loss": 15.487298583984375, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 21.39566421508789, - "learning_rate": 4.990753178119193e-05, - "loss": 14.596461486816406, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 16.463520050048828, - "learning_rate": 4.989068710933732e-05, - "loss": 12.615646362304688, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 14.016276359558105, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.938156890869141, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 20.464670181274414, - "learning_rate": 4.985278284547495e-05, - "loss": 10.395332336425781, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 17.77933692932129, - "learning_rate": 4.983172539352465e-05, - "loss": 10.609977722167969, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 12.832728385925293, - "learning_rate": 4.980926595367956e-05, - "loss": 8.691991424560547, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 13.371121406555176, - "learning_rate": 4.978540579398941e-05, - "loss": 8.057951354980469, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 16.301082611083984, - "learning_rate": 4.9760146261587936e-05, - "loss": 7.973057556152344, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 12.96883487701416, - "learning_rate": 4.973348878261681e-05, - "loss": 7.619200897216797, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 12.284865379333496, - "learning_rate": 4.970543486214516e-05, - "loss": 7.1647804260253904, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 21.732789993286133, - "learning_rate": 4.967598608408454e-05, - "loss": 6.469166564941406, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 10.566523551940918, - "learning_rate": 4.9645144111099515e-05, - "loss": 5.752484130859375, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 11.216104507446289, - "learning_rate": 4.961291068451385e-05, - "loss": 5.699852752685547, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 9.675246238708496, - "learning_rate": 4.9579287624212093e-05, - "loss": 5.267290878295898, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 11.593718528747559, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.2138721466064455, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.523468971252441, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.134304428100586, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 10.710789680480957, - "learning_rate": 4.94701000160798e-05, - "loss": 4.5618743896484375, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.197118759155273, - "learning_rate": 4.943093818728689e-05, - "loss": 4.101953125, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 8.9686279296875, - "learning_rate": 4.939039699886213e-05, - "loss": 3.642832565307617, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 8.167902946472168, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.1743133544921873, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 9.323243141174316, - "learning_rate": 4.930518577661388e-05, - "loss": 3.6948711395263674, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 10.735540390014648, - "learning_rate": 4.926052055377716e-05, - "loss": 3.1577367782592773, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 12.350885391235352, - "learning_rate": 4.921448559301048e-05, - "loss": 3.0701583862304687, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 9.401058197021484, - "learning_rate": 4.916708349342674e-05, - "loss": 2.8100170135498046, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 7.154083728790283, - "learning_rate": 4.911831693132686e-05, - "loss": 2.472774696350098, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 7.130964279174805, - "learning_rate": 4.906818866004871e-05, - "loss": 2.597759246826172, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 7.370577812194824, - "learning_rate": 4.901670150981164e-05, - "loss": 2.3146718978881835, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 9.05713176727295, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1284273147583006, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 7.408658027648926, - "learning_rate": 4.890966227678239e-05, - "loss": 2.052642822265625, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 17.97675895690918, - "learning_rate": 4.885411623737651e-05, - "loss": 2.0922914505004884, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 11.770293235778809, - "learning_rate": 4.87972234054431e-05, - "loss": 1.9368261337280273, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 5.871363162994385, - "learning_rate": 4.87389869931255e-05, - "loss": 2.0074506759643556, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 7.695994853973389, - "learning_rate": 4.8679410288425e-05, - "loss": 1.9740360260009766, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 5.52763557434082, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7269515991210938, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.56577205657959, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.6195665359497071, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 7.687935829162598, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.576361083984375, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.840238094329834, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.704756736755371, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 4.519632816314697, - "learning_rate": 4.836154274553069e-05, - "loss": 2.219845199584961, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 5.06703519821167, - "learning_rate": 4.829399755867636e-05, - "loss": 2.371290397644043, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 5.222914695739746, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.2122323036193847, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 3.4668195247650146, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.334171199798584, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 4.191354274749756, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2463009834289551, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 5.007280349731445, - "learning_rate": 4.801070437462659e-05, - "loss": 1.0995778083801269, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 5.577613830566406, - "learning_rate": 4.793662296282393e-05, - "loss": 1.1267754554748535, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 3.4261252880096436, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.1237424850463866, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 4.648379802703857, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.3997676849365235, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 3.584986448287964, - "learning_rate": 4.770662587550643e-05, - "loss": 0.9382729530334473, - "step": 7460 - }, - { - "epoch": 2.3485872989391248, - "grad_norm": 5.842440605163574, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9716793060302734, - "step": 7480 - }, - { - "epoch": 2.354868151172262, - "grad_norm": 4.669426918029785, - "learning_rate": 4.754687724066917e-05, - "loss": 0.9839717864990234, - "step": 7500 - }, - { - "epoch": 2.3611490034053997, - "grad_norm": 3.096353769302368, - "learning_rate": 4.746509117076991e-05, - "loss": 0.8640604972839355, - "step": 7520 - }, - { - "epoch": 2.367429855638537, - "grad_norm": 4.052355766296387, - "learning_rate": 4.7382036732065104e-05, - "loss": 0.9168565750122071, - "step": 7540 - }, - { - "epoch": 2.373710707871674, - "grad_norm": 3.556760787963867, - "learning_rate": 4.72977186137705e-05, - "loss": 0.8850119590759278, - "step": 7560 - }, - { - "epoch": 2.379991560104812, - "grad_norm": 4.15118408203125, - "learning_rate": 4.721214157644861e-05, - "loss": 0.866028118133545, - "step": 7580 - }, - { - "epoch": 2.386272412337949, - "grad_norm": 5.301335334777832, - "learning_rate": 4.712531045173996e-05, - "loss": 0.7654078483581543, - "step": 7600 - }, - { - "epoch": 2.3925532645710863, - "grad_norm": 5.883398532867432, - "learning_rate": 4.703723014209025e-05, - "loss": 0.7791702270507812, - "step": 7620 - }, - { - "epoch": 2.398834116804224, - "grad_norm": 2.041224241256714, - "learning_rate": 4.694790562047361e-05, - "loss": 0.774465274810791, - "step": 7640 - }, - { - "epoch": 2.405114969037361, - "grad_norm": 3.479743719100952, - "learning_rate": 4.685734193011182e-05, - "loss": 0.7150816440582275, - "step": 7660 - }, - { - "epoch": 2.4113958212704985, - "grad_norm": 3.2080256938934326, - "learning_rate": 4.676554418418953e-05, - "loss": 0.7573625087738037, - "step": 7680 - }, - { - "epoch": 2.417676673503636, - "grad_norm": 2.7411186695098877, - "learning_rate": 4.667251756556563e-05, - "loss": 0.9754419326782227, - "step": 7700 - }, - { - "epoch": 2.4239575257367734, - "grad_norm": 3.309274911880493, - "learning_rate": 4.657826732648061e-05, - "loss": 0.6611151695251465, - "step": 7720 - }, - { - "epoch": 2.4302383779699106, - "grad_norm": 4.727346420288086, - "learning_rate": 4.648279878825998e-05, - "loss": 0.6431382179260254, - "step": 7740 - }, - { - "epoch": 2.4365192302030483, - "grad_norm": 2.4785397052764893, - "learning_rate": 4.63861173410139e-05, - "loss": 0.6248305320739747, - "step": 7760 - }, - { - "epoch": 2.4428000824361855, - "grad_norm": 4.628482818603516, - "learning_rate": 4.628822844333278e-05, - "loss": 0.5700630664825439, - "step": 7780 - }, - { - "epoch": 2.4490809346693228, - "grad_norm": 4.001479625701904, - "learning_rate": 4.618913762197918e-05, - "loss": 0.5712510108947754, - "step": 7800 - }, - { - "epoch": 2.4553617869024604, - "grad_norm": 2.186591625213623, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.6234359741210938, - "step": 7820 - }, - { - "epoch": 2.4616426391355977, - "grad_norm": 2.8685686588287354, - "learning_rate": 4.59873726542891e-05, - "loss": 0.6517808437347412, - "step": 7840 - }, - { - "epoch": 2.467923491368735, - "grad_norm": 3.237225294113159, - "learning_rate": 4.58847098995107e-05, - "loss": 0.5230174541473389, - "step": 7860 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.842398356675507e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/1/latest.json b/slots/1/latest.json deleted file mode 100644 index ab97bdd025d1e1d76cab2173afa0ae4508fe34d7..0000000000000000000000000000000000000000 --- a/slots/1/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:1", "checkpoint": "slots/1/checkpoint-7862", "step": 7862, "updated_at": 1776417247} diff --git a/slots/10/checkpoint-8362/config.json b/slots/10/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/10/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/10/checkpoint-8362/trainer_state.json b/slots/10/checkpoint-8362/trainer_state.json deleted file mode 100644 index 1ff2bbd1579f212c1fa7fb30eb021a9e12c53b41..0000000000000000000000000000000000000000 --- a/slots/10/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.5003729243814035, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 11.95787525177002, - "learning_rate": 4.999985708624277e-05, - "loss": 5.881575012207032, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 11.939610481262207, - "learning_rate": 4.999851618030373e-05, - "loss": 5.647177886962891, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 48.17500305175781, - "learning_rate": 4.9995763869538124e-05, - "loss": 6.1393989562988285, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 14.723703384399414, - "learning_rate": 4.999160030934015e-05, - "loss": 5.649447250366211, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 14.260713577270508, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.456048202514649, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 12.285365104675293, - "learning_rate": 4.997904046060316e-05, - "loss": 5.332961654663086, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 33.52470016479492, - "learning_rate": 4.997064488118748e-05, - "loss": 6.02587776184082, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 17.36154556274414, - "learning_rate": 4.996083947054607e-05, - "loss": 5.352649688720703, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.644362449645996, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.589469528198242, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 14.174250602722168, - "learning_rate": 4.993700144958927e-05, - "loss": 5.2284095764160154, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 14.923873901367188, - "learning_rate": 4.992297018515767e-05, - "loss": 4.571537780761719, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 13.855690956115723, - "learning_rate": 4.990753178119193e-05, - "loss": 5.206478500366211, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 16.607261657714844, - "learning_rate": 4.989068710933732e-05, - "loss": 5.598894882202148, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 10.553526878356934, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.615887832641602, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 9.943771362304688, - "learning_rate": 4.985278284547495e-05, - "loss": 3.9751415252685547, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 11.704524993896484, - "learning_rate": 4.983172539352465e-05, - "loss": 4.089920425415039, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 13.140851020812988, - "learning_rate": 4.980926595367956e-05, - "loss": 4.461263275146484, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.009072303771973, - "learning_rate": 4.978540579398941e-05, - "loss": 3.8523796081542967, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 9.157245635986328, - "learning_rate": 4.9760146261587936e-05, - "loss": 3.737962341308594, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 11.106775283813477, - "learning_rate": 4.973348878261681e-05, - "loss": 4.242771148681641, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.31446361541748, - "learning_rate": 4.970543486214516e-05, - "loss": 3.7330020904541015, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 8.840044975280762, - "learning_rate": 4.967598608408454e-05, - "loss": 3.2959358215332033, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 8.993168830871582, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.447793197631836, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 9.880019187927246, - "learning_rate": 4.961291068451385e-05, - "loss": 3.674238586425781, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 10.820444107055664, - "learning_rate": 4.9579287624212093e-05, - "loss": 3.376015472412109, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 7.960184574127197, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.3296524047851563, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 7.695173740386963, - "learning_rate": 4.9507880274181854e-05, - "loss": 3.172230911254883, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 9.787315368652344, - "learning_rate": 4.94701000160798e-05, - "loss": 3.032790946960449, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 7.822601318359375, - "learning_rate": 4.943093818728689e-05, - "loss": 3.7993682861328124, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 6.423282146453857, - "learning_rate": 4.939039699886213e-05, - "loss": 3.1471582412719727, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 8.177947044372559, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.807335090637207, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 12.727806091308594, - "learning_rate": 4.930518577661388e-05, - "loss": 2.70469970703125, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 12.240203857421875, - "learning_rate": 4.926052055377716e-05, - "loss": 2.772799491882324, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 10.889534950256348, - "learning_rate": 4.921448559301048e-05, - "loss": 2.525356674194336, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 10.73294734954834, - "learning_rate": 4.916708349342674e-05, - "loss": 2.575057792663574, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 6.766477584838867, - "learning_rate": 4.911831693132686e-05, - "loss": 2.450628471374512, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 6.866926193237305, - "learning_rate": 4.906818866004871e-05, - "loss": 2.294377899169922, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 13.464261054992676, - "learning_rate": 4.901670150981164e-05, - "loss": 2.1687923431396485, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 8.798266410827637, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1829912185668947, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.770901679992676, - "learning_rate": 4.890966227678239e-05, - "loss": 1.9910266876220704, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 6.037090301513672, - "learning_rate": 4.885411623737651e-05, - "loss": 2.093680572509766, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 6.327531814575195, - "learning_rate": 4.87972234054431e-05, - "loss": 2.20494384765625, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 5.572488307952881, - "learning_rate": 4.87389869931255e-05, - "loss": 1.9014860153198243, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 6.42570161819458, - "learning_rate": 4.8679410288425e-05, - "loss": 1.685124397277832, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 5.767203330993652, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7935949325561524, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 11.936880111694336, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.7972061157226562, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.11804723739624, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.7099994659423827, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 4.457609176635742, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.8403182983398438, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 6.604846954345703, - "learning_rate": 4.836154274553069e-05, - "loss": 1.799737548828125, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 4.532509803771973, - "learning_rate": 4.829399755867636e-05, - "loss": 1.677235984802246, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 4.908101558685303, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.6233207702636718, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 4.57976770401001, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.4081362724304198, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.677663326263428, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.475815200805664, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 5.180493354797363, - "learning_rate": 4.801070437462659e-05, - "loss": 1.256404972076416, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 5.710392475128174, - "learning_rate": 4.793662296282393e-05, - "loss": 1.344097900390625, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 14.397268295288086, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.3774223327636719, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 5.110644340515137, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.3270224571228026, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 7.093288898468018, - "learning_rate": 4.770662587550643e-05, - "loss": 1.4591200828552247, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 4.576109409332275, - "learning_rate": 4.762739032415868e-05, - "loss": 1.40781831741333, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 8.339225769042969, - "learning_rate": 4.754687724066917e-05, - "loss": 1.479560661315918, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 7.494958877563477, - "learning_rate": 4.746509117076991e-05, - "loss": 1.3163747787475586, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 6.01884651184082, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.2254773139953614, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 7.199056625366211, - "learning_rate": 4.72977186137705e-05, - "loss": 1.0894113540649415, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 8.82214069366455, - "learning_rate": 4.721214157644861e-05, - "loss": 1.3237985610961913, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 4.850705146789551, - "learning_rate": 4.712531045173996e-05, - "loss": 0.8934127807617187, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 5.598201751708984, - "learning_rate": 4.703723014209025e-05, - "loss": 1.342045211791992, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 5.424136161804199, - "learning_rate": 4.694790562047361e-05, - "loss": 1.1706354141235351, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 7.543458461761475, - "learning_rate": 4.685734193011182e-05, - "loss": 0.9434962272644043, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 2.857658624649048, - "learning_rate": 4.676554418418953e-05, - "loss": 0.989952278137207, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 6.373669147491455, - "learning_rate": 4.667251756556563e-05, - "loss": 0.8908876419067383, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 6.381263256072998, - "learning_rate": 4.657826732648061e-05, - "loss": 1.0061418533325195, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.1551015377044678, - "learning_rate": 4.648279878825998e-05, - "loss": 0.7668521881103516, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.2378339767456055, - "learning_rate": 4.63861173410139e-05, - "loss": 0.9835851669311524, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.329948425292969, - "learning_rate": 4.628822844333278e-05, - "loss": 0.9755271911621094, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 3.7822136878967285, - "learning_rate": 4.618913762197918e-05, - "loss": 0.9501211166381835, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.9014272689819336, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.7127848625183105, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 5.8602190017700195, - "learning_rate": 4.59873726542891e-05, - "loss": 0.9189926147460937, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 2.9552366733551025, - "learning_rate": 4.58847098995107e-05, - "loss": 0.8160135269165039, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 4.013094425201416, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8563443183898926, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 2.5212459564208984, - "learning_rate": 4.56758528292216e-05, - "loss": 0.6215355873107911, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 4.694630146026611, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.6466864585876465, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 2.3018579483032227, - "learning_rate": 4.546232642794249e-05, - "loss": 0.9325425148010253, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 2.284792184829712, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.8970767021179199, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 3.947718381881714, - "learning_rate": 4.52441789174058e-05, - "loss": 0.7865421295166015, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 4.801159858703613, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.8510586738586425, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 4.125448226928711, - "learning_rate": 4.502145956295211e-05, - "loss": 0.7897457599639892, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 5.959787845611572, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.8026494026184082, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 2.8297674655914307, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.7088653564453125, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 1.5840117931365967, - "learning_rate": 4.46789186299435e-05, - "loss": 0.5798449993133545, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 3.8085098266601562, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.7292616367340088, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 2.9961509704589844, - "learning_rate": 4.444499194921691e-05, - "loss": 0.6142529487609864, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 5.169718265533447, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6897237777709961, - "step": 8140 - }, - { - "epoch": 3.415791057665886, - "grad_norm": 3.65032958984375, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.6636746883392334, - "step": 8160 - }, - { - "epoch": 3.424165499914947, - "grad_norm": 3.502251386642456, - "learning_rate": 4.408588490950914e-05, - "loss": 0.5737998962402344, - "step": 8180 - }, - { - "epoch": 3.4325399421640084, - "grad_norm": 4.158402919769287, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.5830648422241211, - "step": 8200 - }, - { - "epoch": 3.440914384413069, - "grad_norm": 4.869093418121338, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.6292632579803467, - "step": 8220 - }, - { - "epoch": 3.4492888266621304, - "grad_norm": 3.0105438232421875, - "learning_rate": 4.371708001159045e-05, - "loss": 0.6642576694488526, - "step": 8240 - }, - { - "epoch": 3.4576632689111917, - "grad_norm": 1.6431137323379517, - "learning_rate": 4.359202221530606e-05, - "loss": 0.6456145286560059, - "step": 8260 - }, - { - "epoch": 3.466037711160253, - "grad_norm": 4.555006980895996, - "learning_rate": 4.34659147219236e-05, - "loss": 0.5698325634002686, - "step": 8280 - }, - { - "epoch": 3.4744121534093138, - "grad_norm": 2.3838744163513184, - "learning_rate": 4.333876465141465e-05, - "loss": 0.4672196388244629, - "step": 8300 - }, - { - "epoch": 3.482786595658375, - "grad_norm": 1.8454971313476562, - "learning_rate": 4.321057918261423e-05, - "loss": 0.64217209815979, - "step": 8320 - }, - { - "epoch": 3.4911610379074363, - "grad_norm": 2.3971400260925293, - "learning_rate": 4.3081365552815465e-05, - "loss": 0.5033417701721191, - "step": 8340 - }, - { - "epoch": 3.4995354801564975, - "grad_norm": 3.0280821323394775, - "learning_rate": 4.2951131057360984e-05, - "loss": 0.6511260986328125, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966704514664832e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/10/latest.json b/slots/10/latest.json deleted file mode 100644 index 37a64400c64bedcc0bbbcd114ef40d2cd469ade8..0000000000000000000000000000000000000000 --- a/slots/10/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:10", "checkpoint": "slots/10/checkpoint-8362", "step": 8362, "updated_at": 1776417881} diff --git a/slots/11/latest.json b/slots/11/latest.json deleted file mode 100644 index a6661c4c45273f89bd2e7ec4261ef95d7102a6b8..0000000000000000000000000000000000000000 --- a/slots/11/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:11", "checkpoint": "slots/11/checkpoint-7246", "step": 7246, "updated_at": 1776415939} diff --git a/slots/12/checkpoint-8362/config.json b/slots/12/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/12/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/12/checkpoint-8362/trainer_state.json b/slots/12/checkpoint-8362/trainer_state.json deleted file mode 100644 index 68f4e535a59a32a08eacd0aec847efc0ccdc46d8..0000000000000000000000000000000000000000 --- a/slots/12/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.5003729243814035, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.268299102783203, - "learning_rate": 4.999985708624277e-05, - "loss": 4.392966461181641, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 23.723892211914062, - "learning_rate": 4.999851618030373e-05, - "loss": 5.242445755004883, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 15.284976959228516, - "learning_rate": 4.9995763869538124e-05, - "loss": 4.878763198852539, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 21.47739601135254, - "learning_rate": 4.999160030934015e-05, - "loss": 5.070553588867187, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.51418399810791, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.079425048828125, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 17.037799835205078, - "learning_rate": 4.997904046060316e-05, - "loss": 5.397942352294922, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 11.956343650817871, - "learning_rate": 4.997064488118748e-05, - "loss": 5.866295242309571, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 15.19040298461914, - "learning_rate": 4.996083947054607e-05, - "loss": 6.805714416503906, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 12.783180236816406, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.310309219360351, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 15.791704177856445, - "learning_rate": 4.993700144958927e-05, - "loss": 5.249310302734375, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 15.071763038635254, - "learning_rate": 4.992297018515767e-05, - "loss": 5.418260955810547, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 19.809253692626953, - "learning_rate": 4.990753178119193e-05, - "loss": 4.681427383422852, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 12.614968299865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.647319412231445, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.064179420471191, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.246531677246094, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 15.25048828125, - "learning_rate": 4.985278284547495e-05, - "loss": 4.512583541870117, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 14.296823501586914, - "learning_rate": 4.983172539352465e-05, - "loss": 4.193087005615235, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 10.55685043334961, - "learning_rate": 4.980926595367956e-05, - "loss": 4.492794799804687, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 15.188244819641113, - "learning_rate": 4.978540579398941e-05, - "loss": 4.525092697143554, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 23.44154167175293, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.181382751464843, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 11.647577285766602, - "learning_rate": 4.973348878261681e-05, - "loss": 4.1204273223876955, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.451223373413086, - "learning_rate": 4.970543486214516e-05, - "loss": 3.5496986389160154, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 15.924224853515625, - "learning_rate": 4.967598608408454e-05, - "loss": 3.6359764099121095, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 14.668912887573242, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.783356857299805, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 13.026039123535156, - "learning_rate": 4.961291068451385e-05, - "loss": 3.3328643798828126, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 14.517637252807617, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.825040817260742, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 16.220218658447266, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.251118469238281, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 10.804553985595703, - "learning_rate": 4.9507880274181854e-05, - "loss": 3.2774829864501953, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 16.004716873168945, - "learning_rate": 4.94701000160798e-05, - "loss": 3.2009490966796874, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 8.181561470031738, - "learning_rate": 4.943093818728689e-05, - "loss": 2.7726932525634767, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 10.360410690307617, - "learning_rate": 4.939039699886213e-05, - "loss": 3.2188484191894533, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.99072265625, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.66240234375, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 16.23861312866211, - "learning_rate": 4.930518577661388e-05, - "loss": 2.855569839477539, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 9.361970901489258, - "learning_rate": 4.926052055377716e-05, - "loss": 2.8716863632202148, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 10.665338516235352, - "learning_rate": 4.921448559301048e-05, - "loss": 2.866599464416504, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 6.958986759185791, - "learning_rate": 4.916708349342674e-05, - "loss": 2.719050407409668, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 15.789422035217285, - "learning_rate": 4.911831693132686e-05, - "loss": 2.807525062561035, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 11.294109344482422, - "learning_rate": 4.906818866004871e-05, - "loss": 3.0798269271850587, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 9.417093276977539, - "learning_rate": 4.901670150981164e-05, - "loss": 2.639542579650879, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 9.54472827911377, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.214591217041016, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.778777599334717, - "learning_rate": 4.890966227678239e-05, - "loss": 2.497542381286621, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 18.648605346679688, - "learning_rate": 4.885411623737651e-05, - "loss": 2.118986701965332, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 8.938231468200684, - "learning_rate": 4.87972234054431e-05, - "loss": 2.4246755599975587, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 14.108246803283691, - "learning_rate": 4.87389869931255e-05, - "loss": 2.476232147216797, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 20.093215942382812, - "learning_rate": 4.8679410288425e-05, - "loss": 2.005863571166992, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 6.399996280670166, - "learning_rate": 4.86184966550152e-05, - "loss": 2.176909637451172, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 6.5276103019714355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.833466148376465, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 7.747783660888672, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.028280258178711, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 8.722662925720215, - "learning_rate": 4.8427768950332264e-05, - "loss": 2.0328079223632813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 7.213231563568115, - "learning_rate": 4.836154274553069e-05, - "loss": 1.7321701049804688, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 8.361398696899414, - "learning_rate": 4.829399755867636e-05, - "loss": 1.7763862609863281, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 4.578886032104492, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.8702903747558595, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 7.65287446975708, - "learning_rate": 4.8154965567345073e-05, - "loss": 2.1905149459838866, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.528337478637695, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.5154589653015136, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.492247581481934, - "learning_rate": 4.801070437462659e-05, - "loss": 1.6617794036865234, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 7.519628524780273, - "learning_rate": 4.793662296282393e-05, - "loss": 1.5524075508117676, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.682668685913086, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.613736915588379, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 8.832022666931152, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.5550686836242675, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.3445448875427246, - "learning_rate": 4.770662587550643e-05, - "loss": 1.4483519554138184, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 8.750604629516602, - "learning_rate": 4.762739032415868e-05, - "loss": 1.5183201789855958, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.896105766296387, - "learning_rate": 4.754687724066917e-05, - "loss": 1.298677349090576, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 6.540890693664551, - "learning_rate": 4.746509117076991e-05, - "loss": 1.3707378387451172, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 5.634416580200195, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.2700369834899903, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 5.313226699829102, - "learning_rate": 4.72977186137705e-05, - "loss": 1.3874421119689941, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 4.164125919342041, - "learning_rate": 4.721214157644861e-05, - "loss": 1.3053173065185546, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 3.9634926319122314, - "learning_rate": 4.712531045173996e-05, - "loss": 1.1017135620117187, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 19.879589080810547, - "learning_rate": 4.703723014209025e-05, - "loss": 1.5074621200561524, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 7.186710834503174, - "learning_rate": 4.694790562047361e-05, - "loss": 1.3602344512939453, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 6.228899955749512, - "learning_rate": 4.685734193011182e-05, - "loss": 1.097489070892334, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 4.770694255828857, - "learning_rate": 4.676554418418953e-05, - "loss": 1.1659790992736816, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 4.967444896697998, - "learning_rate": 4.667251756556563e-05, - "loss": 1.1331518173217774, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.312889099121094, - "learning_rate": 4.657826732648061e-05, - "loss": 1.2721074104309082, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.759788751602173, - "learning_rate": 4.648279878825998e-05, - "loss": 0.9295409202575684, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 7.215264797210693, - "learning_rate": 4.63861173410139e-05, - "loss": 1.196054744720459, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673015117645264, - "learning_rate": 4.628822844333278e-05, - "loss": 1.0879426002502441, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 6.737372875213623, - "learning_rate": 4.618913762197918e-05, - "loss": 1.0760483741760254, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 3.58354115486145, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.9610136985778809, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 3.435591220855713, - "learning_rate": 4.59873726542891e-05, - "loss": 1.0568583488464356, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 16.85455322265625, - "learning_rate": 4.58847098995107e-05, - "loss": 1.020804214477539, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 3.1317970752716064, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8810656547546387, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 5.6492133140563965, - "learning_rate": 4.56758528292216e-05, - "loss": 0.834989356994629, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 5.5128045082092285, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.8088371276855468, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 3.422081232070923, - "learning_rate": 4.546232642794249e-05, - "loss": 1.0146251678466798, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 4.3992838859558105, - "learning_rate": 4.5353827256577894e-05, - "loss": 1.0001090049743653, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 10.456899642944336, - "learning_rate": 4.52441789174058e-05, - "loss": 0.9053573608398438, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 2.4118423461914062, - "learning_rate": 4.5133387601121476e-05, - "loss": 1.0744664192199707, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 5.180037021636963, - "learning_rate": 4.502145956295211e-05, - "loss": 0.8649028778076172, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 3.958254337310791, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.9063331604003906, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 4.954156398773193, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.7377657413482666, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 3.4824178218841553, - "learning_rate": 4.46789186299435e-05, - "loss": 0.7941797733306885, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 4.906705379486084, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.7057065010070801, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 5.808215141296387, - "learning_rate": 4.444499194921691e-05, - "loss": 0.8522198677062989, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 5.282096862792969, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6269611835479736, - "step": 8140 - }, - { - "epoch": 3.415791057665886, - "grad_norm": 3.557736396789551, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.6752423286437989, - "step": 8160 - }, - { - "epoch": 3.424165499914947, - "grad_norm": 3.3064231872558594, - "learning_rate": 4.408588490950914e-05, - "loss": 0.7134774684906006, - "step": 8180 - }, - { - "epoch": 3.4325399421640084, - "grad_norm": 5.352931499481201, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.862491226196289, - "step": 8200 - }, - { - "epoch": 3.440914384413069, - "grad_norm": 5.765937805175781, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.6733663558959961, - "step": 8220 - }, - { - "epoch": 3.4492888266621304, - "grad_norm": 5.167966365814209, - "learning_rate": 4.371708001159045e-05, - "loss": 0.7372004508972168, - "step": 8240 - }, - { - "epoch": 3.4576632689111917, - "grad_norm": 2.806338310241699, - "learning_rate": 4.359202221530606e-05, - "loss": 0.5634305477142334, - "step": 8260 - }, - { - "epoch": 3.466037711160253, - "grad_norm": 2.7135820388793945, - "learning_rate": 4.34659147219236e-05, - "loss": 0.6759109020233154, - "step": 8280 - }, - { - "epoch": 3.4744121534093138, - "grad_norm": 3.20377516746521, - "learning_rate": 4.333876465141465e-05, - "loss": 0.8337508201599121, - "step": 8300 - }, - { - "epoch": 3.482786595658375, - "grad_norm": 4.656363010406494, - "learning_rate": 4.321057918261423e-05, - "loss": 0.6865992069244384, - "step": 8320 - }, - { - "epoch": 3.4911610379074363, - "grad_norm": 4.720609188079834, - "learning_rate": 4.3081365552815465e-05, - "loss": 0.7185826301574707, - "step": 8340 - }, - { - "epoch": 3.4995354801564975, - "grad_norm": 2.037018060684204, - "learning_rate": 4.2951131057360984e-05, - "loss": 0.8600939750671387, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966704514664832e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/12/latest.json b/slots/12/latest.json deleted file mode 100644 index ba8ffbad806b5d4c46e1c468cca3edc247166d6c..0000000000000000000000000000000000000000 --- a/slots/12/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:12", "checkpoint": "slots/12/checkpoint-8362", "step": 8362, "updated_at": 1776417985} diff --git a/slots/13/checkpoint-7861/config.json b/slots/13/checkpoint-7861/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/13/checkpoint-7861/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/13/checkpoint-7861/trainer_state.json b/slots/13/checkpoint-7861/trainer_state.json deleted file mode 100644 index 245dbec01aced111d0449873743f7f221b87d5a7..0000000000000000000000000000000000000000 --- a/slots/13/checkpoint-7861/trainer_state.json +++ /dev/null @@ -1,2792 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.468237533980392, - "eval_steps": 500, - "global_step": 7861, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 35.15928649902344, - "learning_rate": 4.999985708624277e-05, - "loss": 42.4168701171875, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 24.441553115844727, - "learning_rate": 4.999851618030373e-05, - "loss": 34.59369812011719, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 27.142696380615234, - "learning_rate": 4.9995763869538124e-05, - "loss": 29.539852905273438, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 20.90863037109375, - "learning_rate": 4.999160030934015e-05, - "loss": 26.295675659179686, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 19.69276237487793, - "learning_rate": 4.9986025734782515e-05, - "loss": 26.864825439453124, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 20.81488037109375, - "learning_rate": 4.997904046060316e-05, - "loss": 21.519651794433592, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 19.697011947631836, - "learning_rate": 4.997064488118748e-05, - "loss": 20.201951599121095, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 19.01127815246582, - "learning_rate": 4.996083947054607e-05, - "loss": 18.92204895019531, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 21.61073112487793, - "learning_rate": 4.9949624782287944e-05, - "loss": 18.937400817871094, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 23.090789794921875, - "learning_rate": 4.993700144958927e-05, - "loss": 15.520936584472656, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.096433639526367, - "learning_rate": 4.992297018515767e-05, - "loss": 15.487298583984375, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 21.39566421508789, - "learning_rate": 4.990753178119193e-05, - "loss": 14.596461486816406, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 16.463520050048828, - "learning_rate": 4.989068710933732e-05, - "loss": 12.615646362304688, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 14.016276359558105, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.938156890869141, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 20.464670181274414, - "learning_rate": 4.985278284547495e-05, - "loss": 10.395332336425781, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 17.77933692932129, - "learning_rate": 4.983172539352465e-05, - "loss": 10.609977722167969, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 12.832728385925293, - "learning_rate": 4.980926595367956e-05, - "loss": 8.691991424560547, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 13.371121406555176, - "learning_rate": 4.978540579398941e-05, - "loss": 8.057951354980469, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 16.301082611083984, - "learning_rate": 4.9760146261587936e-05, - "loss": 7.973057556152344, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 12.96883487701416, - "learning_rate": 4.973348878261681e-05, - "loss": 7.619200897216797, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 12.284865379333496, - "learning_rate": 4.970543486214516e-05, - "loss": 7.1647804260253904, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 21.732789993286133, - "learning_rate": 4.967598608408454e-05, - "loss": 6.469166564941406, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 10.566523551940918, - "learning_rate": 4.9645144111099515e-05, - "loss": 5.752484130859375, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 11.216104507446289, - "learning_rate": 4.961291068451385e-05, - "loss": 5.699852752685547, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 9.675246238708496, - "learning_rate": 4.9579287624212093e-05, - "loss": 5.267290878295898, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 11.593718528747559, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.2138721466064455, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.523468971252441, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.134304428100586, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 10.710789680480957, - "learning_rate": 4.94701000160798e-05, - "loss": 4.5618743896484375, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.197118759155273, - "learning_rate": 4.943093818728689e-05, - "loss": 4.101953125, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 8.9686279296875, - "learning_rate": 4.939039699886213e-05, - "loss": 3.642832565307617, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 8.167902946472168, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.1743133544921873, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 9.323243141174316, - "learning_rate": 4.930518577661388e-05, - "loss": 3.6948711395263674, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 10.735540390014648, - "learning_rate": 4.926052055377716e-05, - "loss": 3.1577367782592773, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 12.350885391235352, - "learning_rate": 4.921448559301048e-05, - "loss": 3.0701583862304687, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 9.401058197021484, - "learning_rate": 4.916708349342674e-05, - "loss": 2.8100170135498046, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 7.154083728790283, - "learning_rate": 4.911831693132686e-05, - "loss": 2.472774696350098, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 7.130964279174805, - "learning_rate": 4.906818866004871e-05, - "loss": 2.597759246826172, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 7.370577812194824, - "learning_rate": 4.901670150981164e-05, - "loss": 2.3146718978881835, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 9.05713176727295, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1284273147583006, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 7.408658027648926, - "learning_rate": 4.890966227678239e-05, - "loss": 2.052642822265625, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 17.97675895690918, - "learning_rate": 4.885411623737651e-05, - "loss": 2.0922914505004884, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 11.770293235778809, - "learning_rate": 4.87972234054431e-05, - "loss": 1.9368261337280273, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 5.871363162994385, - "learning_rate": 4.87389869931255e-05, - "loss": 2.0074506759643556, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 7.695994853973389, - "learning_rate": 4.8679410288425e-05, - "loss": 1.9740360260009766, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 5.52763557434082, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7269515991210938, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.56577205657959, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.6195665359497071, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 7.687935829162598, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.576361083984375, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.840238094329834, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.704756736755371, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 4.519632816314697, - "learning_rate": 4.836154274553069e-05, - "loss": 2.219845199584961, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 5.06703519821167, - "learning_rate": 4.829399755867636e-05, - "loss": 2.371290397644043, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 5.222914695739746, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.2122323036193847, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 3.4668195247650146, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.334171199798584, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 4.191354274749756, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2463009834289551, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 5.007280349731445, - "learning_rate": 4.801070437462659e-05, - "loss": 1.0995778083801269, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 5.577613830566406, - "learning_rate": 4.793662296282393e-05, - "loss": 1.1267754554748535, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 3.4261252880096436, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.1237424850463866, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 4.648379802703857, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.3997676849365235, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 3.584986448287964, - "learning_rate": 4.770662587550643e-05, - "loss": 0.9382729530334473, - "step": 7460 - }, - { - "epoch": 2.3485872989391248, - "grad_norm": 5.842440605163574, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9716793060302734, - "step": 7480 - }, - { - "epoch": 2.354868151172262, - "grad_norm": 4.669426918029785, - "learning_rate": 4.754687724066917e-05, - "loss": 0.9839717864990234, - "step": 7500 - }, - { - "epoch": 2.3611490034053997, - "grad_norm": 3.096353769302368, - "learning_rate": 4.746509117076991e-05, - "loss": 0.8640604972839355, - "step": 7520 - }, - { - "epoch": 2.367429855638537, - "grad_norm": 4.052355766296387, - "learning_rate": 4.7382036732065104e-05, - "loss": 0.9168565750122071, - "step": 7540 - }, - { - "epoch": 2.373710707871674, - "grad_norm": 3.556760787963867, - "learning_rate": 4.72977186137705e-05, - "loss": 0.8850119590759278, - "step": 7560 - }, - { - "epoch": 2.379991560104812, - "grad_norm": 4.15118408203125, - "learning_rate": 4.721214157644861e-05, - "loss": 0.866028118133545, - "step": 7580 - }, - { - "epoch": 2.386272412337949, - "grad_norm": 5.301335334777832, - "learning_rate": 4.712531045173996e-05, - "loss": 0.7654078483581543, - "step": 7600 - }, - { - "epoch": 2.3925532645710863, - "grad_norm": 5.883398532867432, - "learning_rate": 4.703723014209025e-05, - "loss": 0.7791702270507812, - "step": 7620 - }, - { - "epoch": 2.398834116804224, - "grad_norm": 2.041224241256714, - "learning_rate": 4.694790562047361e-05, - "loss": 0.774465274810791, - "step": 7640 - }, - { - "epoch": 2.405114969037361, - "grad_norm": 3.479743719100952, - "learning_rate": 4.685734193011182e-05, - "loss": 0.7150816440582275, - "step": 7660 - }, - { - "epoch": 2.4113958212704985, - "grad_norm": 3.2080256938934326, - "learning_rate": 4.676554418418953e-05, - "loss": 0.7573625087738037, - "step": 7680 - }, - { - "epoch": 2.417676673503636, - "grad_norm": 2.7411186695098877, - "learning_rate": 4.667251756556563e-05, - "loss": 0.9754419326782227, - "step": 7700 - }, - { - "epoch": 2.4239575257367734, - "grad_norm": 3.309274911880493, - "learning_rate": 4.657826732648061e-05, - "loss": 0.6611151695251465, - "step": 7720 - }, - { - "epoch": 2.4302383779699106, - "grad_norm": 4.727346420288086, - "learning_rate": 4.648279878825998e-05, - "loss": 0.6431382179260254, - "step": 7740 - }, - { - "epoch": 2.4365192302030483, - "grad_norm": 2.4785397052764893, - "learning_rate": 4.63861173410139e-05, - "loss": 0.6248305320739747, - "step": 7760 - }, - { - "epoch": 2.4428000824361855, - "grad_norm": 4.628482818603516, - "learning_rate": 4.628822844333278e-05, - "loss": 0.5700630664825439, - "step": 7780 - }, - { - "epoch": 2.4490809346693228, - "grad_norm": 4.001479625701904, - "learning_rate": 4.618913762197918e-05, - "loss": 0.5712510108947754, - "step": 7800 - }, - { - "epoch": 2.4553617869024604, - "grad_norm": 2.186591625213623, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.6234359741210938, - "step": 7820 - }, - { - "epoch": 2.4616426391355977, - "grad_norm": 2.8685686588287354, - "learning_rate": 4.59873726542891e-05, - "loss": 0.6517808437347412, - "step": 7840 - }, - { - "epoch": 2.467923491368735, - "grad_norm": 3.237225294113159, - "learning_rate": 4.58847098995107e-05, - "loss": 0.5230174541473389, - "step": 7860 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.842149713279104e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/13/latest.json b/slots/13/latest.json deleted file mode 100644 index 589753a0750210b189f9878dad738aeef0ef80d4..0000000000000000000000000000000000000000 --- a/slots/13/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:13", "checkpoint": "slots/13/checkpoint-7861", "step": 7861, "updated_at": 1776417208} diff --git a/slots/14/latest.json b/slots/14/latest.json deleted file mode 100644 index a1d4b1903cc8c7f9ac37036d32cba25f244a0818..0000000000000000000000000000000000000000 --- a/slots/14/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:14", "checkpoint": "slots/14/checkpoint-7364", "step": 7364, "updated_at": 1776416329} diff --git a/slots/15/checkpoint-8362/config.json b/slots/15/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/15/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/15/checkpoint-8362/trainer_state.json b/slots/15/checkpoint-8362/trainer_state.json deleted file mode 100644 index 002140bbf0e542892d8157313acfb0a9e7c34bb9..0000000000000000000000000000000000000000 --- a/slots/15/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.5003729243814035, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.981147766113281, - "learning_rate": 4.999985708624277e-05, - "loss": 6.141958999633789, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.432202339172363, - "learning_rate": 4.999851618030373e-05, - "loss": 6.936581420898437, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 14.735859870910645, - "learning_rate": 4.9995763869538124e-05, - "loss": 7.491865539550782, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 12.628633499145508, - "learning_rate": 4.999160030934015e-05, - "loss": 7.847127532958984, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 14.50905990600586, - "learning_rate": 4.9986025734782515e-05, - "loss": 6.550693511962891, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 18.310335159301758, - "learning_rate": 4.997904046060316e-05, - "loss": 6.068325042724609, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 12.4039888381958, - "learning_rate": 4.997064488118748e-05, - "loss": 5.966919708251953, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 14.189772605895996, - "learning_rate": 4.996083947054607e-05, - "loss": 5.816884613037109, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.678384780883789, - "learning_rate": 4.9949624782287944e-05, - "loss": 6.372280120849609, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 13.517382621765137, - "learning_rate": 4.993700144958927e-05, - "loss": 5.818735504150391, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 11.332883834838867, - "learning_rate": 4.992297018515767e-05, - "loss": 5.053961181640625, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 29.601367950439453, - "learning_rate": 4.990753178119193e-05, - "loss": 5.2946617126464846, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.445046424865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.721809005737304, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 14.08784294128418, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.68652572631836, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 10.382508277893066, - "learning_rate": 4.985278284547495e-05, - "loss": 4.90959358215332, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 9.340018272399902, - "learning_rate": 4.983172539352465e-05, - "loss": 4.94355354309082, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 11.943198204040527, - "learning_rate": 4.980926595367956e-05, - "loss": 4.442295074462891, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.623554229736328, - "learning_rate": 4.978540579398941e-05, - "loss": 4.63629150390625, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 9.113373756408691, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.220952987670898, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.733516693115234, - "learning_rate": 4.973348878261681e-05, - "loss": 4.143902206420899, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.420555114746094, - "learning_rate": 4.970543486214516e-05, - "loss": 4.141384506225586, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 7.3338212966918945, - "learning_rate": 4.967598608408454e-05, - "loss": 3.72052001953125, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 10.468330383300781, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.4560214996337892, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 10.967416763305664, - "learning_rate": 4.961291068451385e-05, - "loss": 3.2705413818359377, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 12.951521873474121, - "learning_rate": 4.9579287624212093e-05, - "loss": 3.3012428283691406, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 9.387336730957031, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.2594974517822264, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.693846702575684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.9371822357177733, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 11.925010681152344, - "learning_rate": 4.94701000160798e-05, - "loss": 2.991471290588379, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 9.22618579864502, - "learning_rate": 4.943093818728689e-05, - "loss": 3.6338191986083985, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 9.440751075744629, - "learning_rate": 4.939039699886213e-05, - "loss": 2.745065689086914, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.722249984741211, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.652340126037598, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 10.900734901428223, - "learning_rate": 4.930518577661388e-05, - "loss": 2.8322593688964846, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 6.965799808502197, - "learning_rate": 4.926052055377716e-05, - "loss": 2.7872573852539064, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 6.74927282333374, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9263822555541994, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 8.862627029418945, - "learning_rate": 4.916708349342674e-05, - "loss": 2.499791717529297, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 10.037351608276367, - "learning_rate": 4.911831693132686e-05, - "loss": 2.275549125671387, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 6.2945075035095215, - "learning_rate": 4.906818866004871e-05, - "loss": 2.404999542236328, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 4.414289951324463, - "learning_rate": 4.901670150981164e-05, - "loss": 2.1930471420288087, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 5.771992206573486, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.3143028259277343, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.430806636810303, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1435062408447267, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 4.818403244018555, - "learning_rate": 4.885411623737651e-05, - "loss": 1.8850128173828125, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 6.249518394470215, - "learning_rate": 4.87972234054431e-05, - "loss": 2.0096624374389647, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.708676815032959, - "learning_rate": 4.87389869931255e-05, - "loss": 1.9306636810302735, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 10.92924690246582, - "learning_rate": 4.8679410288425e-05, - "loss": 1.60665283203125, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 8.802605628967285, - "learning_rate": 4.86184966550152e-05, - "loss": 1.6958501815795899, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 11.568385124206543, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.7447893142700195, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.640422344207764, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.7216526031494142, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.1430816650390625, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9419631958007812, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 4.505365371704102, - "learning_rate": 4.836154274553069e-05, - "loss": 1.4452963829040528, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 4.836411952972412, - "learning_rate": 4.829399755867636e-05, - "loss": 1.5338068962097169, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 6.093108177185059, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.553403663635254, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 6.984614372253418, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.5232481956481934, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.966179370880127, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.9450897216796874, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.660058975219727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4149410247802734, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.126588821411133, - "learning_rate": 4.793662296282393e-05, - "loss": 1.0894225120544434, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 9.536477088928223, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.519902515411377, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 5.378171443939209, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1999958038330079, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 4.443706512451172, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2191826820373535, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 4.879043102264404, - "learning_rate": 4.762739032415868e-05, - "loss": 1.2340889930725099, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.470728874206543, - "learning_rate": 4.754687724066917e-05, - "loss": 1.423619270324707, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 5.47915506362915, - "learning_rate": 4.746509117076991e-05, - "loss": 1.251408863067627, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 4.566497325897217, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.0826839447021483, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 6.905686855316162, - "learning_rate": 4.72977186137705e-05, - "loss": 1.069216251373291, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 2.964932680130005, - "learning_rate": 4.721214157644861e-05, - "loss": 1.1391191482543945, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 6.019253253936768, - "learning_rate": 4.712531045173996e-05, - "loss": 0.9639101028442383, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 3.0169851779937744, - "learning_rate": 4.703723014209025e-05, - "loss": 0.934177017211914, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 5.5374369621276855, - "learning_rate": 4.694790562047361e-05, - "loss": 0.921787166595459, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 3.7313361167907715, - "learning_rate": 4.685734193011182e-05, - "loss": 0.9222763061523438, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 3.172354221343994, - "learning_rate": 4.676554418418953e-05, - "loss": 0.9454127311706543, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 2.3172996044158936, - "learning_rate": 4.667251756556563e-05, - "loss": 1.050184154510498, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.589139938354492, - "learning_rate": 4.657826732648061e-05, - "loss": 0.7619458675384522, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 4.144975662231445, - "learning_rate": 4.648279878825998e-05, - "loss": 0.8281337738037109, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.205533266067505, - "learning_rate": 4.63861173410139e-05, - "loss": 0.8384788513183594, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673986911773682, - "learning_rate": 4.628822844333278e-05, - "loss": 0.7749058723449707, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 4.221221446990967, - "learning_rate": 4.618913762197918e-05, - "loss": 1.0583919525146483, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.1249191761016846, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.6937400341033936, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 3.2157654762268066, - "learning_rate": 4.59873726542891e-05, - "loss": 0.7503646850585938, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 2.9630126953125, - "learning_rate": 4.58847098995107e-05, - "loss": 0.7044057846069336, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 2.9990007877349854, - "learning_rate": 4.578086800353284e-05, - "loss": 0.6530625343322753, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 2.869370698928833, - "learning_rate": 4.56758528292216e-05, - "loss": 0.6877331733703613, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 2.248493194580078, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.7664458751678467, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 2.4392411708831787, - "learning_rate": 4.546232642794249e-05, - "loss": 0.6721152305603028, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 1.9127728939056396, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.6500507354736328, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 2.9150123596191406, - "learning_rate": 4.52441789174058e-05, - "loss": 0.6752653598785401, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 4.481291770935059, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.6834932327270508, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 1.990797519683838, - "learning_rate": 4.502145956295211e-05, - "loss": 0.6353804588317871, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 6.725353717803955, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.6204021453857422, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 3.649500846862793, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.6903295516967773, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 3.0070176124572754, - "learning_rate": 4.46789186299435e-05, - "loss": 0.5644550800323487, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 3.847533941268921, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.6999043464660645, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 3.3568928241729736, - "learning_rate": 4.444499194921691e-05, - "loss": 0.5565146446228028, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 3.4787251949310303, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6062619686126709, - "step": 8140 - }, - { - "epoch": 3.415791057665886, - "grad_norm": 9.04430866241455, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.4710696220397949, - "step": 8160 - }, - { - "epoch": 3.424165499914947, - "grad_norm": 2.6757805347442627, - "learning_rate": 4.408588490950914e-05, - "loss": 0.5070647716522216, - "step": 8180 - }, - { - "epoch": 3.4325399421640084, - "grad_norm": 8.834247589111328, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.8732002258300782, - "step": 8200 - }, - { - "epoch": 3.440914384413069, - "grad_norm": 5.973219871520996, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.5055295944213867, - "step": 8220 - }, - { - "epoch": 3.4492888266621304, - "grad_norm": 9.602012634277344, - "learning_rate": 4.371708001159045e-05, - "loss": 0.6795970439910889, - "step": 8240 - }, - { - "epoch": 3.4576632689111917, - "grad_norm": 1.5608420372009277, - "learning_rate": 4.359202221530606e-05, - "loss": 0.5012212753295898, - "step": 8260 - }, - { - "epoch": 3.466037711160253, - "grad_norm": 6.789797306060791, - "learning_rate": 4.34659147219236e-05, - "loss": 0.5281818866729736, - "step": 8280 - }, - { - "epoch": 3.4744121534093138, - "grad_norm": 4.547278881072998, - "learning_rate": 4.333876465141465e-05, - "loss": 0.5114842891693115, - "step": 8300 - }, - { - "epoch": 3.482786595658375, - "grad_norm": 1.368796467781067, - "learning_rate": 4.321057918261423e-05, - "loss": 0.40313048362731935, - "step": 8320 - }, - { - "epoch": 3.4911610379074363, - "grad_norm": 1.7168347835540771, - "learning_rate": 4.3081365552815465e-05, - "loss": 0.45522661209106446, - "step": 8340 - }, - { - "epoch": 3.4995354801564975, - "grad_norm": 1.5703785419464111, - "learning_rate": 4.2951131057360984e-05, - "loss": 0.6539173126220703, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966704514664832e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/15/latest.json b/slots/15/latest.json deleted file mode 100644 index 723fa231bcbdb558a6a5f79b496fa674b856ca8f..0000000000000000000000000000000000000000 --- a/slots/15/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:15", "checkpoint": "slots/15/checkpoint-8362", "step": 8362, "updated_at": 1776417975} diff --git a/slots/16/checkpoint-8362/config.json b/slots/16/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/16/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/16/checkpoint-8362/trainer_state.json b/slots/16/checkpoint-8362/trainer_state.json deleted file mode 100644 index 601da08817367d9a8b05cc4a8ffa8ba2fb5641ba..0000000000000000000000000000000000000000 --- a/slots/16/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.6255728824204834, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 35.15928649902344, - "learning_rate": 4.999985708624277e-05, - "loss": 42.4168701171875, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 24.441553115844727, - "learning_rate": 4.999851618030373e-05, - "loss": 34.59369812011719, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 27.142696380615234, - "learning_rate": 4.9995763869538124e-05, - "loss": 29.539852905273438, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 20.90863037109375, - "learning_rate": 4.999160030934015e-05, - "loss": 26.295675659179686, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 19.69276237487793, - "learning_rate": 4.9986025734782515e-05, - "loss": 26.864825439453124, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 20.81488037109375, - "learning_rate": 4.997904046060316e-05, - "loss": 21.519651794433592, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 19.697011947631836, - "learning_rate": 4.997064488118748e-05, - "loss": 20.201951599121095, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 19.01127815246582, - "learning_rate": 4.996083947054607e-05, - "loss": 18.92204895019531, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 21.61073112487793, - "learning_rate": 4.9949624782287944e-05, - "loss": 18.937400817871094, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 23.090789794921875, - "learning_rate": 4.993700144958927e-05, - "loss": 15.520936584472656, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.096433639526367, - "learning_rate": 4.992297018515767e-05, - "loss": 15.487298583984375, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 21.39566421508789, - "learning_rate": 4.990753178119193e-05, - "loss": 14.596461486816406, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 16.463520050048828, - "learning_rate": 4.989068710933732e-05, - "loss": 12.615646362304688, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 14.016276359558105, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.938156890869141, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 20.464670181274414, - "learning_rate": 4.985278284547495e-05, - "loss": 10.395332336425781, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 17.77933692932129, - "learning_rate": 4.983172539352465e-05, - "loss": 10.609977722167969, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 12.832728385925293, - "learning_rate": 4.980926595367956e-05, - "loss": 8.691991424560547, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 13.371121406555176, - "learning_rate": 4.978540579398941e-05, - "loss": 8.057951354980469, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 16.301082611083984, - "learning_rate": 4.9760146261587936e-05, - "loss": 7.973057556152344, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 12.96883487701416, - "learning_rate": 4.973348878261681e-05, - "loss": 7.619200897216797, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 12.284865379333496, - "learning_rate": 4.970543486214516e-05, - "loss": 7.1647804260253904, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 21.732789993286133, - "learning_rate": 4.967598608408454e-05, - "loss": 6.469166564941406, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 10.566523551940918, - "learning_rate": 4.9645144111099515e-05, - "loss": 5.752484130859375, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 11.216104507446289, - "learning_rate": 4.961291068451385e-05, - "loss": 5.699852752685547, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 9.675246238708496, - "learning_rate": 4.9579287624212093e-05, - "loss": 5.267290878295898, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 11.593718528747559, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.2138721466064455, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.523468971252441, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.134304428100586, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 10.710789680480957, - "learning_rate": 4.94701000160798e-05, - "loss": 4.5618743896484375, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.197118759155273, - "learning_rate": 4.943093818728689e-05, - "loss": 4.101953125, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 8.9686279296875, - "learning_rate": 4.939039699886213e-05, - "loss": 3.642832565307617, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 8.167902946472168, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.1743133544921873, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 9.323243141174316, - "learning_rate": 4.930518577661388e-05, - "loss": 3.6948711395263674, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 10.735540390014648, - "learning_rate": 4.926052055377716e-05, - "loss": 3.1577367782592773, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 12.350885391235352, - "learning_rate": 4.921448559301048e-05, - "loss": 3.0701583862304687, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 9.401058197021484, - "learning_rate": 4.916708349342674e-05, - "loss": 2.8100170135498046, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 7.154083728790283, - "learning_rate": 4.911831693132686e-05, - "loss": 2.472774696350098, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 7.130964279174805, - "learning_rate": 4.906818866004871e-05, - "loss": 2.597759246826172, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 7.370577812194824, - "learning_rate": 4.901670150981164e-05, - "loss": 2.3146718978881835, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 9.05713176727295, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1284273147583006, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 7.408658027648926, - "learning_rate": 4.890966227678239e-05, - "loss": 2.052642822265625, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 17.97675895690918, - "learning_rate": 4.885411623737651e-05, - "loss": 2.0922914505004884, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 11.770293235778809, - "learning_rate": 4.87972234054431e-05, - "loss": 1.9368261337280273, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 5.871363162994385, - "learning_rate": 4.87389869931255e-05, - "loss": 2.0074506759643556, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 7.695994853973389, - "learning_rate": 4.8679410288425e-05, - "loss": 1.9740360260009766, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 5.52763557434082, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7269515991210938, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.56577205657959, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.6195665359497071, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 7.687935829162598, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.576361083984375, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.840238094329834, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.704756736755371, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 4.519632816314697, - "learning_rate": 4.836154274553069e-05, - "loss": 2.219845199584961, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 5.06703519821167, - "learning_rate": 4.829399755867636e-05, - "loss": 2.371290397644043, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 5.222914695739746, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.2122323036193847, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 3.4668195247650146, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.334171199798584, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 4.191354274749756, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2463009834289551, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 5.007280349731445, - "learning_rate": 4.801070437462659e-05, - "loss": 1.0995778083801269, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 5.577613830566406, - "learning_rate": 4.793662296282393e-05, - "loss": 1.1267754554748535, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 3.4261252880096436, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.1237424850463866, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 4.648379802703857, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.3997676849365235, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 3.584986448287964, - "learning_rate": 4.770662587550643e-05, - "loss": 0.9382729530334473, - "step": 7460 - }, - { - "epoch": 2.3485872989391248, - "grad_norm": 5.842440605163574, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9716793060302734, - "step": 7480 - }, - { - "epoch": 2.354868151172262, - "grad_norm": 4.669426918029785, - "learning_rate": 4.754687724066917e-05, - "loss": 0.9839717864990234, - "step": 7500 - }, - { - "epoch": 2.3611490034053997, - "grad_norm": 3.096353769302368, - "learning_rate": 4.746509117076991e-05, - "loss": 0.8640604972839355, - "step": 7520 - }, - { - "epoch": 2.367429855638537, - "grad_norm": 4.052355766296387, - "learning_rate": 4.7382036732065104e-05, - "loss": 0.9168565750122071, - "step": 7540 - }, - { - "epoch": 2.373710707871674, - "grad_norm": 3.556760787963867, - "learning_rate": 4.72977186137705e-05, - "loss": 0.8850119590759278, - "step": 7560 - }, - { - "epoch": 2.379991560104812, - "grad_norm": 4.15118408203125, - "learning_rate": 4.721214157644861e-05, - "loss": 0.866028118133545, - "step": 7580 - }, - { - "epoch": 2.386272412337949, - "grad_norm": 5.301335334777832, - "learning_rate": 4.712531045173996e-05, - "loss": 0.7654078483581543, - "step": 7600 - }, - { - "epoch": 2.3925532645710863, - "grad_norm": 5.883398532867432, - "learning_rate": 4.703723014209025e-05, - "loss": 0.7791702270507812, - "step": 7620 - }, - { - "epoch": 2.398834116804224, - "grad_norm": 2.041224241256714, - "learning_rate": 4.694790562047361e-05, - "loss": 0.774465274810791, - "step": 7640 - }, - { - "epoch": 2.405114969037361, - "grad_norm": 3.479743719100952, - "learning_rate": 4.685734193011182e-05, - "loss": 0.7150816440582275, - "step": 7660 - }, - { - "epoch": 2.4113958212704985, - "grad_norm": 3.2080256938934326, - "learning_rate": 4.676554418418953e-05, - "loss": 0.7573625087738037, - "step": 7680 - }, - { - "epoch": 2.417676673503636, - "grad_norm": 2.7411186695098877, - "learning_rate": 4.667251756556563e-05, - "loss": 0.9754419326782227, - "step": 7700 - }, - { - "epoch": 2.4239575257367734, - "grad_norm": 3.309274911880493, - "learning_rate": 4.657826732648061e-05, - "loss": 0.6611151695251465, - "step": 7720 - }, - { - "epoch": 2.4302383779699106, - "grad_norm": 4.727346420288086, - "learning_rate": 4.648279878825998e-05, - "loss": 0.6431382179260254, - "step": 7740 - }, - { - "epoch": 2.4365192302030483, - "grad_norm": 2.4785397052764893, - "learning_rate": 4.63861173410139e-05, - "loss": 0.6248305320739747, - "step": 7760 - }, - { - "epoch": 2.4428000824361855, - "grad_norm": 4.628482818603516, - "learning_rate": 4.628822844333278e-05, - "loss": 0.5700630664825439, - "step": 7780 - }, - { - "epoch": 2.4490809346693228, - "grad_norm": 4.001479625701904, - "learning_rate": 4.618913762197918e-05, - "loss": 0.5712510108947754, - "step": 7800 - }, - { - "epoch": 2.4553617869024604, - "grad_norm": 2.186591625213623, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.6234359741210938, - "step": 7820 - }, - { - "epoch": 2.4616426391355977, - "grad_norm": 2.8685686588287354, - "learning_rate": 4.59873726542891e-05, - "loss": 0.6517808437347412, - "step": 7840 - }, - { - "epoch": 2.467923491368735, - "grad_norm": 3.237225294113159, - "learning_rate": 4.58847098995107e-05, - "loss": 0.5230174541473389, - "step": 7860 - }, - { - "epoch": 2.4742043436018726, - "grad_norm": 2.532601833343506, - "learning_rate": 4.578086800353284e-05, - "loss": 0.5691303730010986, - "step": 7880 - }, - { - "epoch": 2.48048519583501, - "grad_norm": 3.152569532394409, - "learning_rate": 4.56758528292216e-05, - "loss": 0.5500536918640136, - "step": 7900 - }, - { - "epoch": 2.486766048068147, - "grad_norm": 3.253985643386841, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.43275976181030273, - "step": 7920 - }, - { - "epoch": 2.4930469003012847, - "grad_norm": 2.00400447845459, - "learning_rate": 4.546232642794249e-05, - "loss": 0.4545224666595459, - "step": 7940 - }, - { - "epoch": 2.499327752534422, - "grad_norm": 1.4281991720199585, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.5263880729675293, - "step": 7960 - }, - { - "epoch": 2.505608604767559, - "grad_norm": 2.0894293785095215, - "learning_rate": 4.52441789174058e-05, - "loss": 0.4275042533874512, - "step": 7980 - }, - { - "epoch": 2.511889457000697, - "grad_norm": 2.0100977420806885, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.43469738960266113, - "step": 8000 - }, - { - "epoch": 2.518170309233834, - "grad_norm": 1.315463900566101, - "learning_rate": 4.502145956295211e-05, - "loss": 0.42215423583984374, - "step": 8020 - }, - { - "epoch": 2.5244511614669713, - "grad_norm": 6.972764492034912, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.3768086194992065, - "step": 8040 - }, - { - "epoch": 2.530732013700109, - "grad_norm": 2.0225272178649902, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.33483200073242186, - "step": 8060 - }, - { - "epoch": 2.5370128659332463, - "grad_norm": 2.306771993637085, - "learning_rate": 4.46789186299435e-05, - "loss": 0.3931446075439453, - "step": 8080 - }, - { - "epoch": 2.5432937181663835, - "grad_norm": 4.199439525604248, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.37719106674194336, - "step": 8100 - }, - { - "epoch": 2.549574570399521, - "grad_norm": 1.1673650741577148, - "learning_rate": 4.444499194921691e-05, - "loss": 0.3632986545562744, - "step": 8120 - }, - { - "epoch": 2.5558554226326584, - "grad_norm": 2.109755039215088, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.4211446762084961, - "step": 8140 - }, - { - "epoch": 2.5621362748657956, - "grad_norm": 1.5791980028152466, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.5339166164398194, - "step": 8160 - }, - { - "epoch": 2.5684171270989333, - "grad_norm": 0.9690497517585754, - "learning_rate": 4.408588490950914e-05, - "loss": 0.3071375608444214, - "step": 8180 - }, - { - "epoch": 2.5746979793320706, - "grad_norm": 2.0959460735321045, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.344182562828064, - "step": 8200 - }, - { - "epoch": 2.580978831565208, - "grad_norm": 1.8708431720733643, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.3655919313430786, - "step": 8220 - }, - { - "epoch": 2.5872596837983455, - "grad_norm": 2.37229061126709, - "learning_rate": 4.371708001159045e-05, - "loss": 0.2837576627731323, - "step": 8240 - }, - { - "epoch": 2.5935405360314827, - "grad_norm": 21.922826766967773, - "learning_rate": 4.359202221530606e-05, - "loss": 0.3853543043136597, - "step": 8260 - }, - { - "epoch": 2.59982138826462, - "grad_norm": 4.813821315765381, - "learning_rate": 4.34659147219236e-05, - "loss": 0.2917461633682251, - "step": 8280 - }, - { - "epoch": 2.6061022404977576, - "grad_norm": 2.855118751525879, - "learning_rate": 4.333876465141465e-05, - "loss": 0.3352597713470459, - "step": 8300 - }, - { - "epoch": 2.612383092730895, - "grad_norm": 3.0607855319976807, - "learning_rate": 4.321057918261423e-05, - "loss": 0.3263031005859375, - "step": 8320 - }, - { - "epoch": 2.618663944964032, - "grad_norm": 1.7499445676803589, - "learning_rate": 4.3081365552815465e-05, - "loss": 0.4285446643829346, - "step": 8340 - }, - { - "epoch": 2.6249447971971698, - "grad_norm": 1.2600862979888916, - "learning_rate": 4.2951131057360984e-05, - "loss": 0.25559823513031005, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966720054877107e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/16/latest.json b/slots/16/latest.json deleted file mode 100644 index 593eda10104b3fcd544642ee1c3283bb9e4ed288..0000000000000000000000000000000000000000 --- a/slots/16/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:16", "checkpoint": "slots/16/checkpoint-8362", "step": 8362, "updated_at": 1776418083} diff --git a/slots/17/checkpoint-8148/config.json b/slots/17/checkpoint-8148/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/17/checkpoint-8148/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/17/checkpoint-8148/trainer_state.json b/slots/17/checkpoint-8148/trainer_state.json deleted file mode 100644 index 329269c6df9dbe7d319c6a73b762926e8265f828..0000000000000000000000000000000000000000 --- a/slots/17/checkpoint-8148/trainer_state.json +++ /dev/null @@ -1,2890 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.4107663923164493, - "eval_steps": 500, - "global_step": 8148, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.268299102783203, - "learning_rate": 4.999985708624277e-05, - "loss": 4.392966461181641, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 23.723892211914062, - "learning_rate": 4.999851618030373e-05, - "loss": 5.242445755004883, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 15.284976959228516, - "learning_rate": 4.9995763869538124e-05, - "loss": 4.878763198852539, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 21.47739601135254, - "learning_rate": 4.999160030934015e-05, - "loss": 5.070553588867187, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.51418399810791, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.079425048828125, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 17.037799835205078, - "learning_rate": 4.997904046060316e-05, - "loss": 5.397942352294922, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 11.956343650817871, - "learning_rate": 4.997064488118748e-05, - "loss": 5.866295242309571, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 15.19040298461914, - "learning_rate": 4.996083947054607e-05, - "loss": 6.805714416503906, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 12.783180236816406, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.310309219360351, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 15.791704177856445, - "learning_rate": 4.993700144958927e-05, - "loss": 5.249310302734375, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 15.071763038635254, - "learning_rate": 4.992297018515767e-05, - "loss": 5.418260955810547, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 19.809253692626953, - "learning_rate": 4.990753178119193e-05, - "loss": 4.681427383422852, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 12.614968299865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.647319412231445, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.064179420471191, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.246531677246094, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 15.25048828125, - "learning_rate": 4.985278284547495e-05, - "loss": 4.512583541870117, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 14.296823501586914, - "learning_rate": 4.983172539352465e-05, - "loss": 4.193087005615235, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 10.55685043334961, - "learning_rate": 4.980926595367956e-05, - "loss": 4.492794799804687, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 15.188244819641113, - "learning_rate": 4.978540579398941e-05, - "loss": 4.525092697143554, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 23.44154167175293, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.181382751464843, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 11.647577285766602, - "learning_rate": 4.973348878261681e-05, - "loss": 4.1204273223876955, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.451223373413086, - "learning_rate": 4.970543486214516e-05, - "loss": 3.5496986389160154, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 15.924224853515625, - "learning_rate": 4.967598608408454e-05, - "loss": 3.6359764099121095, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 14.668912887573242, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.783356857299805, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 13.026039123535156, - "learning_rate": 4.961291068451385e-05, - "loss": 3.3328643798828126, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 14.517637252807617, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.825040817260742, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 16.220218658447266, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.251118469238281, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 10.804553985595703, - "learning_rate": 4.9507880274181854e-05, - "loss": 3.2774829864501953, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 16.004716873168945, - "learning_rate": 4.94701000160798e-05, - "loss": 3.2009490966796874, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 8.181561470031738, - "learning_rate": 4.943093818728689e-05, - "loss": 2.7726932525634767, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 10.360410690307617, - "learning_rate": 4.939039699886213e-05, - "loss": 3.2188484191894533, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.99072265625, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.66240234375, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 16.23861312866211, - "learning_rate": 4.930518577661388e-05, - "loss": 2.855569839477539, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 9.361970901489258, - "learning_rate": 4.926052055377716e-05, - "loss": 2.8716863632202148, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 10.665338516235352, - "learning_rate": 4.921448559301048e-05, - "loss": 2.866599464416504, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 6.958986759185791, - "learning_rate": 4.916708349342674e-05, - "loss": 2.719050407409668, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 15.789422035217285, - "learning_rate": 4.911831693132686e-05, - "loss": 2.807525062561035, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 11.294109344482422, - "learning_rate": 4.906818866004871e-05, - "loss": 3.0798269271850587, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 9.417093276977539, - "learning_rate": 4.901670150981164e-05, - "loss": 2.639542579650879, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 9.54472827911377, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.214591217041016, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.778777599334717, - "learning_rate": 4.890966227678239e-05, - "loss": 2.497542381286621, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 18.648605346679688, - "learning_rate": 4.885411623737651e-05, - "loss": 2.118986701965332, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 8.938231468200684, - "learning_rate": 4.87972234054431e-05, - "loss": 2.4246755599975587, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 14.108246803283691, - "learning_rate": 4.87389869931255e-05, - "loss": 2.476232147216797, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 20.093215942382812, - "learning_rate": 4.8679410288425e-05, - "loss": 2.005863571166992, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 6.399996280670166, - "learning_rate": 4.86184966550152e-05, - "loss": 2.176909637451172, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 6.5276103019714355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.833466148376465, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 7.747783660888672, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.028280258178711, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 8.722662925720215, - "learning_rate": 4.8427768950332264e-05, - "loss": 2.0328079223632813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 7.213231563568115, - "learning_rate": 4.836154274553069e-05, - "loss": 1.7321701049804688, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 8.361398696899414, - "learning_rate": 4.829399755867636e-05, - "loss": 1.7763862609863281, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 4.578886032104492, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.8702903747558595, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 7.65287446975708, - "learning_rate": 4.8154965567345073e-05, - "loss": 2.1905149459838866, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.528337478637695, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.5154589653015136, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.492247581481934, - "learning_rate": 4.801070437462659e-05, - "loss": 1.6617794036865234, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 7.519628524780273, - "learning_rate": 4.793662296282393e-05, - "loss": 1.5524075508117676, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.682668685913086, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.613736915588379, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 8.832022666931152, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.5550686836242675, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.3445448875427246, - "learning_rate": 4.770662587550643e-05, - "loss": 1.4483519554138184, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 8.750604629516602, - "learning_rate": 4.762739032415868e-05, - "loss": 1.5183201789855958, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.896105766296387, - "learning_rate": 4.754687724066917e-05, - "loss": 1.298677349090576, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 6.540890693664551, - "learning_rate": 4.746509117076991e-05, - "loss": 1.3707378387451172, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 5.634416580200195, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.2700369834899903, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 5.313226699829102, - "learning_rate": 4.72977186137705e-05, - "loss": 1.3874421119689941, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 4.164125919342041, - "learning_rate": 4.721214157644861e-05, - "loss": 1.3053173065185546, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 3.9634926319122314, - "learning_rate": 4.712531045173996e-05, - "loss": 1.1017135620117187, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 19.879589080810547, - "learning_rate": 4.703723014209025e-05, - "loss": 1.5074621200561524, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 7.186710834503174, - "learning_rate": 4.694790562047361e-05, - "loss": 1.3602344512939453, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 6.228899955749512, - "learning_rate": 4.685734193011182e-05, - "loss": 1.097489070892334, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 4.770694255828857, - "learning_rate": 4.676554418418953e-05, - "loss": 1.1659790992736816, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 4.967444896697998, - "learning_rate": 4.667251756556563e-05, - "loss": 1.1331518173217774, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.312889099121094, - "learning_rate": 4.657826732648061e-05, - "loss": 1.2721074104309082, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.759788751602173, - "learning_rate": 4.648279878825998e-05, - "loss": 0.9295409202575684, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 7.215264797210693, - "learning_rate": 4.63861173410139e-05, - "loss": 1.196054744720459, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673015117645264, - "learning_rate": 4.628822844333278e-05, - "loss": 1.0879426002502441, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 6.737372875213623, - "learning_rate": 4.618913762197918e-05, - "loss": 1.0760483741760254, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 3.58354115486145, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.9610136985778809, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 3.435591220855713, - "learning_rate": 4.59873726542891e-05, - "loss": 1.0568583488464356, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 16.85455322265625, - "learning_rate": 4.58847098995107e-05, - "loss": 1.020804214477539, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 3.1317970752716064, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8810656547546387, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 5.6492133140563965, - "learning_rate": 4.56758528292216e-05, - "loss": 0.834989356994629, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 5.5128045082092285, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.8088371276855468, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 3.422081232070923, - "learning_rate": 4.546232642794249e-05, - "loss": 1.0146251678466798, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 4.3992838859558105, - "learning_rate": 4.5353827256577894e-05, - "loss": 1.0001090049743653, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 10.456899642944336, - "learning_rate": 4.52441789174058e-05, - "loss": 0.9053573608398438, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 2.4118423461914062, - "learning_rate": 4.5133387601121476e-05, - "loss": 1.0744664192199707, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 5.180037021636963, - "learning_rate": 4.502145956295211e-05, - "loss": 0.8649028778076172, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 3.958254337310791, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.9063331604003906, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 4.954156398773193, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.7377657413482666, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 3.4824178218841553, - "learning_rate": 4.46789186299435e-05, - "loss": 0.7941797733306885, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 4.906705379486084, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.7057065010070801, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 5.808215141296387, - "learning_rate": 4.444499194921691e-05, - "loss": 0.8522198677062989, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 5.282096862792969, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6269611835479736, - "step": 8140 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.913494827834547e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/17/latest.json b/slots/17/latest.json deleted file mode 100644 index a787752bff3c53b665a496705e47ecac229a4643..0000000000000000000000000000000000000000 --- a/slots/17/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:17", "checkpoint": "slots/17/checkpoint-8148", "step": 8148, "updated_at": 1776417583} diff --git a/slots/18/checkpoint-8362/config.json b/slots/18/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/18/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/18/checkpoint-8362/trainer_state.json b/slots/18/checkpoint-8362/trainer_state.json deleted file mode 100644 index cd9a181b12fa508f9bf2339ff45829083c1cb3de..0000000000000000000000000000000000000000 --- a/slots/18/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.5003729243814035, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 10.619898796081543, - "learning_rate": 4.999985708624277e-05, - "loss": 5.514506530761719, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.31767463684082, - "learning_rate": 4.999851618030373e-05, - "loss": 5.064883041381836, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 13.286064147949219, - "learning_rate": 4.9995763869538124e-05, - "loss": 5.178463363647461, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 13.165849685668945, - "learning_rate": 4.999160030934015e-05, - "loss": 5.5684253692626955, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.71381950378418, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.4804939270019535, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 15.386504173278809, - "learning_rate": 4.997904046060316e-05, - "loss": 5.793609619140625, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 30.56761360168457, - "learning_rate": 4.997064488118748e-05, - "loss": 5.199003982543945, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 10.182106971740723, - "learning_rate": 4.996083947054607e-05, - "loss": 5.1698661804199215, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.703635215759277, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.147470855712891, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 17.958473205566406, - "learning_rate": 4.993700144958927e-05, - "loss": 4.737754058837891, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 16.8748779296875, - "learning_rate": 4.992297018515767e-05, - "loss": 4.211444091796875, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 46.32542037963867, - "learning_rate": 4.990753178119193e-05, - "loss": 4.386423492431641, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.126466751098633, - "learning_rate": 4.989068710933732e-05, - "loss": 4.631509017944336, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.159765243530273, - "learning_rate": 4.9872437120636304e-05, - "loss": 3.881015396118164, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 12.283050537109375, - "learning_rate": 4.985278284547495e-05, - "loss": 3.940886306762695, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 12.025839805603027, - "learning_rate": 4.983172539352465e-05, - "loss": 3.9277923583984373, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 12.16906452178955, - "learning_rate": 4.980926595367956e-05, - "loss": 4.246135330200195, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.341684341430664, - "learning_rate": 4.978540579398941e-05, - "loss": 4.455572891235351, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 8.899483680725098, - "learning_rate": 4.9760146261587936e-05, - "loss": 3.8522106170654298, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.554529190063477, - "learning_rate": 4.973348878261681e-05, - "loss": 3.7468845367431642, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 9.822662353515625, - "learning_rate": 4.970543486214516e-05, - "loss": 3.0998603820800783, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 10.157042503356934, - "learning_rate": 4.967598608408454e-05, - "loss": 3.316312789916992, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 9.876077651977539, - "learning_rate": 4.9645144111099515e-05, - "loss": 4.0845588684082035, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 11.132689476013184, - "learning_rate": 4.961291068451385e-05, - "loss": 3.1008926391601563, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 9.930462837219238, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.9224714279174804, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 7.317378997802734, - "learning_rate": 4.9544276828536906e-05, - "loss": 2.853228759765625, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.085936546325684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.5410511016845705, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 7.214178085327148, - "learning_rate": 4.94701000160798e-05, - "loss": 3.0386709213256835, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 7.6133952140808105, - "learning_rate": 4.943093818728689e-05, - "loss": 2.6715574264526367, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 7.800808429718018, - "learning_rate": 4.939039699886213e-05, - "loss": 2.398478126525879, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 5.941168785095215, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.8046531677246094, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 14.817143440246582, - "learning_rate": 4.930518577661388e-05, - "loss": 2.6330236434936523, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 8.76217269897461, - "learning_rate": 4.926052055377716e-05, - "loss": 2.199250602722168, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 11.228110313415527, - "learning_rate": 4.921448559301048e-05, - "loss": 2.290775489807129, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 7.807727813720703, - "learning_rate": 4.916708349342674e-05, - "loss": 2.2031742095947267, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 8.53569507598877, - "learning_rate": 4.911831693132686e-05, - "loss": 1.9725076675415039, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 7.547488689422607, - "learning_rate": 4.906818866004871e-05, - "loss": 2.019377326965332, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 5.556341648101807, - "learning_rate": 4.901670150981164e-05, - "loss": 2.0958667755126954, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 10.818191528320312, - "learning_rate": 4.8963858387556665e-05, - "loss": 1.8875907897949218, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 5.718668460845947, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1760227203369142, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 12.06888484954834, - "learning_rate": 4.885411623737651e-05, - "loss": 1.980560302734375, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 5.500242233276367, - "learning_rate": 4.87972234054431e-05, - "loss": 1.7397878646850586, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.90056848526001, - "learning_rate": 4.87389869931255e-05, - "loss": 1.8117895126342773, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 4.671687126159668, - "learning_rate": 4.8679410288425e-05, - "loss": 1.8994810104370117, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 4.093003749847412, - "learning_rate": 4.86184966550152e-05, - "loss": 1.5177894592285157, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 7.3171610832214355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.4882768630981444, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.505290985107422, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.8049379348754884, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.227847576141357, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.3910842895507813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 6.138688564300537, - "learning_rate": 4.836154274553069e-05, - "loss": 1.5944181442260743, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 9.806740760803223, - "learning_rate": 4.829399755867636e-05, - "loss": 1.9773569107055664, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 17.26584243774414, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.3594069480895996, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 4.6048479080200195, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.365511989593506, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 3.5992016792297363, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2431794166564942, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 3.939440965652466, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4344836235046388, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.212338447570801, - "learning_rate": 4.793662296282393e-05, - "loss": 1.340198040008545, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.5816969871521, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.0717330932617188, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 6.169701099395752, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1681714057922363, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.759722948074341, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2151671409606934, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 2.909806966781616, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9399504661560059, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 5.237748146057129, - "learning_rate": 4.754687724066917e-05, - "loss": 1.2489306449890136, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 3.2770180702209473, - "learning_rate": 4.746509117076991e-05, - "loss": 1.25095853805542, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 2.7698371410369873, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.3378321647644043, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 3.8602523803710938, - "learning_rate": 4.72977186137705e-05, - "loss": 1.3043223381042481, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 2.804724931716919, - "learning_rate": 4.721214157644861e-05, - "loss": 1.0128710746765137, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 5.5239577293396, - "learning_rate": 4.712531045173996e-05, - "loss": 0.9921689033508301, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 11.35846996307373, - "learning_rate": 4.7041663749229935e-05, - "loss": 1.3766287803649901, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 10.031026840209961, - "learning_rate": 4.6952401319017915e-05, - "loss": 0.9078640937805176, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 2.9733171463012695, - "learning_rate": 4.686189946623564e-05, - "loss": 0.8636553764343262, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 4.90701150894165, - "learning_rate": 4.677016330057647e-05, - "loss": 0.8574177742004394, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 9.797515869140625, - "learning_rate": 4.667719800142248e-05, - "loss": 1.1090837478637696, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.111540794372559, - "learning_rate": 4.6583008817552094e-05, - "loss": 0.8992094039916992, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.002032518386841, - "learning_rate": 4.6487601066843674e-05, - "loss": 0.7496014595031738, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.6695992946624756, - "learning_rate": 4.639098013597534e-05, - "loss": 0.6838997840881348, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.092342853546143, - "learning_rate": 4.629315148012079e-05, - "loss": 0.689493465423584, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 6.389557361602783, - "learning_rate": 4.619412062264134e-05, - "loss": 0.9438887596130371, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.32489275932312, - "learning_rate": 4.609389315477405e-05, - "loss": 0.7676928043365479, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 2.865983724594116, - "learning_rate": 4.599247473531607e-05, - "loss": 0.8114580154418946, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 8.078557014465332, - "learning_rate": 4.588987109030511e-05, - "loss": 0.6942833423614502, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 3.2191479206085205, - "learning_rate": 4.578608801269621e-05, - "loss": 0.827882194519043, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 7.289218902587891, - "learning_rate": 4.5681131362034604e-05, - "loss": 0.8882099151611328, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 2.5519909858703613, - "learning_rate": 4.5575007064124937e-05, - "loss": 0.6065804958343506, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 3.498563289642334, - "learning_rate": 4.5467721110696685e-05, - "loss": 0.730232048034668, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 2.0245680809020996, - "learning_rate": 4.5359279559065876e-05, - "loss": 0.9999460220336914, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 2.994798183441162, - "learning_rate": 4.524968853179306e-05, - "loss": 0.6818597316741943, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 3.339902400970459, - "learning_rate": 4.51389542163377e-05, - "loss": 0.7384091854095459, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 2.02184796333313, - "learning_rate": 4.5027082864708735e-05, - "loss": 0.6503865718841553, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 3.314213991165161, - "learning_rate": 4.491408079311169e-05, - "loss": 0.6390689849853516, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 6.766301155090332, - "learning_rate": 4.479995438159201e-05, - "loss": 0.6914161205291748, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 2.9052584171295166, - "learning_rate": 4.468471007367485e-05, - "loss": 0.7612205028533936, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 3.555596351623535, - "learning_rate": 4.456835437600129e-05, - "loss": 0.5866696834564209, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 5.336625576019287, - "learning_rate": 4.445089385796099e-05, - "loss": 0.6077582359313964, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 3.4913692474365234, - "learning_rate": 4.4332335151321235e-05, - "loss": 0.7415694236755371, - "step": 8140 - }, - { - "epoch": 3.415791057665886, - "grad_norm": 1.9187442064285278, - "learning_rate": 4.421268494985256e-05, - "loss": 0.6037701606750489, - "step": 8160 - }, - { - "epoch": 3.424165499914947, - "grad_norm": 5.737117290496826, - "learning_rate": 4.409195000895079e-05, - "loss": 0.7351339340209961, - "step": 8180 - }, - { - "epoch": 3.4325399421640084, - "grad_norm": 1.966489315032959, - "learning_rate": 4.397013714525563e-05, - "loss": 0.5357085704803467, - "step": 8200 - }, - { - "epoch": 3.440914384413069, - "grad_norm": 3.426973342895508, - "learning_rate": 4.384725323626584e-05, - "loss": 0.5511207103729248, - "step": 8220 - }, - { - "epoch": 3.4492888266621304, - "grad_norm": 1.806939959526062, - "learning_rate": 4.3723305219950904e-05, - "loss": 0.6522912979125977, - "step": 8240 - }, - { - "epoch": 3.4576632689111917, - "grad_norm": 4.771038055419922, - "learning_rate": 4.35983000943593e-05, - "loss": 0.5756620407104492, - "step": 8260 - }, - { - "epoch": 3.466037711160253, - "grad_norm": 2.907762050628662, - "learning_rate": 4.347224491722345e-05, - "loss": 0.6083052635192872, - "step": 8280 - }, - { - "epoch": 3.4744121534093138, - "grad_norm": 2.4061028957366943, - "learning_rate": 4.3345146805561156e-05, - "loss": 0.4108818531036377, - "step": 8300 - }, - { - "epoch": 3.482786595658375, - "grad_norm": 2.462939977645874, - "learning_rate": 4.321701293527388e-05, - "loss": 0.4628077507019043, - "step": 8320 - }, - { - "epoch": 3.4911610379074363, - "grad_norm": 2.649022340774536, - "learning_rate": 4.308785054074151e-05, - "loss": 0.46345086097717286, - "step": 8340 - }, - { - "epoch": 3.4995354801564975, - "grad_norm": 4.22981595993042, - "learning_rate": 4.2957666914413975e-05, - "loss": 0.5101219654083252, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966704514664832e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/18/latest.json b/slots/18/latest.json deleted file mode 100644 index a9e5f9618152602a042c1a1931d13c2b2275ae15..0000000000000000000000000000000000000000 --- a/slots/18/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:18", "checkpoint": "slots/18/checkpoint-8362", "step": 8362, "updated_at": 1776418297} diff --git a/slots/19/checkpoint-8362/config.json b/slots/19/checkpoint-8362/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/19/checkpoint-8362/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/19/checkpoint-8362/trainer_state.json b/slots/19/checkpoint-8362/trainer_state.json deleted file mode 100644 index 789db1a10d43e96c6ca1d6520a17ccd691f19b14..0000000000000000000000000000000000000000 --- a/slots/19/checkpoint-8362/trainer_state.json +++ /dev/null @@ -1,2967 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.6255728824204834, - "eval_steps": 500, - "global_step": 8362, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 31.780010223388672, - "learning_rate": 4.999985708624277e-05, - "loss": 47.03160400390625, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 22.647113800048828, - "learning_rate": 4.999851618030373e-05, - "loss": 39.12403564453125, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 24.094038009643555, - "learning_rate": 4.9995763869538124e-05, - "loss": 32.462225341796874, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 25.049236297607422, - "learning_rate": 4.999160030934015e-05, - "loss": 25.110198974609375, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 20.87278938293457, - "learning_rate": 4.9986025734782515e-05, - "loss": 24.496664428710936, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 19.164220809936523, - "learning_rate": 4.997904046060316e-05, - "loss": 21.625892639160156, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 17.22857666015625, - "learning_rate": 4.997064488118748e-05, - "loss": 18.14871826171875, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 17.680524826049805, - "learning_rate": 4.996083947054607e-05, - "loss": 19.17234649658203, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 17.1239070892334, - "learning_rate": 4.9949624782287944e-05, - "loss": 17.536575317382812, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 15.556852340698242, - "learning_rate": 4.993700144958927e-05, - "loss": 14.654452514648437, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.166357040405273, - "learning_rate": 4.992297018515767e-05, - "loss": 12.504849243164063, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 15.202767372131348, - "learning_rate": 4.990753178119193e-05, - "loss": 14.847886657714843, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 14.017868995666504, - "learning_rate": 4.989068710933732e-05, - "loss": 10.980046844482422, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 16.90302848815918, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.107038879394532, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 26.98078155517578, - "learning_rate": 4.985278284547495e-05, - "loss": 10.064899444580078, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 19.16476058959961, - "learning_rate": 4.983172539352465e-05, - "loss": 9.747372436523438, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 14.147124290466309, - "learning_rate": 4.980926595367956e-05, - "loss": 7.905096435546875, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 16.14385986328125, - "learning_rate": 4.978540579398941e-05, - "loss": 8.444129943847656, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 13.877010345458984, - "learning_rate": 4.9760146261587936e-05, - "loss": 6.83173828125, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 13.799625396728516, - "learning_rate": 4.973348878261681e-05, - "loss": 7.934703826904297, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 15.579751014709473, - "learning_rate": 4.970543486214516e-05, - "loss": 6.793711853027344, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 13.521928787231445, - "learning_rate": 4.967598608408454e-05, - "loss": 7.2837577819824215, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 15.898303031921387, - "learning_rate": 4.9645144111099515e-05, - "loss": 6.6693672180175785, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 24.22979164123535, - "learning_rate": 4.961291068451385e-05, - "loss": 5.961820983886719, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 11.246175765991211, - "learning_rate": 4.9579287624212093e-05, - "loss": 4.978904724121094, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 24.801748275756836, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.248088073730469, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.347773551940918, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.6643627166748045, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 12.957473754882812, - "learning_rate": 4.94701000160798e-05, - "loss": 4.121147918701172, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.496082305908203, - "learning_rate": 4.943093818728689e-05, - "loss": 3.8258975982666015, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 12.250146865844727, - "learning_rate": 4.939039699886213e-05, - "loss": 3.8255241394042967, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 12.506306648254395, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.5919628143310547, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 16.467334747314453, - "learning_rate": 4.930518577661388e-05, - "loss": 3.715103530883789, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 8.267026901245117, - "learning_rate": 4.926052055377716e-05, - "loss": 3.194218063354492, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 9.150392532348633, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9999900817871095, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 11.638959884643555, - "learning_rate": 4.916708349342674e-05, - "loss": 2.837971496582031, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 10.091157913208008, - "learning_rate": 4.911831693132686e-05, - "loss": 2.7495031356811523, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 9.66382884979248, - "learning_rate": 4.906818866004871e-05, - "loss": 2.6875839233398438, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 8.437216758728027, - "learning_rate": 4.901670150981164e-05, - "loss": 2.5731342315673826, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 8.248567581176758, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.859779930114746, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 11.652517318725586, - "learning_rate": 4.890966227678239e-05, - "loss": 2.2908452987670898, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 14.507343292236328, - "learning_rate": 4.885411623737651e-05, - "loss": 2.302378463745117, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 9.44881820678711, - "learning_rate": 4.87972234054431e-05, - "loss": 2.3078046798706056, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 9.367117881774902, - "learning_rate": 4.87389869931255e-05, - "loss": 2.1523136138916015, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 8.667911529541016, - "learning_rate": 4.8679410288425e-05, - "loss": 2.2135799407958983, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 16.578901290893555, - "learning_rate": 4.86184966550152e-05, - "loss": 2.1074777603149415, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.302729606628418, - "learning_rate": 4.8556249532052045e-05, - "loss": 2.0359333038330076, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 20.793943405151367, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.2183906555175783, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.577836990356445, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9698509216308593, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 5.63059139251709, - "learning_rate": 4.836154274553069e-05, - "loss": 1.652799415588379, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 6.654581069946289, - "learning_rate": 4.829399755867636e-05, - "loss": 1.740532112121582, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 11.970749855041504, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.6816909790039063, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 7.981494426727295, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.61510009765625, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 6.4518561363220215, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.6320558547973634, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 11.362939834594727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.474619197845459, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 4.846298694610596, - "learning_rate": 4.793662296282393e-05, - "loss": 1.4175973892211915, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 7.837019920349121, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.5572842597961425, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 5.593219757080078, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.439277744293213, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 5.2232160568237305, - "learning_rate": 4.770662587550643e-05, - "loss": 1.3270212173461915, - "step": 7460 - }, - { - "epoch": 2.3485872989391248, - "grad_norm": 6.035870552062988, - "learning_rate": 4.762739032415868e-05, - "loss": 1.3490946769714356, - "step": 7480 - }, - { - "epoch": 2.354868151172262, - "grad_norm": 6.241731643676758, - "learning_rate": 4.754687724066917e-05, - "loss": 1.2901057243347167, - "step": 7500 - }, - { - "epoch": 2.3611490034053997, - "grad_norm": 11.480992317199707, - "learning_rate": 4.746509117076991e-05, - "loss": 1.252013874053955, - "step": 7520 - }, - { - "epoch": 2.367429855638537, - "grad_norm": 3.9125211238861084, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.154312515258789, - "step": 7540 - }, - { - "epoch": 2.373710707871674, - "grad_norm": 5.499811172485352, - "learning_rate": 4.72977186137705e-05, - "loss": 1.0750234603881836, - "step": 7560 - }, - { - "epoch": 2.379991560104812, - "grad_norm": 5.505170822143555, - "learning_rate": 4.721214157644861e-05, - "loss": 1.490649127960205, - "step": 7580 - }, - { - "epoch": 2.386272412337949, - "grad_norm": 5.891989231109619, - "learning_rate": 4.712531045173996e-05, - "loss": 1.1875076293945312, - "step": 7600 - }, - { - "epoch": 2.3925532645710863, - "grad_norm": 5.132697582244873, - "learning_rate": 4.703723014209025e-05, - "loss": 1.0753458976745605, - "step": 7620 - }, - { - "epoch": 2.398834116804224, - "grad_norm": 5.394096851348877, - "learning_rate": 4.694790562047361e-05, - "loss": 1.0024099349975586, - "step": 7640 - }, - { - "epoch": 2.405114969037361, - "grad_norm": 4.041260719299316, - "learning_rate": 4.685734193011182e-05, - "loss": 1.2125065803527832, - "step": 7660 - }, - { - "epoch": 2.4113958212704985, - "grad_norm": 4.972134113311768, - "learning_rate": 4.676554418418953e-05, - "loss": 0.9521866798400879, - "step": 7680 - }, - { - "epoch": 2.417676673503636, - "grad_norm": 10.228482246398926, - "learning_rate": 4.667251756556563e-05, - "loss": 1.0725263595581054, - "step": 7700 - }, - { - "epoch": 2.4239575257367734, - "grad_norm": 3.5744264125823975, - "learning_rate": 4.657826732648061e-05, - "loss": 0.9225356101989746, - "step": 7720 - }, - { - "epoch": 2.4302383779699106, - "grad_norm": 4.8591203689575195, - "learning_rate": 4.648279878825998e-05, - "loss": 0.8432273864746094, - "step": 7740 - }, - { - "epoch": 2.4365192302030483, - "grad_norm": 3.154543161392212, - "learning_rate": 4.63861173410139e-05, - "loss": 0.9964156150817871, - "step": 7760 - }, - { - "epoch": 2.4428000824361855, - "grad_norm": 3.979494571685791, - "learning_rate": 4.628822844333278e-05, - "loss": 0.8577518463134766, - "step": 7780 - }, - { - "epoch": 2.4490809346693228, - "grad_norm": 5.186657428741455, - "learning_rate": 4.618913762197918e-05, - "loss": 0.8347867965698242, - "step": 7800 - }, - { - "epoch": 2.4553617869024604, - "grad_norm": 2.6562411785125732, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.8252105712890625, - "step": 7820 - }, - { - "epoch": 2.4616426391355977, - "grad_norm": 3.5569987297058105, - "learning_rate": 4.59873726542891e-05, - "loss": 1.0673627853393555, - "step": 7840 - }, - { - "epoch": 2.467923491368735, - "grad_norm": 4.606166839599609, - "learning_rate": 4.58847098995107e-05, - "loss": 0.9000311851501465, - "step": 7860 - }, - { - "epoch": 2.4742043436018726, - "grad_norm": 5.916482925415039, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8248819351196289, - "step": 7880 - }, - { - "epoch": 2.48048519583501, - "grad_norm": 2.7883312702178955, - "learning_rate": 4.56758528292216e-05, - "loss": 0.7192736148834229, - "step": 7900 - }, - { - "epoch": 2.486766048068147, - "grad_norm": 7.564154148101807, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.7242763042449951, - "step": 7920 - }, - { - "epoch": 2.4930469003012847, - "grad_norm": 4.136919975280762, - "learning_rate": 4.546232642794249e-05, - "loss": 0.7495751857757569, - "step": 7940 - }, - { - "epoch": 2.499327752534422, - "grad_norm": 5.9343743324279785, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.6677919387817383, - "step": 7960 - }, - { - "epoch": 2.505608604767559, - "grad_norm": 5.331826210021973, - "learning_rate": 4.52441789174058e-05, - "loss": 0.611272144317627, - "step": 7980 - }, - { - "epoch": 2.511889457000697, - "grad_norm": 2.7531578540802, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.7472806930541992, - "step": 8000 - }, - { - "epoch": 2.518170309233834, - "grad_norm": 2.8748040199279785, - "learning_rate": 4.502145956295211e-05, - "loss": 0.7553040027618408, - "step": 8020 - }, - { - "epoch": 2.5244511614669713, - "grad_norm": 4.344831943511963, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.6073373317718506, - "step": 8040 - }, - { - "epoch": 2.530732013700109, - "grad_norm": 5.624392032623291, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.7095471858978272, - "step": 8060 - }, - { - "epoch": 2.5370128659332463, - "grad_norm": 6.282398700714111, - "learning_rate": 4.46789186299435e-05, - "loss": 0.5880768299102783, - "step": 8080 - }, - { - "epoch": 2.5432937181663835, - "grad_norm": 5.560777187347412, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.6635794639587402, - "step": 8100 - }, - { - "epoch": 2.549574570399521, - "grad_norm": 4.565273761749268, - "learning_rate": 4.444499194921691e-05, - "loss": 0.560918664932251, - "step": 8120 - }, - { - "epoch": 2.5558554226326584, - "grad_norm": 2.701143980026245, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6100841999053955, - "step": 8140 - }, - { - "epoch": 2.5621362748657956, - "grad_norm": 3.9895026683807373, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.5091825962066651, - "step": 8160 - }, - { - "epoch": 2.5684171270989333, - "grad_norm": 3.236727237701416, - "learning_rate": 4.408588490950914e-05, - "loss": 0.8316182136535645, - "step": 8180 - }, - { - "epoch": 2.5746979793320706, - "grad_norm": 2.363232135772705, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.48111910820007325, - "step": 8200 - }, - { - "epoch": 2.580978831565208, - "grad_norm": 2.9194321632385254, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.5176596164703369, - "step": 8220 - }, - { - "epoch": 2.5872596837983455, - "grad_norm": 1.993800163269043, - "learning_rate": 4.371708001159045e-05, - "loss": 0.47290701866149903, - "step": 8240 - }, - { - "epoch": 2.5935405360314827, - "grad_norm": 2.4063210487365723, - "learning_rate": 4.359202221530606e-05, - "loss": 0.49527649879455565, - "step": 8260 - }, - { - "epoch": 2.59982138826462, - "grad_norm": 2.8830325603485107, - "learning_rate": 4.34659147219236e-05, - "loss": 0.47321524620056155, - "step": 8280 - }, - { - "epoch": 2.6061022404977576, - "grad_norm": 2.797807455062866, - "learning_rate": 4.333876465141465e-05, - "loss": 0.5300829410552979, - "step": 8300 - }, - { - "epoch": 2.612383092730895, - "grad_norm": 2.4733376502990723, - "learning_rate": 4.321057918261423e-05, - "loss": 0.5099183559417725, - "step": 8320 - }, - { - "epoch": 2.618663944964032, - "grad_norm": 3.297940492630005, - "learning_rate": 4.3081365552815465e-05, - "loss": 0.49715418815612794, - "step": 8340 - }, - { - "epoch": 2.6249447971971698, - "grad_norm": 1.5188289880752563, - "learning_rate": 4.2951131057360984e-05, - "loss": 0.5388905525207519, - "step": 8360 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": true - }, - "attributes": {} - } - }, - "total_flos": 1.966720054877107e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/19/latest.json b/slots/19/latest.json deleted file mode 100644 index 6a3045d7c34a82888e17ed7c8a47331e98b3c34a..0000000000000000000000000000000000000000 --- a/slots/19/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:19", "checkpoint": "slots/19/checkpoint-8362", "step": 8362, "updated_at": 1776417985} diff --git a/slots/2/checkpoint-7543/config.json b/slots/2/checkpoint-7543/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/2/checkpoint-7543/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/2/checkpoint-7543/trainer_state.json b/slots/2/checkpoint-7543/trainer_state.json deleted file mode 100644 index 1ad1095dcadaac41243ac3c046b158dc4406050b..0000000000000000000000000000000000000000 --- a/slots/2/checkpoint-7543/trainer_state.json +++ /dev/null @@ -1,2680 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.1574395142823497, - "eval_steps": 500, - "global_step": 7543, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 10.619898796081543, - "learning_rate": 4.999985708624277e-05, - "loss": 5.514506530761719, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.31767463684082, - "learning_rate": 4.999851618030373e-05, - "loss": 5.064883041381836, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 13.286064147949219, - "learning_rate": 4.9995763869538124e-05, - "loss": 5.178463363647461, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 13.165849685668945, - "learning_rate": 4.999160030934015e-05, - "loss": 5.5684253692626955, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.71381950378418, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.4804939270019535, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 15.386504173278809, - "learning_rate": 4.997904046060316e-05, - "loss": 5.793609619140625, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 30.56761360168457, - "learning_rate": 4.997064488118748e-05, - "loss": 5.199003982543945, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 10.182106971740723, - "learning_rate": 4.996083947054607e-05, - "loss": 5.1698661804199215, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.703635215759277, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.147470855712891, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 17.958473205566406, - "learning_rate": 4.993700144958927e-05, - "loss": 4.737754058837891, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 16.8748779296875, - "learning_rate": 4.992297018515767e-05, - "loss": 4.211444091796875, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 46.32542037963867, - "learning_rate": 4.990753178119193e-05, - "loss": 4.386423492431641, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.126466751098633, - "learning_rate": 4.989068710933732e-05, - "loss": 4.631509017944336, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.159765243530273, - "learning_rate": 4.9872437120636304e-05, - "loss": 3.881015396118164, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 12.283050537109375, - "learning_rate": 4.985278284547495e-05, - "loss": 3.940886306762695, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 12.025839805603027, - "learning_rate": 4.983172539352465e-05, - "loss": 3.9277923583984373, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 12.16906452178955, - "learning_rate": 4.980926595367956e-05, - "loss": 4.246135330200195, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.341684341430664, - "learning_rate": 4.978540579398941e-05, - "loss": 4.455572891235351, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 8.899483680725098, - "learning_rate": 4.9760146261587936e-05, - "loss": 3.8522106170654298, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.554529190063477, - "learning_rate": 4.973348878261681e-05, - "loss": 3.7468845367431642, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 9.822662353515625, - "learning_rate": 4.970543486214516e-05, - "loss": 3.0998603820800783, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 10.157042503356934, - "learning_rate": 4.967598608408454e-05, - "loss": 3.316312789916992, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 9.876077651977539, - "learning_rate": 4.9645144111099515e-05, - "loss": 4.0845588684082035, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 11.132689476013184, - "learning_rate": 4.961291068451385e-05, - "loss": 3.1008926391601563, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 9.930462837219238, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.9224714279174804, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 7.317378997802734, - "learning_rate": 4.9544276828536906e-05, - "loss": 2.853228759765625, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.085936546325684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.5410511016845705, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 7.214178085327148, - "learning_rate": 4.94701000160798e-05, - "loss": 3.0386709213256835, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 7.6133952140808105, - "learning_rate": 4.943093818728689e-05, - "loss": 2.6715574264526367, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 7.800808429718018, - "learning_rate": 4.939039699886213e-05, - "loss": 2.398478126525879, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 5.941168785095215, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.8046531677246094, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 14.817143440246582, - "learning_rate": 4.930518577661388e-05, - "loss": 2.6330236434936523, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 8.76217269897461, - "learning_rate": 4.926052055377716e-05, - "loss": 2.199250602722168, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 11.228110313415527, - "learning_rate": 4.921448559301048e-05, - "loss": 2.290775489807129, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 7.807727813720703, - "learning_rate": 4.916708349342674e-05, - "loss": 2.2031742095947267, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 8.53569507598877, - "learning_rate": 4.911831693132686e-05, - "loss": 1.9725076675415039, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 7.547488689422607, - "learning_rate": 4.906818866004871e-05, - "loss": 2.019377326965332, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 5.556341648101807, - "learning_rate": 4.901670150981164e-05, - "loss": 2.0958667755126954, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 10.818191528320312, - "learning_rate": 4.8963858387556665e-05, - "loss": 1.8875907897949218, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 5.718668460845947, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1760227203369142, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 12.06888484954834, - "learning_rate": 4.885411623737651e-05, - "loss": 1.980560302734375, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 5.500242233276367, - "learning_rate": 4.87972234054431e-05, - "loss": 1.7397878646850586, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.90056848526001, - "learning_rate": 4.87389869931255e-05, - "loss": 1.8117895126342773, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 4.671687126159668, - "learning_rate": 4.8679410288425e-05, - "loss": 1.8994810104370117, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 4.093003749847412, - "learning_rate": 4.86184966550152e-05, - "loss": 1.5177894592285157, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 7.3171610832214355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.4882768630981444, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.505290985107422, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.8049379348754884, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.227847576141357, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.3910842895507813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 6.138688564300537, - "learning_rate": 4.836154274553069e-05, - "loss": 1.5944181442260743, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 9.806740760803223, - "learning_rate": 4.829399755867636e-05, - "loss": 1.9773569107055664, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 17.26584243774414, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.3594069480895996, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 4.6048479080200195, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.365511989593506, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 3.5992016792297363, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2431794166564942, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 3.939440965652466, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4344836235046388, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.212338447570801, - "learning_rate": 4.793662296282393e-05, - "loss": 1.340198040008545, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.5816969871521, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.0717330932617188, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 6.169701099395752, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1681714057922363, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.759722948074341, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2151671409606934, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 2.909806966781616, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9399504661560059, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 5.237748146057129, - "learning_rate": 4.754687724066917e-05, - "loss": 1.2489306449890136, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 3.2770180702209473, - "learning_rate": 4.746509117076991e-05, - "loss": 1.25095853805542, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 2.7698371410369873, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.3378321647644043, - "step": 7540 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.7630655730106112e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/2/latest.json b/slots/2/latest.json deleted file mode 100644 index 3a77928e96d8cc34ba9c62ba62ce238d5383587f..0000000000000000000000000000000000000000 --- a/slots/2/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:2", "checkpoint": "slots/2/checkpoint-7543", "step": 7543, "updated_at": 1776416725} diff --git a/slots/3/checkpoint-8335/config.json b/slots/3/checkpoint-8335/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/3/checkpoint-8335/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/3/checkpoint-8335/trainer_state.json b/slots/3/checkpoint-8335/trainer_state.json deleted file mode 100644 index 7b99c10a65b4cc296a1f9d3910d55af419d4be37..0000000000000000000000000000000000000000 --- a/slots/3/checkpoint-8335/trainer_state.json +++ /dev/null @@ -1,2953 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.489067427345171, - "eval_steps": 500, - "global_step": 8335, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.981147766113281, - "learning_rate": 4.999985708624277e-05, - "loss": 6.141958999633789, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.432202339172363, - "learning_rate": 4.999851618030373e-05, - "loss": 6.936581420898437, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 14.735859870910645, - "learning_rate": 4.9995763869538124e-05, - "loss": 7.491865539550782, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 12.628633499145508, - "learning_rate": 4.999160030934015e-05, - "loss": 7.847127532958984, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 14.50905990600586, - "learning_rate": 4.9986025734782515e-05, - "loss": 6.550693511962891, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 18.310335159301758, - "learning_rate": 4.997904046060316e-05, - "loss": 6.068325042724609, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 12.4039888381958, - "learning_rate": 4.997064488118748e-05, - "loss": 5.966919708251953, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 14.189772605895996, - "learning_rate": 4.996083947054607e-05, - "loss": 5.816884613037109, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.678384780883789, - "learning_rate": 4.9949624782287944e-05, - "loss": 6.372280120849609, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 13.517382621765137, - "learning_rate": 4.993700144958927e-05, - "loss": 5.818735504150391, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 11.332883834838867, - "learning_rate": 4.992297018515767e-05, - "loss": 5.053961181640625, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 29.601367950439453, - "learning_rate": 4.990753178119193e-05, - "loss": 5.2946617126464846, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.445046424865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.721809005737304, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 14.08784294128418, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.68652572631836, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 10.382508277893066, - "learning_rate": 4.985278284547495e-05, - "loss": 4.90959358215332, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 9.340018272399902, - "learning_rate": 4.983172539352465e-05, - "loss": 4.94355354309082, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 11.943198204040527, - "learning_rate": 4.980926595367956e-05, - "loss": 4.442295074462891, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.623554229736328, - "learning_rate": 4.978540579398941e-05, - "loss": 4.63629150390625, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 9.113373756408691, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.220952987670898, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.733516693115234, - "learning_rate": 4.973348878261681e-05, - "loss": 4.143902206420899, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.420555114746094, - "learning_rate": 4.970543486214516e-05, - "loss": 4.141384506225586, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 7.3338212966918945, - "learning_rate": 4.967598608408454e-05, - "loss": 3.72052001953125, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 10.468330383300781, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.4560214996337892, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 10.967416763305664, - "learning_rate": 4.961291068451385e-05, - "loss": 3.2705413818359377, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 12.951521873474121, - "learning_rate": 4.9579287624212093e-05, - "loss": 3.3012428283691406, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 9.387336730957031, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.2594974517822264, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.693846702575684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.9371822357177733, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 11.925010681152344, - "learning_rate": 4.94701000160798e-05, - "loss": 2.991471290588379, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 9.22618579864502, - "learning_rate": 4.943093818728689e-05, - "loss": 3.6338191986083985, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 9.440751075744629, - "learning_rate": 4.939039699886213e-05, - "loss": 2.745065689086914, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.722249984741211, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.652340126037598, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 10.900734901428223, - "learning_rate": 4.930518577661388e-05, - "loss": 2.8322593688964846, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 6.965799808502197, - "learning_rate": 4.926052055377716e-05, - "loss": 2.7872573852539064, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 6.74927282333374, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9263822555541994, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 8.862627029418945, - "learning_rate": 4.916708349342674e-05, - "loss": 2.499791717529297, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 10.037351608276367, - "learning_rate": 4.911831693132686e-05, - "loss": 2.275549125671387, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 6.2945075035095215, - "learning_rate": 4.906818866004871e-05, - "loss": 2.404999542236328, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 4.414289951324463, - "learning_rate": 4.901670150981164e-05, - "loss": 2.1930471420288087, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 5.771992206573486, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.3143028259277343, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.430806636810303, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1435062408447267, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 4.818403244018555, - "learning_rate": 4.885411623737651e-05, - "loss": 1.8850128173828125, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 6.249518394470215, - "learning_rate": 4.87972234054431e-05, - "loss": 2.0096624374389647, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.708676815032959, - "learning_rate": 4.87389869931255e-05, - "loss": 1.9306636810302735, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 10.92924690246582, - "learning_rate": 4.8679410288425e-05, - "loss": 1.60665283203125, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 8.802605628967285, - "learning_rate": 4.86184966550152e-05, - "loss": 1.6958501815795899, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 11.568385124206543, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.7447893142700195, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.640422344207764, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.7216526031494142, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.1430816650390625, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9419631958007812, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 4.505365371704102, - "learning_rate": 4.836154274553069e-05, - "loss": 1.4452963829040528, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 4.836411952972412, - "learning_rate": 4.829399755867636e-05, - "loss": 1.5338068962097169, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 6.093108177185059, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.553403663635254, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 6.984614372253418, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.5232481956481934, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.966179370880127, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.9450897216796874, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.660058975219727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4149410247802734, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.126588821411133, - "learning_rate": 4.793662296282393e-05, - "loss": 1.0894225120544434, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 9.536477088928223, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.519902515411377, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 5.378171443939209, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1999958038330079, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 4.443706512451172, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2191826820373535, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 4.879043102264404, - "learning_rate": 4.762739032415868e-05, - "loss": 1.2340889930725099, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.470728874206543, - "learning_rate": 4.754687724066917e-05, - "loss": 1.423619270324707, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 5.47915506362915, - "learning_rate": 4.746509117076991e-05, - "loss": 1.251408863067627, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 4.566497325897217, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.0826839447021483, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 6.905686855316162, - "learning_rate": 4.72977186137705e-05, - "loss": 1.069216251373291, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 2.964932680130005, - "learning_rate": 4.721214157644861e-05, - "loss": 1.1391191482543945, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 6.019253253936768, - "learning_rate": 4.712531045173996e-05, - "loss": 0.9639101028442383, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 3.0169851779937744, - "learning_rate": 4.703723014209025e-05, - "loss": 0.934177017211914, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 5.5374369621276855, - "learning_rate": 4.694790562047361e-05, - "loss": 0.921787166595459, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 3.7313361167907715, - "learning_rate": 4.685734193011182e-05, - "loss": 0.9222763061523438, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 3.172354221343994, - "learning_rate": 4.676554418418953e-05, - "loss": 0.9454127311706543, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 2.3172996044158936, - "learning_rate": 4.667251756556563e-05, - "loss": 1.050184154510498, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.589139938354492, - "learning_rate": 4.657826732648061e-05, - "loss": 0.7619458675384522, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 4.144975662231445, - "learning_rate": 4.648279878825998e-05, - "loss": 0.8281337738037109, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.205533266067505, - "learning_rate": 4.63861173410139e-05, - "loss": 0.8384788513183594, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673986911773682, - "learning_rate": 4.628822844333278e-05, - "loss": 0.7749058723449707, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 4.221221446990967, - "learning_rate": 4.618913762197918e-05, - "loss": 1.0583919525146483, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.1249191761016846, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.6937400341033936, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 3.2157654762268066, - "learning_rate": 4.59873726542891e-05, - "loss": 0.7503646850585938, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 2.9630126953125, - "learning_rate": 4.58847098995107e-05, - "loss": 0.7044057846069336, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 2.9990007877349854, - "learning_rate": 4.578086800353284e-05, - "loss": 0.6530625343322753, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 2.869370698928833, - "learning_rate": 4.56758528292216e-05, - "loss": 0.6877331733703613, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 2.248493194580078, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.7664458751678467, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 2.4392411708831787, - "learning_rate": 4.546232642794249e-05, - "loss": 0.6721152305603028, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 1.9127728939056396, - "learning_rate": 4.5353827256577894e-05, - "loss": 0.6500507354736328, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 2.9150123596191406, - "learning_rate": 4.52441789174058e-05, - "loss": 0.6752653598785401, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 4.481291770935059, - "learning_rate": 4.5133387601121476e-05, - "loss": 0.6834932327270508, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 1.990797519683838, - "learning_rate": 4.502145956295211e-05, - "loss": 0.6353804588317871, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 6.725353717803955, - "learning_rate": 4.4908401122303705e-05, - "loss": 0.6204021453857422, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 3.649500846862793, - "learning_rate": 4.4794218662404296e-05, - "loss": 0.6903295516967773, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 3.0070176124572754, - "learning_rate": 4.46789186299435e-05, - "loss": 0.5644550800323487, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 3.847533941268921, - "learning_rate": 4.4562507534708574e-05, - "loss": 0.6999043464660645, - "step": 8100 - }, - { - "epoch": 3.399042173167764, - "grad_norm": 3.3568928241729736, - "learning_rate": 4.444499194921691e-05, - "loss": 0.5565146446228028, - "step": 8120 - }, - { - "epoch": 3.4074166154168246, - "grad_norm": 3.4787251949310303, - "learning_rate": 4.4326378508344854e-05, - "loss": 0.6062619686126709, - "step": 8140 - }, - { - "epoch": 3.415791057665886, - "grad_norm": 9.04430866241455, - "learning_rate": 4.4206673908953235e-05, - "loss": 0.4710696220397949, - "step": 8160 - }, - { - "epoch": 3.424165499914947, - "grad_norm": 2.6757805347442627, - "learning_rate": 4.408588490950914e-05, - "loss": 0.5070647716522216, - "step": 8180 - }, - { - "epoch": 3.4325399421640084, - "grad_norm": 8.834247589111328, - "learning_rate": 4.3964018329704414e-05, - "loss": 0.8732002258300782, - "step": 8200 - }, - { - "epoch": 3.440914384413069, - "grad_norm": 5.973219871520996, - "learning_rate": 4.3841081050070595e-05, - "loss": 0.5055295944213867, - "step": 8220 - }, - { - "epoch": 3.4492888266621304, - "grad_norm": 9.602012634277344, - "learning_rate": 4.371708001159045e-05, - "loss": 0.6795970439910889, - "step": 8240 - }, - { - "epoch": 3.4576632689111917, - "grad_norm": 1.5608420372009277, - "learning_rate": 4.359202221530606e-05, - "loss": 0.5012212753295898, - "step": 8260 - }, - { - "epoch": 3.466037711160253, - "grad_norm": 6.789797306060791, - "learning_rate": 4.34659147219236e-05, - "loss": 0.5281818866729736, - "step": 8280 - }, - { - "epoch": 3.4744121534093138, - "grad_norm": 4.547278881072998, - "learning_rate": 4.333876465141465e-05, - "loss": 0.5114842891693115, - "step": 8300 - }, - { - "epoch": 3.482786595658375, - "grad_norm": 1.368796467781067, - "learning_rate": 4.321057918261423e-05, - "loss": 0.40313048362731935, - "step": 8320 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.9599911429619456e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/3/latest.json b/slots/3/latest.json deleted file mode 100644 index 89cae2b09606dcd85ec52cda723920a4bfd4ec97..0000000000000000000000000000000000000000 --- a/slots/3/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:3", "checkpoint": "slots/3/checkpoint-8335", "step": 8335, "updated_at": 1776417901} diff --git a/slots/4/checkpoint-7947/config.json b/slots/4/checkpoint-7947/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/4/checkpoint-7947/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/4/checkpoint-7947/trainer_state.json b/slots/4/checkpoint-7947/trainer_state.json deleted file mode 100644 index 5bd90cb82c665697a67d36d8b6fd266293e29c8a..0000000000000000000000000000000000000000 --- a/slots/4/checkpoint-7947/trainer_state.json +++ /dev/null @@ -1,2820 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.326603247713385, - "eval_steps": 500, - "global_step": 7947, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 11.95787525177002, - "learning_rate": 4.999985708624277e-05, - "loss": 5.881575012207032, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 11.939610481262207, - "learning_rate": 4.999851618030373e-05, - "loss": 5.647177886962891, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 48.17500305175781, - "learning_rate": 4.9995763869538124e-05, - "loss": 6.1393989562988285, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 14.723703384399414, - "learning_rate": 4.999160030934015e-05, - "loss": 5.649447250366211, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 14.260713577270508, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.456048202514649, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 12.285365104675293, - "learning_rate": 4.997904046060316e-05, - "loss": 5.332961654663086, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 33.52470016479492, - "learning_rate": 4.997064488118748e-05, - "loss": 6.02587776184082, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 17.36154556274414, - "learning_rate": 4.996083947054607e-05, - "loss": 5.352649688720703, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.644362449645996, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.589469528198242, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 14.174250602722168, - "learning_rate": 4.993700144958927e-05, - "loss": 5.2284095764160154, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 14.923873901367188, - "learning_rate": 4.992297018515767e-05, - "loss": 4.571537780761719, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 13.855690956115723, - "learning_rate": 4.990753178119193e-05, - "loss": 5.206478500366211, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 16.607261657714844, - "learning_rate": 4.989068710933732e-05, - "loss": 5.598894882202148, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 10.553526878356934, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.615887832641602, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 9.943771362304688, - "learning_rate": 4.985278284547495e-05, - "loss": 3.9751415252685547, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 11.704524993896484, - "learning_rate": 4.983172539352465e-05, - "loss": 4.089920425415039, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 13.140851020812988, - "learning_rate": 4.980926595367956e-05, - "loss": 4.461263275146484, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.009072303771973, - "learning_rate": 4.978540579398941e-05, - "loss": 3.8523796081542967, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 9.157245635986328, - "learning_rate": 4.9760146261587936e-05, - "loss": 3.737962341308594, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 11.106775283813477, - "learning_rate": 4.973348878261681e-05, - "loss": 4.242771148681641, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.31446361541748, - "learning_rate": 4.970543486214516e-05, - "loss": 3.7330020904541015, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 8.840044975280762, - "learning_rate": 4.967598608408454e-05, - "loss": 3.2959358215332033, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 8.993168830871582, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.447793197631836, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 9.880019187927246, - "learning_rate": 4.961291068451385e-05, - "loss": 3.674238586425781, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 10.820444107055664, - "learning_rate": 4.9579287624212093e-05, - "loss": 3.376015472412109, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 7.960184574127197, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.3296524047851563, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 7.695173740386963, - "learning_rate": 4.9507880274181854e-05, - "loss": 3.172230911254883, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 9.787315368652344, - "learning_rate": 4.94701000160798e-05, - "loss": 3.032790946960449, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 7.822601318359375, - "learning_rate": 4.943093818728689e-05, - "loss": 3.7993682861328124, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 6.423282146453857, - "learning_rate": 4.939039699886213e-05, - "loss": 3.1471582412719727, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 8.177947044372559, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.807335090637207, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 12.727806091308594, - "learning_rate": 4.930518577661388e-05, - "loss": 2.70469970703125, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 12.240203857421875, - "learning_rate": 4.926052055377716e-05, - "loss": 2.772799491882324, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 10.889534950256348, - "learning_rate": 4.921448559301048e-05, - "loss": 2.525356674194336, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 10.73294734954834, - "learning_rate": 4.916708349342674e-05, - "loss": 2.575057792663574, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 6.766477584838867, - "learning_rate": 4.911831693132686e-05, - "loss": 2.450628471374512, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 6.866926193237305, - "learning_rate": 4.906818866004871e-05, - "loss": 2.294377899169922, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 13.464261054992676, - "learning_rate": 4.901670150981164e-05, - "loss": 2.1687923431396485, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 8.798266410827637, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1829912185668947, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.770901679992676, - "learning_rate": 4.890966227678239e-05, - "loss": 1.9910266876220704, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 6.037090301513672, - "learning_rate": 4.885411623737651e-05, - "loss": 2.093680572509766, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 6.327531814575195, - "learning_rate": 4.87972234054431e-05, - "loss": 2.20494384765625, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 5.572488307952881, - "learning_rate": 4.87389869931255e-05, - "loss": 1.9014860153198243, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 6.42570161819458, - "learning_rate": 4.8679410288425e-05, - "loss": 1.685124397277832, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 5.767203330993652, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7935949325561524, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 11.936880111694336, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.7972061157226562, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.11804723739624, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.7099994659423827, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 4.457609176635742, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.8403182983398438, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 6.604846954345703, - "learning_rate": 4.836154274553069e-05, - "loss": 1.799737548828125, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 4.532509803771973, - "learning_rate": 4.829399755867636e-05, - "loss": 1.677235984802246, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 4.908101558685303, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.6233207702636718, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 4.57976770401001, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.4081362724304198, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.677663326263428, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.475815200805664, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 5.180493354797363, - "learning_rate": 4.801070437462659e-05, - "loss": 1.256404972076416, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 5.710392475128174, - "learning_rate": 4.793662296282393e-05, - "loss": 1.344097900390625, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 14.397268295288086, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.3774223327636719, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 5.110644340515137, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.3270224571228026, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 7.093288898468018, - "learning_rate": 4.770662587550643e-05, - "loss": 1.4591200828552247, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 4.576109409332275, - "learning_rate": 4.762739032415868e-05, - "loss": 1.40781831741333, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 8.339225769042969, - "learning_rate": 4.754687724066917e-05, - "loss": 1.479560661315918, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 7.494958877563477, - "learning_rate": 4.746509117076991e-05, - "loss": 1.3163747787475586, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 6.01884651184082, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.2254773139953614, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 7.199056625366211, - "learning_rate": 4.72977186137705e-05, - "loss": 1.0894113540649415, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 8.82214069366455, - "learning_rate": 4.721214157644861e-05, - "loss": 1.3237985610961913, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 4.850705146789551, - "learning_rate": 4.712531045173996e-05, - "loss": 0.8934127807617187, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 5.598201751708984, - "learning_rate": 4.703723014209025e-05, - "loss": 1.342045211791992, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 5.424136161804199, - "learning_rate": 4.694790562047361e-05, - "loss": 1.1706354141235351, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 7.543458461761475, - "learning_rate": 4.685734193011182e-05, - "loss": 0.9434962272644043, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 2.857658624649048, - "learning_rate": 4.676554418418953e-05, - "loss": 0.989952278137207, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 6.373669147491455, - "learning_rate": 4.667251756556563e-05, - "loss": 0.8908876419067383, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 6.381263256072998, - "learning_rate": 4.657826732648061e-05, - "loss": 1.0061418533325195, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.1551015377044678, - "learning_rate": 4.648279878825998e-05, - "loss": 0.7668521881103516, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.2378339767456055, - "learning_rate": 4.63861173410139e-05, - "loss": 0.9835851669311524, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.329948425292969, - "learning_rate": 4.628822844333278e-05, - "loss": 0.9755271911621094, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 3.7822136878967285, - "learning_rate": 4.618913762197918e-05, - "loss": 0.9501211166381835, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.9014272689819336, - "learning_rate": 4.6088850471575695e-05, - "loss": 0.7127848625183105, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 5.8602190017700195, - "learning_rate": 4.59873726542891e-05, - "loss": 0.9189926147460937, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 2.9552366733551025, - "learning_rate": 4.58847098995107e-05, - "loss": 0.8160135269165039, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 4.013094425201416, - "learning_rate": 4.578086800353284e-05, - "loss": 0.8563443183898926, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 2.5212459564208984, - "learning_rate": 4.56758528292216e-05, - "loss": 0.6215355873107911, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 4.694630146026611, - "learning_rate": 4.5569670305685866e-05, - "loss": 0.6466864585876465, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 2.3018579483032227, - "learning_rate": 4.546232642794249e-05, - "loss": 0.9325425148010253, - "step": 7940 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.863517505157504e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/4/latest.json b/slots/4/latest.json deleted file mode 100644 index 56ff5371264d650f3b363353a2514e471fdbefa5..0000000000000000000000000000000000000000 --- a/slots/4/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:4", "checkpoint": "slots/4/checkpoint-7947", "step": 7947, "updated_at": 1776417202} diff --git a/slots/5/checkpoint-7783/config.json b/slots/5/checkpoint-7783/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/5/checkpoint-7783/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/5/checkpoint-7783/trainer_state.json b/slots/5/checkpoint-7783/trainer_state.json deleted file mode 100644 index a65ee401ca881b0639858786b1574bbe679c1cdc..0000000000000000000000000000000000000000 --- a/slots/5/checkpoint-7783/trainer_state.json +++ /dev/null @@ -1,2764 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.2579328212710834, - "eval_steps": 500, - "global_step": 7783, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.268299102783203, - "learning_rate": 4.999985708624277e-05, - "loss": 4.392966461181641, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 23.723892211914062, - "learning_rate": 4.999851618030373e-05, - "loss": 5.242445755004883, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 15.284976959228516, - "learning_rate": 4.9995763869538124e-05, - "loss": 4.878763198852539, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 21.47739601135254, - "learning_rate": 4.999160030934015e-05, - "loss": 5.070553588867187, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.51418399810791, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.079425048828125, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 17.037799835205078, - "learning_rate": 4.997904046060316e-05, - "loss": 5.397942352294922, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 11.956343650817871, - "learning_rate": 4.997064488118748e-05, - "loss": 5.866295242309571, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 15.19040298461914, - "learning_rate": 4.996083947054607e-05, - "loss": 6.805714416503906, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 12.783180236816406, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.310309219360351, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 15.791704177856445, - "learning_rate": 4.993700144958927e-05, - "loss": 5.249310302734375, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 15.071763038635254, - "learning_rate": 4.992297018515767e-05, - "loss": 5.418260955810547, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 19.809253692626953, - "learning_rate": 4.990753178119193e-05, - "loss": 4.681427383422852, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 12.614968299865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.647319412231445, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.064179420471191, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.246531677246094, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 15.25048828125, - "learning_rate": 4.985278284547495e-05, - "loss": 4.512583541870117, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 14.296823501586914, - "learning_rate": 4.983172539352465e-05, - "loss": 4.193087005615235, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 10.55685043334961, - "learning_rate": 4.980926595367956e-05, - "loss": 4.492794799804687, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 15.188244819641113, - "learning_rate": 4.978540579398941e-05, - "loss": 4.525092697143554, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 23.44154167175293, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.181382751464843, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 11.647577285766602, - "learning_rate": 4.973348878261681e-05, - "loss": 4.1204273223876955, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.451223373413086, - "learning_rate": 4.970543486214516e-05, - "loss": 3.5496986389160154, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 15.924224853515625, - "learning_rate": 4.967598608408454e-05, - "loss": 3.6359764099121095, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 14.668912887573242, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.783356857299805, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 13.026039123535156, - "learning_rate": 4.961291068451385e-05, - "loss": 3.3328643798828126, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 14.517637252807617, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.825040817260742, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 16.220218658447266, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.251118469238281, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 10.804553985595703, - "learning_rate": 4.9507880274181854e-05, - "loss": 3.2774829864501953, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 16.004716873168945, - "learning_rate": 4.94701000160798e-05, - "loss": 3.2009490966796874, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 8.181561470031738, - "learning_rate": 4.943093818728689e-05, - "loss": 2.7726932525634767, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 10.360410690307617, - "learning_rate": 4.939039699886213e-05, - "loss": 3.2188484191894533, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.99072265625, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.66240234375, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 16.23861312866211, - "learning_rate": 4.930518577661388e-05, - "loss": 2.855569839477539, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 9.361970901489258, - "learning_rate": 4.926052055377716e-05, - "loss": 2.8716863632202148, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 10.665338516235352, - "learning_rate": 4.921448559301048e-05, - "loss": 2.866599464416504, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 6.958986759185791, - "learning_rate": 4.916708349342674e-05, - "loss": 2.719050407409668, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 15.789422035217285, - "learning_rate": 4.911831693132686e-05, - "loss": 2.807525062561035, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 11.294109344482422, - "learning_rate": 4.906818866004871e-05, - "loss": 3.0798269271850587, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 9.417093276977539, - "learning_rate": 4.901670150981164e-05, - "loss": 2.639542579650879, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 9.54472827911377, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.214591217041016, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.778777599334717, - "learning_rate": 4.890966227678239e-05, - "loss": 2.497542381286621, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 18.648605346679688, - "learning_rate": 4.885411623737651e-05, - "loss": 2.118986701965332, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 8.938231468200684, - "learning_rate": 4.87972234054431e-05, - "loss": 2.4246755599975587, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 14.108246803283691, - "learning_rate": 4.87389869931255e-05, - "loss": 2.476232147216797, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 20.093215942382812, - "learning_rate": 4.8679410288425e-05, - "loss": 2.005863571166992, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 6.399996280670166, - "learning_rate": 4.86184966550152e-05, - "loss": 2.176909637451172, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 6.5276103019714355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.833466148376465, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 7.747783660888672, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.028280258178711, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 8.722662925720215, - "learning_rate": 4.8427768950332264e-05, - "loss": 2.0328079223632813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 7.213231563568115, - "learning_rate": 4.836154274553069e-05, - "loss": 1.7321701049804688, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 8.361398696899414, - "learning_rate": 4.829399755867636e-05, - "loss": 1.7763862609863281, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 4.578886032104492, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.8702903747558595, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 7.65287446975708, - "learning_rate": 4.8154965567345073e-05, - "loss": 2.1905149459838866, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.528337478637695, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.5154589653015136, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.492247581481934, - "learning_rate": 4.801070437462659e-05, - "loss": 1.6617794036865234, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 7.519628524780273, - "learning_rate": 4.793662296282393e-05, - "loss": 1.5524075508117676, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.682668685913086, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.613736915588379, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 8.832022666931152, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.5550686836242675, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.3445448875427246, - "learning_rate": 4.770662587550643e-05, - "loss": 1.4483519554138184, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 8.750604629516602, - "learning_rate": 4.762739032415868e-05, - "loss": 1.5183201789855958, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.896105766296387, - "learning_rate": 4.754687724066917e-05, - "loss": 1.298677349090576, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 6.540890693664551, - "learning_rate": 4.746509117076991e-05, - "loss": 1.3707378387451172, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 5.634416580200195, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.2700369834899903, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 5.313226699829102, - "learning_rate": 4.72977186137705e-05, - "loss": 1.3874421119689941, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 4.164125919342041, - "learning_rate": 4.721214157644861e-05, - "loss": 1.3053173065185546, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 3.9634926319122314, - "learning_rate": 4.712531045173996e-05, - "loss": 1.1017135620117187, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 19.879589080810547, - "learning_rate": 4.703723014209025e-05, - "loss": 1.5074621200561524, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 7.186710834503174, - "learning_rate": 4.694790562047361e-05, - "loss": 1.3602344512939453, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 6.228899955749512, - "learning_rate": 4.685734193011182e-05, - "loss": 1.097489070892334, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 4.770694255828857, - "learning_rate": 4.676554418418953e-05, - "loss": 1.1659790992736816, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 4.967444896697998, - "learning_rate": 4.667251756556563e-05, - "loss": 1.1331518173217774, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.312889099121094, - "learning_rate": 4.657826732648061e-05, - "loss": 1.2721074104309082, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.759788751602173, - "learning_rate": 4.648279878825998e-05, - "loss": 0.9295409202575684, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 7.215264797210693, - "learning_rate": 4.63861173410139e-05, - "loss": 1.196054744720459, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673015117645264, - "learning_rate": 4.628822844333278e-05, - "loss": 1.0879426002502441, - "step": 7780 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.822739988147379e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/5/latest.json b/slots/5/latest.json deleted file mode 100644 index 1eb803087291903f0488eaa256816f38b0f810f3..0000000000000000000000000000000000000000 --- a/slots/5/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:5", "checkpoint": "slots/5/checkpoint-7783", "step": 7783, "updated_at": 1776416933} diff --git a/slots/6/checkpoint-7465/config.json b/slots/6/checkpoint-7465/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/6/checkpoint-7465/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/6/checkpoint-7465/trainer_state.json b/slots/6/checkpoint-7465/trainer_state.json deleted file mode 100644 index 677fc53d8223273d088abbb3103405a72faf8be0..0000000000000000000000000000000000000000 --- a/slots/6/checkpoint-7465/trainer_state.json +++ /dev/null @@ -1,2652 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.343876659764272, - "eval_steps": 500, - "global_step": 7465, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 31.780010223388672, - "learning_rate": 4.999985708624277e-05, - "loss": 47.03160400390625, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 22.647113800048828, - "learning_rate": 4.999851618030373e-05, - "loss": 39.12403564453125, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 24.094038009643555, - "learning_rate": 4.9995763869538124e-05, - "loss": 32.462225341796874, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 25.049236297607422, - "learning_rate": 4.999160030934015e-05, - "loss": 25.110198974609375, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 20.87278938293457, - "learning_rate": 4.9986025734782515e-05, - "loss": 24.496664428710936, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 19.164220809936523, - "learning_rate": 4.997904046060316e-05, - "loss": 21.625892639160156, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 17.22857666015625, - "learning_rate": 4.997064488118748e-05, - "loss": 18.14871826171875, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 17.680524826049805, - "learning_rate": 4.996083947054607e-05, - "loss": 19.17234649658203, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 17.1239070892334, - "learning_rate": 4.9949624782287944e-05, - "loss": 17.536575317382812, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 15.556852340698242, - "learning_rate": 4.993700144958927e-05, - "loss": 14.654452514648437, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.166357040405273, - "learning_rate": 4.992297018515767e-05, - "loss": 12.504849243164063, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 15.202767372131348, - "learning_rate": 4.990753178119193e-05, - "loss": 14.847886657714843, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 14.017868995666504, - "learning_rate": 4.989068710933732e-05, - "loss": 10.980046844482422, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 16.90302848815918, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.107038879394532, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 26.98078155517578, - "learning_rate": 4.985278284547495e-05, - "loss": 10.064899444580078, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 19.16476058959961, - "learning_rate": 4.983172539352465e-05, - "loss": 9.747372436523438, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 14.147124290466309, - "learning_rate": 4.980926595367956e-05, - "loss": 7.905096435546875, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 16.14385986328125, - "learning_rate": 4.978540579398941e-05, - "loss": 8.444129943847656, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 13.877010345458984, - "learning_rate": 4.9760146261587936e-05, - "loss": 6.83173828125, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 13.799625396728516, - "learning_rate": 4.973348878261681e-05, - "loss": 7.934703826904297, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 15.579751014709473, - "learning_rate": 4.970543486214516e-05, - "loss": 6.793711853027344, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 13.521928787231445, - "learning_rate": 4.967598608408454e-05, - "loss": 7.2837577819824215, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 15.898303031921387, - "learning_rate": 4.9645144111099515e-05, - "loss": 6.6693672180175785, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 24.22979164123535, - "learning_rate": 4.961291068451385e-05, - "loss": 5.961820983886719, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 11.246175765991211, - "learning_rate": 4.9579287624212093e-05, - "loss": 4.978904724121094, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 24.801748275756836, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.248088073730469, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.347773551940918, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.6643627166748045, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 12.957473754882812, - "learning_rate": 4.94701000160798e-05, - "loss": 4.121147918701172, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.496082305908203, - "learning_rate": 4.943093818728689e-05, - "loss": 3.8258975982666015, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 12.250146865844727, - "learning_rate": 4.939039699886213e-05, - "loss": 3.8255241394042967, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 12.506306648254395, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.5919628143310547, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 16.467334747314453, - "learning_rate": 4.930518577661388e-05, - "loss": 3.715103530883789, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 8.267026901245117, - "learning_rate": 4.926052055377716e-05, - "loss": 3.194218063354492, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 9.150392532348633, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9999900817871095, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 11.638959884643555, - "learning_rate": 4.916708349342674e-05, - "loss": 2.837971496582031, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 10.091157913208008, - "learning_rate": 4.911831693132686e-05, - "loss": 2.7495031356811523, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 9.66382884979248, - "learning_rate": 4.906818866004871e-05, - "loss": 2.6875839233398438, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 8.437216758728027, - "learning_rate": 4.901670150981164e-05, - "loss": 2.5731342315673826, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 8.248567581176758, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.859779930114746, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 11.652517318725586, - "learning_rate": 4.890966227678239e-05, - "loss": 2.2908452987670898, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 14.507343292236328, - "learning_rate": 4.885411623737651e-05, - "loss": 2.302378463745117, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 9.44881820678711, - "learning_rate": 4.87972234054431e-05, - "loss": 2.3078046798706056, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 9.367117881774902, - "learning_rate": 4.87389869931255e-05, - "loss": 2.1523136138916015, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 8.667911529541016, - "learning_rate": 4.8679410288425e-05, - "loss": 2.2135799407958983, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 16.578901290893555, - "learning_rate": 4.86184966550152e-05, - "loss": 2.1074777603149415, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.302729606628418, - "learning_rate": 4.8556249532052045e-05, - "loss": 2.0359333038330076, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 20.793943405151367, - "learning_rate": 4.8492672433979744e-05, - "loss": 2.2183906555175783, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.577836990356445, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9698509216308593, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 5.63059139251709, - "learning_rate": 4.836154274553069e-05, - "loss": 1.652799415588379, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 6.654581069946289, - "learning_rate": 4.829399755867636e-05, - "loss": 1.740532112121582, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 11.970749855041504, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.6816909790039063, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 7.981494426727295, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.61510009765625, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 6.4518561363220215, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.6320558547973634, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 11.362939834594727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.474619197845459, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 4.846298694610596, - "learning_rate": 4.793662296282393e-05, - "loss": 1.4175973892211915, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 7.837019920349121, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.5572842597961425, - "step": 7420 - }, - { - "epoch": 2.33602559447285, - "grad_norm": 5.593219757080078, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.439277744293213, - "step": 7440 - }, - { - "epoch": 2.3423064467059875, - "grad_norm": 5.2232160568237305, - "learning_rate": 4.770662587550643e-05, - "loss": 1.3270212173461915, - "step": 7460 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.7436869283034368e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/6/latest.json b/slots/6/latest.json deleted file mode 100644 index d287010e1b901efab6001c558d60d69fab85378d..0000000000000000000000000000000000000000 --- a/slots/6/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:6", "checkpoint": "slots/6/checkpoint-7465", "step": 7465, "updated_at": 1776416394} diff --git a/slots/7/checkpoint-7428/config.json b/slots/7/checkpoint-7428/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/7/checkpoint-7428/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/7/checkpoint-7428/trainer_state.json b/slots/7/checkpoint-7428/trainer_state.json deleted file mode 100644 index 7e401e550e4930eb39694ac0c816d0bc0eff21f5..0000000000000000000000000000000000000000 --- a/slots/7/checkpoint-7428/trainer_state.json +++ /dev/null @@ -1,2638 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 2.3322570831329674, - "eval_steps": 500, - "global_step": 7428, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 1.984523587544285, - "grad_norm": 35.15928649902344, - "learning_rate": 4.999985708624277e-05, - "loss": 42.4168701171875, - "step": 6320 - }, - { - "epoch": 1.9908044397774223, - "grad_norm": 24.441553115844727, - "learning_rate": 4.999851618030373e-05, - "loss": 34.59369812011719, - "step": 6340 - }, - { - "epoch": 1.9970852920105597, - "grad_norm": 27.142696380615234, - "learning_rate": 4.9995763869538124e-05, - "loss": 29.539852905273438, - "step": 6360 - }, - { - "epoch": 2.003140426116569, - "grad_norm": 20.90863037109375, - "learning_rate": 4.999160030934015e-05, - "loss": 26.295675659179686, - "step": 6380 - }, - { - "epoch": 2.009421278349706, - "grad_norm": 19.69276237487793, - "learning_rate": 4.9986025734782515e-05, - "loss": 26.864825439453124, - "step": 6400 - }, - { - "epoch": 2.0157021305828433, - "grad_norm": 20.81488037109375, - "learning_rate": 4.997904046060316e-05, - "loss": 21.519651794433592, - "step": 6420 - }, - { - "epoch": 2.021982982815981, - "grad_norm": 19.697011947631836, - "learning_rate": 4.997064488118748e-05, - "loss": 20.201951599121095, - "step": 6440 - }, - { - "epoch": 2.0282638350491182, - "grad_norm": 19.01127815246582, - "learning_rate": 4.996083947054607e-05, - "loss": 18.92204895019531, - "step": 6460 - }, - { - "epoch": 2.0345446872822555, - "grad_norm": 21.61073112487793, - "learning_rate": 4.9949624782287944e-05, - "loss": 18.937400817871094, - "step": 6480 - }, - { - "epoch": 2.040825539515393, - "grad_norm": 23.090789794921875, - "learning_rate": 4.993700144958927e-05, - "loss": 15.520936584472656, - "step": 6500 - }, - { - "epoch": 2.0471063917485304, - "grad_norm": 17.096433639526367, - "learning_rate": 4.992297018515767e-05, - "loss": 15.487298583984375, - "step": 6520 - }, - { - "epoch": 2.0533872439816676, - "grad_norm": 21.39566421508789, - "learning_rate": 4.990753178119193e-05, - "loss": 14.596461486816406, - "step": 6540 - }, - { - "epoch": 2.0596680962148053, - "grad_norm": 16.463520050048828, - "learning_rate": 4.989068710933732e-05, - "loss": 12.615646362304688, - "step": 6560 - }, - { - "epoch": 2.0659489484479425, - "grad_norm": 14.016276359558105, - "learning_rate": 4.9872437120636304e-05, - "loss": 10.938156890869141, - "step": 6580 - }, - { - "epoch": 2.0722298006810798, - "grad_norm": 20.464670181274414, - "learning_rate": 4.985278284547495e-05, - "loss": 10.395332336425781, - "step": 6600 - }, - { - "epoch": 2.0785106529142174, - "grad_norm": 17.77933692932129, - "learning_rate": 4.983172539352465e-05, - "loss": 10.609977722167969, - "step": 6620 - }, - { - "epoch": 2.0847915051473547, - "grad_norm": 12.832728385925293, - "learning_rate": 4.980926595367956e-05, - "loss": 8.691991424560547, - "step": 6640 - }, - { - "epoch": 2.091072357380492, - "grad_norm": 13.371121406555176, - "learning_rate": 4.978540579398941e-05, - "loss": 8.057951354980469, - "step": 6660 - }, - { - "epoch": 2.0973532096136296, - "grad_norm": 16.301082611083984, - "learning_rate": 4.9760146261587936e-05, - "loss": 7.973057556152344, - "step": 6680 - }, - { - "epoch": 2.103634061846767, - "grad_norm": 12.96883487701416, - "learning_rate": 4.973348878261681e-05, - "loss": 7.619200897216797, - "step": 6700 - }, - { - "epoch": 2.109914914079904, - "grad_norm": 12.284865379333496, - "learning_rate": 4.970543486214516e-05, - "loss": 7.1647804260253904, - "step": 6720 - }, - { - "epoch": 2.1161957663130417, - "grad_norm": 21.732789993286133, - "learning_rate": 4.967598608408454e-05, - "loss": 6.469166564941406, - "step": 6740 - }, - { - "epoch": 2.122476618546179, - "grad_norm": 10.566523551940918, - "learning_rate": 4.9645144111099515e-05, - "loss": 5.752484130859375, - "step": 6760 - }, - { - "epoch": 2.128757470779316, - "grad_norm": 11.216104507446289, - "learning_rate": 4.961291068451385e-05, - "loss": 5.699852752685547, - "step": 6780 - }, - { - "epoch": 2.135038323012454, - "grad_norm": 9.675246238708496, - "learning_rate": 4.9579287624212093e-05, - "loss": 5.267290878295898, - "step": 6800 - }, - { - "epoch": 2.141319175245591, - "grad_norm": 11.593718528747559, - "learning_rate": 4.9544276828536906e-05, - "loss": 5.2138721466064455, - "step": 6820 - }, - { - "epoch": 2.1476000274787284, - "grad_norm": 11.523468971252441, - "learning_rate": 4.9507880274181854e-05, - "loss": 4.134304428100586, - "step": 6840 - }, - { - "epoch": 2.153880879711866, - "grad_norm": 10.710789680480957, - "learning_rate": 4.94701000160798e-05, - "loss": 4.5618743896484375, - "step": 6860 - }, - { - "epoch": 2.1601617319450033, - "grad_norm": 10.197118759155273, - "learning_rate": 4.943093818728689e-05, - "loss": 4.101953125, - "step": 6880 - }, - { - "epoch": 2.1664425841781405, - "grad_norm": 8.9686279296875, - "learning_rate": 4.939039699886213e-05, - "loss": 3.642832565307617, - "step": 6900 - }, - { - "epoch": 2.172723436411278, - "grad_norm": 8.167902946472168, - "learning_rate": 4.9348478739742526e-05, - "loss": 3.1743133544921873, - "step": 6920 - }, - { - "epoch": 2.1790042886444154, - "grad_norm": 9.323243141174316, - "learning_rate": 4.930518577661388e-05, - "loss": 3.6948711395263674, - "step": 6940 - }, - { - "epoch": 2.1852851408775527, - "grad_norm": 10.735540390014648, - "learning_rate": 4.926052055377716e-05, - "loss": 3.1577367782592773, - "step": 6960 - }, - { - "epoch": 2.1915659931106903, - "grad_norm": 12.350885391235352, - "learning_rate": 4.921448559301048e-05, - "loss": 3.0701583862304687, - "step": 6980 - }, - { - "epoch": 2.1978468453438276, - "grad_norm": 9.401058197021484, - "learning_rate": 4.916708349342674e-05, - "loss": 2.8100170135498046, - "step": 7000 - }, - { - "epoch": 2.204127697576965, - "grad_norm": 7.154083728790283, - "learning_rate": 4.911831693132686e-05, - "loss": 2.472774696350098, - "step": 7020 - }, - { - "epoch": 2.2104085498101025, - "grad_norm": 7.130964279174805, - "learning_rate": 4.906818866004871e-05, - "loss": 2.597759246826172, - "step": 7040 - }, - { - "epoch": 2.2166894020432397, - "grad_norm": 7.370577812194824, - "learning_rate": 4.901670150981164e-05, - "loss": 2.3146718978881835, - "step": 7060 - }, - { - "epoch": 2.222970254276377, - "grad_norm": 9.05713176727295, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.1284273147583006, - "step": 7080 - }, - { - "epoch": 2.2292511065095146, - "grad_norm": 7.408658027648926, - "learning_rate": 4.890966227678239e-05, - "loss": 2.052642822265625, - "step": 7100 - }, - { - "epoch": 2.235531958742652, - "grad_norm": 17.97675895690918, - "learning_rate": 4.885411623737651e-05, - "loss": 2.0922914505004884, - "step": 7120 - }, - { - "epoch": 2.241812810975789, - "grad_norm": 11.770293235778809, - "learning_rate": 4.87972234054431e-05, - "loss": 1.9368261337280273, - "step": 7140 - }, - { - "epoch": 2.248093663208927, - "grad_norm": 5.871363162994385, - "learning_rate": 4.87389869931255e-05, - "loss": 2.0074506759643556, - "step": 7160 - }, - { - "epoch": 2.254374515442064, - "grad_norm": 7.695994853973389, - "learning_rate": 4.8679410288425e-05, - "loss": 1.9740360260009766, - "step": 7180 - }, - { - "epoch": 2.2606553676752013, - "grad_norm": 5.52763557434082, - "learning_rate": 4.86184966550152e-05, - "loss": 1.7269515991210938, - "step": 7200 - }, - { - "epoch": 2.266936219908339, - "grad_norm": 5.56577205657959, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.6195665359497071, - "step": 7220 - }, - { - "epoch": 2.273217072141476, - "grad_norm": 7.687935829162598, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.576361083984375, - "step": 7240 - }, - { - "epoch": 2.2794979243746134, - "grad_norm": 7.840238094329834, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.704756736755371, - "step": 7260 - }, - { - "epoch": 2.285778776607751, - "grad_norm": 4.519632816314697, - "learning_rate": 4.836154274553069e-05, - "loss": 2.219845199584961, - "step": 7280 - }, - { - "epoch": 2.2920596288408883, - "grad_norm": 5.06703519821167, - "learning_rate": 4.829399755867636e-05, - "loss": 2.371290397644043, - "step": 7300 - }, - { - "epoch": 2.2983404810740256, - "grad_norm": 5.222914695739746, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.2122323036193847, - "step": 7320 - }, - { - "epoch": 2.3046213333071632, - "grad_norm": 3.4668195247650146, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.334171199798584, - "step": 7340 - }, - { - "epoch": 2.3109021855403005, - "grad_norm": 4.191354274749756, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2463009834289551, - "step": 7360 - }, - { - "epoch": 2.3171830377734377, - "grad_norm": 5.007280349731445, - "learning_rate": 4.801070437462659e-05, - "loss": 1.0995778083801269, - "step": 7380 - }, - { - "epoch": 2.3234638900065754, - "grad_norm": 5.577613830566406, - "learning_rate": 4.793662296282393e-05, - "loss": 1.1267754554748535, - "step": 7400 - }, - { - "epoch": 2.3297447422397126, - "grad_norm": 3.4261252880096436, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.1237424850463866, - "step": 7420 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 3, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.7344871226365184e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/7/latest.json b/slots/7/latest.json deleted file mode 100644 index 002bb88aa44f4fc98243fca74308e2999f7afdc8..0000000000000000000000000000000000000000 --- a/slots/7/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:7", "checkpoint": "slots/7/checkpoint-7428", "step": 7428, "updated_at": 1776416414} diff --git a/slots/8/checkpoint-8113/config.json b/slots/8/checkpoint-8113/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/8/checkpoint-8113/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/8/checkpoint-8113/trainer_state.json b/slots/8/checkpoint-8113/trainer_state.json deleted file mode 100644 index 1e553c35beeb690ee83b09fe99e580f9ab190bba..0000000000000000000000000000000000000000 --- a/slots/8/checkpoint-8113/trainer_state.json +++ /dev/null @@ -1,2876 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.3961111183805923, - "eval_steps": 500, - "global_step": 8113, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 10.619898796081543, - "learning_rate": 4.999985708624277e-05, - "loss": 5.514506530761719, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.31767463684082, - "learning_rate": 4.999851618030373e-05, - "loss": 5.064883041381836, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 13.286064147949219, - "learning_rate": 4.9995763869538124e-05, - "loss": 5.178463363647461, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 13.165849685668945, - "learning_rate": 4.999160030934015e-05, - "loss": 5.5684253692626955, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 12.71381950378418, - "learning_rate": 4.9986025734782515e-05, - "loss": 5.4804939270019535, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 15.386504173278809, - "learning_rate": 4.997904046060316e-05, - "loss": 5.793609619140625, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 30.56761360168457, - "learning_rate": 4.997064488118748e-05, - "loss": 5.199003982543945, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 10.182106971740723, - "learning_rate": 4.996083947054607e-05, - "loss": 5.1698661804199215, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.703635215759277, - "learning_rate": 4.9949624782287944e-05, - "loss": 5.147470855712891, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 17.958473205566406, - "learning_rate": 4.993700144958927e-05, - "loss": 4.737754058837891, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 16.8748779296875, - "learning_rate": 4.992297018515767e-05, - "loss": 4.211444091796875, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 46.32542037963867, - "learning_rate": 4.990753178119193e-05, - "loss": 4.386423492431641, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.126466751098633, - "learning_rate": 4.989068710933732e-05, - "loss": 4.631509017944336, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 12.159765243530273, - "learning_rate": 4.9872437120636304e-05, - "loss": 3.881015396118164, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 12.283050537109375, - "learning_rate": 4.985278284547495e-05, - "loss": 3.940886306762695, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 12.025839805603027, - "learning_rate": 4.983172539352465e-05, - "loss": 3.9277923583984373, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 12.16906452178955, - "learning_rate": 4.980926595367956e-05, - "loss": 4.246135330200195, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.341684341430664, - "learning_rate": 4.978540579398941e-05, - "loss": 4.455572891235351, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 8.899483680725098, - "learning_rate": 4.9760146261587936e-05, - "loss": 3.8522106170654298, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.554529190063477, - "learning_rate": 4.973348878261681e-05, - "loss": 3.7468845367431642, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 9.822662353515625, - "learning_rate": 4.970543486214516e-05, - "loss": 3.0998603820800783, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 10.157042503356934, - "learning_rate": 4.967598608408454e-05, - "loss": 3.316312789916992, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 9.876077651977539, - "learning_rate": 4.9645144111099515e-05, - "loss": 4.0845588684082035, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 11.132689476013184, - "learning_rate": 4.961291068451385e-05, - "loss": 3.1008926391601563, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 9.930462837219238, - "learning_rate": 4.9579287624212093e-05, - "loss": 2.9224714279174804, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 7.317378997802734, - "learning_rate": 4.9544276828536906e-05, - "loss": 2.853228759765625, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.085936546325684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.5410511016845705, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 7.214178085327148, - "learning_rate": 4.94701000160798e-05, - "loss": 3.0386709213256835, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 7.6133952140808105, - "learning_rate": 4.943093818728689e-05, - "loss": 2.6715574264526367, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 7.800808429718018, - "learning_rate": 4.939039699886213e-05, - "loss": 2.398478126525879, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 5.941168785095215, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.8046531677246094, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 14.817143440246582, - "learning_rate": 4.930518577661388e-05, - "loss": 2.6330236434936523, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 8.76217269897461, - "learning_rate": 4.926052055377716e-05, - "loss": 2.199250602722168, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 11.228110313415527, - "learning_rate": 4.921448559301048e-05, - "loss": 2.290775489807129, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 7.807727813720703, - "learning_rate": 4.916708349342674e-05, - "loss": 2.2031742095947267, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 8.53569507598877, - "learning_rate": 4.911831693132686e-05, - "loss": 1.9725076675415039, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 7.547488689422607, - "learning_rate": 4.906818866004871e-05, - "loss": 2.019377326965332, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 5.556341648101807, - "learning_rate": 4.901670150981164e-05, - "loss": 2.0958667755126954, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 10.818191528320312, - "learning_rate": 4.8963858387556665e-05, - "loss": 1.8875907897949218, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 5.718668460845947, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1760227203369142, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 12.06888484954834, - "learning_rate": 4.885411623737651e-05, - "loss": 1.980560302734375, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 5.500242233276367, - "learning_rate": 4.87972234054431e-05, - "loss": 1.7397878646850586, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.90056848526001, - "learning_rate": 4.87389869931255e-05, - "loss": 1.8117895126342773, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 4.671687126159668, - "learning_rate": 4.8679410288425e-05, - "loss": 1.8994810104370117, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 4.093003749847412, - "learning_rate": 4.86184966550152e-05, - "loss": 1.5177894592285157, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 7.3171610832214355, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.4882768630981444, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.505290985107422, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.8049379348754884, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.227847576141357, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.3910842895507813, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 6.138688564300537, - "learning_rate": 4.836154274553069e-05, - "loss": 1.5944181442260743, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 9.806740760803223, - "learning_rate": 4.829399755867636e-05, - "loss": 1.9773569107055664, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 17.26584243774414, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.3594069480895996, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 4.6048479080200195, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.365511989593506, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 3.5992016792297363, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.2431794166564942, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 3.939440965652466, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4344836235046388, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.212338447570801, - "learning_rate": 4.793662296282393e-05, - "loss": 1.340198040008545, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 5.5816969871521, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.0717330932617188, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 6.169701099395752, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1681714057922363, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 3.759722948074341, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2151671409606934, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 2.909806966781616, - "learning_rate": 4.762739032415868e-05, - "loss": 0.9399504661560059, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 5.237748146057129, - "learning_rate": 4.754687724066917e-05, - "loss": 1.2489306449890136, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 3.2770180702209473, - "learning_rate": 4.746509117076991e-05, - "loss": 1.25095853805542, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 2.7698371410369873, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.3378321647644043, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 3.8602523803710938, - "learning_rate": 4.72977186137705e-05, - "loss": 1.3043223381042481, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 2.804724931716919, - "learning_rate": 4.721214157644861e-05, - "loss": 1.0128710746765137, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 5.5239577293396, - "learning_rate": 4.712531045173996e-05, - "loss": 0.9921689033508301, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 11.35846996307373, - "learning_rate": 4.7041663749229935e-05, - "loss": 1.3766287803649901, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 10.031026840209961, - "learning_rate": 4.6952401319017915e-05, - "loss": 0.9078640937805176, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 2.9733171463012695, - "learning_rate": 4.686189946623564e-05, - "loss": 0.8636553764343262, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 4.90701150894165, - "learning_rate": 4.677016330057647e-05, - "loss": 0.8574177742004394, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 9.797515869140625, - "learning_rate": 4.667719800142248e-05, - "loss": 1.1090837478637696, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.111540794372559, - "learning_rate": 4.6583008817552094e-05, - "loss": 0.8992094039916992, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 3.002032518386841, - "learning_rate": 4.6487601066843674e-05, - "loss": 0.7496014595031738, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.6695992946624756, - "learning_rate": 4.639098013597534e-05, - "loss": 0.6838997840881348, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.092342853546143, - "learning_rate": 4.629315148012079e-05, - "loss": 0.689493465423584, - "step": 7780 - }, - { - "epoch": 3.265051097182785, - "grad_norm": 6.389557361602783, - "learning_rate": 4.619412062264134e-05, - "loss": 0.9438887596130371, - "step": 7800 - }, - { - "epoch": 3.2734255394318463, - "grad_norm": 2.32489275932312, - "learning_rate": 4.609389315477405e-05, - "loss": 0.7676928043365479, - "step": 7820 - }, - { - "epoch": 3.2817999816809076, - "grad_norm": 2.865983724594116, - "learning_rate": 4.599247473531607e-05, - "loss": 0.8114580154418946, - "step": 7840 - }, - { - "epoch": 3.290174423929969, - "grad_norm": 8.078557014465332, - "learning_rate": 4.588987109030511e-05, - "loss": 0.6942833423614502, - "step": 7860 - }, - { - "epoch": 3.29854886617903, - "grad_norm": 3.2191479206085205, - "learning_rate": 4.578608801269621e-05, - "loss": 0.827882194519043, - "step": 7880 - }, - { - "epoch": 3.306923308428091, - "grad_norm": 7.289218902587891, - "learning_rate": 4.5681131362034604e-05, - "loss": 0.8882099151611328, - "step": 7900 - }, - { - "epoch": 3.315297750677152, - "grad_norm": 2.5519909858703613, - "learning_rate": 4.5575007064124937e-05, - "loss": 0.6065804958343506, - "step": 7920 - }, - { - "epoch": 3.3236721929262134, - "grad_norm": 3.498563289642334, - "learning_rate": 4.5467721110696685e-05, - "loss": 0.730232048034668, - "step": 7940 - }, - { - "epoch": 3.332046635175274, - "grad_norm": 2.0245680809020996, - "learning_rate": 4.5359279559065876e-05, - "loss": 0.9999460220336914, - "step": 7960 - }, - { - "epoch": 3.3404210774243355, - "grad_norm": 2.994798183441162, - "learning_rate": 4.524968853179306e-05, - "loss": 0.6818597316741943, - "step": 7980 - }, - { - "epoch": 3.3487955196733967, - "grad_norm": 3.339902400970459, - "learning_rate": 4.51389542163377e-05, - "loss": 0.7384091854095459, - "step": 8000 - }, - { - "epoch": 3.357169961922458, - "grad_norm": 2.02184796333313, - "learning_rate": 4.5027082864708735e-05, - "loss": 0.6503865718841553, - "step": 8020 - }, - { - "epoch": 3.3655444041715192, - "grad_norm": 3.314213991165161, - "learning_rate": 4.491408079311169e-05, - "loss": 0.6390689849853516, - "step": 8040 - }, - { - "epoch": 3.37391884642058, - "grad_norm": 6.766301155090332, - "learning_rate": 4.479995438159201e-05, - "loss": 0.6914161205291748, - "step": 8060 - }, - { - "epoch": 3.3822932886696413, - "grad_norm": 2.9052584171295166, - "learning_rate": 4.468471007367485e-05, - "loss": 0.7612205028533936, - "step": 8080 - }, - { - "epoch": 3.3906677309187025, - "grad_norm": 3.555596351623535, - "learning_rate": 4.456835437600129e-05, - "loss": 0.5866696834564209, - "step": 8100 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.904792308960435e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/8/latest.json b/slots/8/latest.json deleted file mode 100644 index dda91b0731ac7f69ce2a3952eb8e5a5887940c16..0000000000000000000000000000000000000000 --- a/slots/8/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:8", "checkpoint": "slots/8/checkpoint-8113", "step": 8113, "updated_at": 1776417792} diff --git a/slots/9/checkpoint-7793/config.json b/slots/9/checkpoint-7793/config.json deleted file mode 100644 index 107138a904f38104a668de36ff574a982f8e1797..0000000000000000000000000000000000000000 --- a/slots/9/checkpoint-7793/config.json +++ /dev/null @@ -1,24 +0,0 @@ -{ - "architectures": [ - "TwinyForCausalLM" - ], - "attention_dropout": 0.0, - "dtype": "float32", - "hidden_dropout": 0.0, - "hidden_size": 768, - "initializer_range": 0.02, - "intermediate_size": 3072, - "max_position_embeddings": 128, - "model_type": "twiny", - "neftune_alpha": 0.0, - "num_attention_heads": 12, - "num_hidden_layers": 1, - "num_key_value_heads": 3, - "qk_norm": true, - "rezero_init": 1.0, - "rms_norm_eps": 1e-06, - "rope_theta": 10000.0, - "transformers_version": "5.0.0", - "use_cache": false, - "vocab_size": 32000 -} diff --git a/slots/9/checkpoint-7793/trainer_state.json b/slots/9/checkpoint-7793/trainer_state.json deleted file mode 100644 index b33647f008a2928c7d2ea38e6d6435c4185a1266..0000000000000000000000000000000000000000 --- a/slots/9/checkpoint-7793/trainer_state.json +++ /dev/null @@ -1,2764 +0,0 @@ -{ - "best_global_step": null, - "best_metric": null, - "best_model_checkpoint": null, - "epoch": 3.262120042395614, - "eval_steps": 500, - "global_step": 7793, - "is_hyper_param_search": false, - "is_local_process_zero": true, - "is_world_process_zero": true, - "log_history": [ - { - "epoch": 2.617016627215141e-05, - "grad_norm": 122.33340454101562, - "learning_rate": 5e-05, - "loss": 419.9996032714844, - "step": 1 - }, - { - "epoch": 0.0005234033254430282, - "grad_norm": 90.82410430908203, - "learning_rate": 4.999999511980259e-05, - "loss": 428.18197471217104, - "step": 20 - }, - { - "epoch": 0.0010468066508860564, - "grad_norm": 59.52641296386719, - "learning_rate": 4.999997943828395e-05, - "loss": 366.3947265625, - "step": 40 - }, - { - "epoch": 0.0015702099763290846, - "grad_norm": 41.61277389526367, - "learning_rate": 4.999995294193233e-05, - "loss": 325.6545166015625, - "step": 60 - }, - { - "epoch": 0.0020936133017721128, - "grad_norm": 46.67556381225586, - "learning_rate": 4.999991563075919e-05, - "loss": 310.3648193359375, - "step": 80 - }, - { - "epoch": 0.002617016627215141, - "grad_norm": 47.65824890136719, - "learning_rate": 4.999986750478065e-05, - "loss": 303.49443359375, - "step": 100 - }, - { - "epoch": 0.003140419952658169, - "grad_norm": 62.748138427734375, - "learning_rate": 4.999980856401755e-05, - "loss": 322.389990234375, - "step": 120 - }, - { - "epoch": 0.007327646556202395, - "grad_norm": 48.20169448852539, - "learning_rate": 4.999895523943923e-05, - "loss": 288.82497336647725, - "step": 140 - }, - { - "epoch": 0.008374453207088451, - "grad_norm": 68.66603088378906, - "learning_rate": 4.99986329623214e-05, - "loss": 293.87509765625, - "step": 160 - }, - { - "epoch": 0.009421259857974507, - "grad_norm": 61.69401931762695, - "learning_rate": 4.999826742820622e-05, - "loss": 324.10361328125, - "step": 180 - }, - { - "epoch": 0.010468066508860565, - "grad_norm": 56.778385162353516, - "learning_rate": 4.999785863772618e-05, - "loss": 274.875537109375, - "step": 200 - }, - { - "epoch": 0.01151487315974662, - "grad_norm": 71.3816146850586, - "learning_rate": 4.999740659158865e-05, - "loss": 281.357080078125, - "step": 220 - }, - { - "epoch": 0.012561679810632677, - "grad_norm": 75.08697509765625, - "learning_rate": 4.9996911290575844e-05, - "loss": 263.852783203125, - "step": 240 - }, - { - "epoch": 0.013608486461518732, - "grad_norm": 53.91341781616211, - "learning_rate": 4.999637273554481e-05, - "loss": 258.546630859375, - "step": 260 - }, - { - "epoch": 0.01465529311240479, - "grad_norm": 56.39162063598633, - "learning_rate": 4.9995790927427446e-05, - "loss": 270.5470703125, - "step": 280 - }, - { - "epoch": 0.015702099763290844, - "grad_norm": 62.93777084350586, - "learning_rate": 4.99951658672305e-05, - "loss": 252.8383056640625, - "step": 300 - }, - { - "epoch": 0.016748906414176902, - "grad_norm": 61.51578903198242, - "learning_rate": 4.9994497556035567e-05, - "loss": 287.8167724609375, - "step": 320 - }, - { - "epoch": 0.01779571306506296, - "grad_norm": 62.75235366821289, - "learning_rate": 4.9993785994999074e-05, - "loss": 258.673779296875, - "step": 340 - }, - { - "epoch": 0.018842519715949014, - "grad_norm": 56.2258186340332, - "learning_rate": 4.999303118535229e-05, - "loss": 252.9609375, - "step": 360 - }, - { - "epoch": 0.01988932636683507, - "grad_norm": 54.998130798339844, - "learning_rate": 4.9992233128401314e-05, - "loss": 241.9411376953125, - "step": 380 - }, - { - "epoch": 0.02093613301772113, - "grad_norm": 63.37522888183594, - "learning_rate": 4.999139182552709e-05, - "loss": 249.7280517578125, - "step": 400 - }, - { - "epoch": 0.021982939668607183, - "grad_norm": 60.007144927978516, - "learning_rate": 4.9990507278185374e-05, - "loss": 236.941015625, - "step": 420 - }, - { - "epoch": 0.02302974631949324, - "grad_norm": 66.06304168701172, - "learning_rate": 4.998957948790677e-05, - "loss": 240.55878363715277, - "step": 440 - }, - { - "epoch": 0.0240765529703793, - "grad_norm": 68.9736099243164, - "learning_rate": 4.998860845629671e-05, - "loss": 247.988720703125, - "step": 460 - }, - { - "epoch": 0.025123359621265353, - "grad_norm": 61.50578308105469, - "learning_rate": 4.998759418503544e-05, - "loss": 233.025537109375, - "step": 480 - }, - { - "epoch": 0.02617016627215141, - "grad_norm": 54.96660614013672, - "learning_rate": 4.998653667587801e-05, - "loss": 243.004736328125, - "step": 500 - }, - { - "epoch": 0.06804234327362181, - "grad_norm": 50.733734130859375, - "learning_rate": 4.990903999473118e-05, - "loss": 250.31749378551137, - "step": 520 - }, - { - "epoch": 0.07065935647645342, - "grad_norm": 48.06721496582031, - "learning_rate": 4.9901899191179905e-05, - "loss": 243.69541015625, - "step": 540 - }, - { - "epoch": 0.07327636967928504, - "grad_norm": 54.863311767578125, - "learning_rate": 4.989448913413268e-05, - "loss": 225.0319580078125, - "step": 560 - }, - { - "epoch": 0.07589338288211664, - "grad_norm": 60.36458969116211, - "learning_rate": 4.988680990371125e-05, - "loss": 222.7537353515625, - "step": 580 - }, - { - "epoch": 0.07851039608494825, - "grad_norm": 57.877986907958984, - "learning_rate": 4.987886158294783e-05, - "loss": 241.103515625, - "step": 600 - }, - { - "epoch": 0.08112740928777985, - "grad_norm": 64.6783676147461, - "learning_rate": 4.987064425778417e-05, - "loss": 215.971142578125, - "step": 620 - }, - { - "epoch": 0.08374442249061147, - "grad_norm": 50.828765869140625, - "learning_rate": 4.986215801707066e-05, - "loss": 213.7265380859375, - "step": 640 - }, - { - "epoch": 0.08636143569344307, - "grad_norm": 55.68048858642578, - "learning_rate": 4.9853402952565383e-05, - "loss": 216.94306640625, - "step": 660 - }, - { - "epoch": 0.08897844889627468, - "grad_norm": 54.125267028808594, - "learning_rate": 4.984437915893305e-05, - "loss": 204.139990234375, - "step": 680 - }, - { - "epoch": 0.0915954620991063, - "grad_norm": 57.27487564086914, - "learning_rate": 4.9835086733744074e-05, - "loss": 202.9946533203125, - "step": 700 - }, - { - "epoch": 0.0942124753019379, - "grad_norm": 65.67598724365234, - "learning_rate": 4.9825525777473424e-05, - "loss": 207.21923828125, - "step": 720 - }, - { - "epoch": 0.09682948850476951, - "grad_norm": 71.1091537475586, - "learning_rate": 4.98156963934996e-05, - "loss": 204.4235107421875, - "step": 740 - }, - { - "epoch": 0.09944650170760111, - "grad_norm": 60.16761779785156, - "learning_rate": 4.9805598688103496e-05, - "loss": 213.841748046875, - "step": 760 - }, - { - "epoch": 0.19596246146437973, - "grad_norm": 268.7782897949219, - "learning_rate": 4.99999986487106e-05, - "loss": 474.80499267578125, - "step": 780 - }, - { - "epoch": 0.2009871399634664, - "grad_norm": 107.96406555175781, - "learning_rate": 4.9999459486180746e-05, - "loss": 528.41904296875, - "step": 800 - }, - { - "epoch": 0.20601181846255306, - "grad_norm": 133.09701538085938, - "learning_rate": 4.9997837968095394e-05, - "loss": 484.614111328125, - "step": 820 - }, - { - "epoch": 0.21103649696163973, - "grad_norm": 102.78153228759766, - "learning_rate": 4.999513551586019e-05, - "loss": 467.132373046875, - "step": 840 - }, - { - "epoch": 0.21606117546072637, - "grad_norm": 92.18177795410156, - "learning_rate": 4.999135224633215e-05, - "loss": 463.333203125, - "step": 860 - }, - { - "epoch": 0.22108585395981303, - "grad_norm": 106.00531005859375, - "learning_rate": 4.998648832310402e-05, - "loss": 452.96220703125, - "step": 880 - }, - { - "epoch": 0.2261105324588997, - "grad_norm": 100.12345123291016, - "learning_rate": 4.9980543956497236e-05, - "loss": 425.08818359375, - "step": 900 - }, - { - "epoch": 0.23113521095798636, - "grad_norm": 96.5484619140625, - "learning_rate": 4.997351940355277e-05, - "loss": 474.501953125, - "step": 920 - }, - { - "epoch": 0.23615988945707303, - "grad_norm": 104.5451889038086, - "learning_rate": 4.996541496802006e-05, - "loss": 447.30810546875, - "step": 940 - }, - { - "epoch": 0.24118456795615967, - "grad_norm": 129.78697204589844, - "learning_rate": 4.995623100034386e-05, - "loss": 442.8482421875, - "step": 960 - }, - { - "epoch": 0.24620924645524633, - "grad_norm": 95.0133285522461, - "learning_rate": 4.994596789764909e-05, - "loss": 433.66689453125, - "step": 980 - }, - { - "epoch": 0.251233924954333, - "grad_norm": 93.045654296875, - "learning_rate": 4.993462610372364e-05, - "loss": 418.920263671875, - "step": 1000 - }, - { - "epoch": 0.25625860345341966, - "grad_norm": 102.6265869140625, - "learning_rate": 4.992220610899924e-05, - "loss": 394.33173828125, - "step": 1020 - }, - { - "epoch": 0.26128328195250633, - "grad_norm": 94.07565307617188, - "learning_rate": 4.990870845053017e-05, - "loss": 388.1531494140625, - "step": 1040 - }, - { - "epoch": 0.266307960451593, - "grad_norm": 105.10103607177734, - "learning_rate": 4.989413371197013e-05, - "loss": 379.9688720703125, - "step": 1060 - }, - { - "epoch": 0.27133263895067966, - "grad_norm": 99.91519927978516, - "learning_rate": 4.987848252354691e-05, - "loss": 388.9978271484375, - "step": 1080 - }, - { - "epoch": 0.2763573174497663, - "grad_norm": 93.43067169189453, - "learning_rate": 4.986175556203521e-05, - "loss": 359.740380859375, - "step": 1100 - }, - { - "epoch": 0.28138199594885294, - "grad_norm": 86.09584045410156, - "learning_rate": 4.984395355072732e-05, - "loss": 394.5279296875, - "step": 1120 - }, - { - "epoch": 0.2864066744479396, - "grad_norm": 92.78561401367188, - "learning_rate": 4.982507725940191e-05, - "loss": 379.4552490234375, - "step": 1140 - }, - { - "epoch": 0.29143135294702627, - "grad_norm": 89.77680969238281, - "learning_rate": 4.980512750429068e-05, - "loss": 382.2197509765625, - "step": 1160 - }, - { - "epoch": 0.29645603144611293, - "grad_norm": 90.80536651611328, - "learning_rate": 4.97841051480431e-05, - "loss": 322.7532958984375, - "step": 1180 - }, - { - "epoch": 0.3014807099451996, - "grad_norm": 95.0079574584961, - "learning_rate": 4.976201109968908e-05, - "loss": 383.434130859375, - "step": 1200 - }, - { - "epoch": 0.30650538844428626, - "grad_norm": 96.08088684082031, - "learning_rate": 4.973884631459971e-05, - "loss": 315.9826904296875, - "step": 1220 - }, - { - "epoch": 0.31153006694337293, - "grad_norm": 95.79961395263672, - "learning_rate": 4.9714611794445906e-05, - "loss": 300.992431640625, - "step": 1240 - }, - { - "epoch": 0.3165547454424596, - "grad_norm": 85.15829467773438, - "learning_rate": 4.968930858715511e-05, - "loss": 345.0525390625, - "step": 1260 - }, - { - "epoch": 0.32157942394154626, - "grad_norm": 97.35791778564453, - "learning_rate": 4.966293778686598e-05, - "loss": 303.704541015625, - "step": 1280 - }, - { - "epoch": 0.3266041024406329, - "grad_norm": 100.1771469116211, - "learning_rate": 4.963550053388106e-05, - "loss": 326.5158447265625, - "step": 1300 - }, - { - "epoch": 0.33162878093971954, - "grad_norm": 90.39754486083984, - "learning_rate": 4.9606998014617535e-05, - "loss": 307.4456787109375, - "step": 1320 - }, - { - "epoch": 0.3366534594388062, - "grad_norm": 95.20532989501953, - "learning_rate": 4.957743146155581e-05, - "loss": 268.6821533203125, - "step": 1340 - }, - { - "epoch": 0.34167813793789287, - "grad_norm": 79.56786346435547, - "learning_rate": 4.954680215318635e-05, - "loss": 271.5860107421875, - "step": 1360 - }, - { - "epoch": 0.34670281643697953, - "grad_norm": 70.59774017333984, - "learning_rate": 4.951511141395432e-05, - "loss": 274.2177734375, - "step": 1380 - }, - { - "epoch": 0.3517274949360662, - "grad_norm": 85.6360092163086, - "learning_rate": 4.948236061420229e-05, - "loss": 282.2251708984375, - "step": 1400 - }, - { - "epoch": 0.35675217343515286, - "grad_norm": 80.13080596923828, - "learning_rate": 4.944855117011107e-05, - "loss": 288.8053466796875, - "step": 1420 - }, - { - "epoch": 0.36177685193423953, - "grad_norm": 72.03929901123047, - "learning_rate": 4.941368454363839e-05, - "loss": 260.295849609375, - "step": 1440 - }, - { - "epoch": 0.3668015304333262, - "grad_norm": 71.65677642822266, - "learning_rate": 4.937776224245573e-05, - "loss": 283.718701171875, - "step": 1460 - }, - { - "epoch": 0.37182620893241286, - "grad_norm": 79.88951110839844, - "learning_rate": 4.934078581988311e-05, - "loss": 237.0281494140625, - "step": 1480 - }, - { - "epoch": 0.3768508874314995, - "grad_norm": 70.0428237915039, - "learning_rate": 4.930275687482191e-05, - "loss": 236.0194091796875, - "step": 1500 - }, - { - "epoch": 0.38187556593058614, - "grad_norm": 92.4908447265625, - "learning_rate": 4.926367705168577e-05, - "loss": 240.9611572265625, - "step": 1520 - }, - { - "epoch": 0.3869002444296728, - "grad_norm": 70.08905792236328, - "learning_rate": 4.922354804032944e-05, - "loss": 230.3265380859375, - "step": 1540 - }, - { - "epoch": 0.39192492292875947, - "grad_norm": 77.18536376953125, - "learning_rate": 4.9182371575975736e-05, - "loss": 242.0444580078125, - "step": 1560 - }, - { - "epoch": 0.39694960142784613, - "grad_norm": 79.4393081665039, - "learning_rate": 4.9140149439140506e-05, - "loss": 247.16416015625, - "step": 1580 - }, - { - "epoch": 0.4019742799269328, - "grad_norm": 70.19295501708984, - "learning_rate": 4.909688345555561e-05, - "loss": 221.3811279296875, - "step": 1600 - }, - { - "epoch": 0.40699895842601946, - "grad_norm": 66.46074676513672, - "learning_rate": 4.9052575496090016e-05, - "loss": 222.3531005859375, - "step": 1620 - }, - { - "epoch": 0.41202363692510613, - "grad_norm": 103.58501434326172, - "learning_rate": 4.900722747666888e-05, - "loss": 237.266455078125, - "step": 1640 - }, - { - "epoch": 0.4170483154241928, - "grad_norm": 84.74609375, - "learning_rate": 4.8960841358190704e-05, - "loss": 220.35146484375, - "step": 1660 - }, - { - "epoch": 0.42207299392327946, - "grad_norm": 75.79342651367188, - "learning_rate": 4.891341914644251e-05, - "loss": 207.2835693359375, - "step": 1680 - }, - { - "epoch": 0.4270976724223661, - "grad_norm": 84.80883026123047, - "learning_rate": 4.886496289201318e-05, - "loss": 199.74576416015626, - "step": 1700 - }, - { - "epoch": 0.43212235092145274, - "grad_norm": 67.72750091552734, - "learning_rate": 4.881547469020471e-05, - "loss": 198.93392333984374, - "step": 1720 - }, - { - "epoch": 0.4371470294205394, - "grad_norm": 76.4952163696289, - "learning_rate": 4.8764956680941676e-05, - "loss": 189.1037109375, - "step": 1740 - }, - { - "epoch": 0.44217170791962607, - "grad_norm": 78.63851928710938, - "learning_rate": 4.8713411048678635e-05, - "loss": 226.8509521484375, - "step": 1760 - }, - { - "epoch": 0.44719638641871273, - "grad_norm": 63.36662292480469, - "learning_rate": 4.866084002230573e-05, - "loss": 202.01885986328125, - "step": 1780 - }, - { - "epoch": 0.4522210649177994, - "grad_norm": 82.46707916259766, - "learning_rate": 4.860724587505223e-05, - "loss": 190.03756103515624, - "step": 1800 - }, - { - "epoch": 0.45724574341688606, - "grad_norm": 79.09195709228516, - "learning_rate": 4.8552630924388335e-05, - "loss": 193.71884765625, - "step": 1820 - }, - { - "epoch": 0.46227042191597273, - "grad_norm": 72.94918060302734, - "learning_rate": 4.8496997531924885e-05, - "loss": 174.4781982421875, - "step": 1840 - }, - { - "epoch": 0.4672951004150594, - "grad_norm": 97.54232025146484, - "learning_rate": 4.8440348103311276e-05, - "loss": 196.74876708984374, - "step": 1860 - }, - { - "epoch": 0.47231977891414606, - "grad_norm": 70.02765655517578, - "learning_rate": 4.838268508813143e-05, - "loss": 183.72100830078125, - "step": 1880 - }, - { - "epoch": 0.4773444574132327, - "grad_norm": 62.996543884277344, - "learning_rate": 4.8324010979797875e-05, - "loss": 171.5578369140625, - "step": 1900 - }, - { - "epoch": 0.48236913591231934, - "grad_norm": 72.1956787109375, - "learning_rate": 4.8264328315443914e-05, - "loss": 167.999951171875, - "step": 1920 - }, - { - "epoch": 0.487393814411406, - "grad_norm": 96.08892822265625, - "learning_rate": 4.820363967581394e-05, - "loss": 159.68211669921874, - "step": 1940 - }, - { - "epoch": 0.49241849291049267, - "grad_norm": 65.87628173828125, - "learning_rate": 4.814194768515182e-05, - "loss": 156.25810546875, - "step": 1960 - }, - { - "epoch": 0.49744317140957933, - "grad_norm": 70.53435516357422, - "learning_rate": 4.807925501108744e-05, - "loss": 158.4431640625, - "step": 1980 - }, - { - "epoch": 0.502467849908666, - "grad_norm": 71.5821304321289, - "learning_rate": 4.801556436452133e-05, - "loss": 157.15218505859374, - "step": 2000 - }, - { - "epoch": 0.5074925284077526, - "grad_norm": 62.593109130859375, - "learning_rate": 4.795087849950747e-05, - "loss": 154.85384521484374, - "step": 2020 - }, - { - "epoch": 0.5125172069068393, - "grad_norm": 70.64319610595703, - "learning_rate": 4.7885200213134164e-05, - "loss": 144.212548828125, - "step": 2040 - }, - { - "epoch": 0.5175418854059259, - "grad_norm": 57.551979064941406, - "learning_rate": 4.781853234540312e-05, - "loss": 150.35511474609376, - "step": 2060 - }, - { - "epoch": 0.5225665639050127, - "grad_norm": 60.21275329589844, - "learning_rate": 4.7750877779106666e-05, - "loss": 139.1343994140625, - "step": 2080 - }, - { - "epoch": 0.5275912424040993, - "grad_norm": 74.57681274414062, - "learning_rate": 4.7682239439703025e-05, - "loss": 152.9317138671875, - "step": 2100 - }, - { - "epoch": 0.532615920903186, - "grad_norm": 62.081398010253906, - "learning_rate": 4.761262029518988e-05, - "loss": 138.762646484375, - "step": 2120 - }, - { - "epoch": 0.5376405994022726, - "grad_norm": 97.3165283203125, - "learning_rate": 4.754202335597601e-05, - "loss": 132.7292236328125, - "step": 2140 - }, - { - "epoch": 0.5426652779013593, - "grad_norm": 68.6803970336914, - "learning_rate": 4.7470451674751106e-05, - "loss": 129.36715087890624, - "step": 2160 - }, - { - "epoch": 0.5476899564004459, - "grad_norm": 63.835575103759766, - "learning_rate": 4.7397908346353796e-05, - "loss": 127.2830078125, - "step": 2180 - }, - { - "epoch": 0.5527146348995327, - "grad_norm": 63.419559478759766, - "learning_rate": 4.73243965076378e-05, - "loss": 129.15245361328124, - "step": 2200 - }, - { - "epoch": 0.5577393133986193, - "grad_norm": 53.640995025634766, - "learning_rate": 4.724991933733629e-05, - "loss": 122.0046630859375, - "step": 2220 - }, - { - "epoch": 0.5627639918977059, - "grad_norm": 59.515647888183594, - "learning_rate": 4.7174480055924455e-05, - "loss": 135.0306396484375, - "step": 2240 - }, - { - "epoch": 0.5677886703967926, - "grad_norm": 61.889373779296875, - "learning_rate": 4.7098081925480216e-05, - "loss": 123.96072998046876, - "step": 2260 - }, - { - "epoch": 0.5728133488958792, - "grad_norm": 52.50545120239258, - "learning_rate": 4.7020728249543196e-05, - "loss": 120.4759521484375, - "step": 2280 - }, - { - "epoch": 0.5778380273949659, - "grad_norm": 64.7458267211914, - "learning_rate": 4.694242237297186e-05, - "loss": 131.4004150390625, - "step": 2300 - }, - { - "epoch": 0.5828627058940525, - "grad_norm": 52.680599212646484, - "learning_rate": 4.686316768179889e-05, - "loss": 124.689404296875, - "step": 2320 - }, - { - "epoch": 0.5878873843931393, - "grad_norm": 77.16497039794922, - "learning_rate": 4.678296760308474e-05, - "loss": 120.44024658203125, - "step": 2340 - }, - { - "epoch": 0.5929120628922259, - "grad_norm": 60.93242263793945, - "learning_rate": 4.670182560476948e-05, - "loss": 120.22978515625, - "step": 2360 - }, - { - "epoch": 0.5979367413913126, - "grad_norm": 59.64492416381836, - "learning_rate": 4.661974519552283e-05, - "loss": 108.0113525390625, - "step": 2380 - }, - { - "epoch": 0.6029614198903992, - "grad_norm": 65.3152847290039, - "learning_rate": 4.653672992459242e-05, - "loss": 112.077978515625, - "step": 2400 - }, - { - "epoch": 0.6079860983894858, - "grad_norm": 58.9239616394043, - "learning_rate": 4.645278338165035e-05, - "loss": 106.50635986328125, - "step": 2420 - }, - { - "epoch": 0.6130107768885725, - "grad_norm": 56.77158737182617, - "learning_rate": 4.636790919663794e-05, - "loss": 102.1367431640625, - "step": 2440 - }, - { - "epoch": 0.6180354553876591, - "grad_norm": 49.77719497680664, - "learning_rate": 4.6282111039608784e-05, - "loss": 97.8525634765625, - "step": 2460 - }, - { - "epoch": 0.6230601338867459, - "grad_norm": 51.9918212890625, - "learning_rate": 4.6195392620570035e-05, - "loss": 105.98614501953125, - "step": 2480 - }, - { - "epoch": 0.6280848123858325, - "grad_norm": 58.79019546508789, - "learning_rate": 4.6107757689322014e-05, - "loss": 102.065966796875, - "step": 2500 - }, - { - "epoch": 0.6331094908849192, - "grad_norm": 49.530391693115234, - "learning_rate": 4.6019210035296034e-05, - "loss": 98.87720947265625, - "step": 2520 - }, - { - "epoch": 0.6381341693840058, - "grad_norm": 53.472782135009766, - "learning_rate": 4.592975348739054e-05, - "loss": 108.6241943359375, - "step": 2540 - }, - { - "epoch": 0.6431588478830925, - "grad_norm": 63.10646438598633, - "learning_rate": 4.584393142668341e-05, - "loss": 93.73190307617188, - "step": 2560 - }, - { - "epoch": 0.6481835263821791, - "grad_norm": 59.84651565551758, - "learning_rate": 4.5752713697269135e-05, - "loss": 96.8425048828125, - "step": 2580 - }, - { - "epoch": 0.6532082048812659, - "grad_norm": 60.635169982910156, - "learning_rate": 4.5660598597571637e-05, - "loss": 103.2667236328125, - "step": 2600 - }, - { - "epoch": 0.6582328833803525, - "grad_norm": 61.401737213134766, - "learning_rate": 4.5567590110749646e-05, - "loss": 87.72919311523438, - "step": 2620 - }, - { - "epoch": 0.6632575618794391, - "grad_norm": 60.122772216796875, - "learning_rate": 4.547369225859297e-05, - "loss": 94.103662109375, - "step": 2640 - }, - { - "epoch": 0.6682822403785258, - "grad_norm": 55.90855026245117, - "learning_rate": 4.537890910134854e-05, - "loss": 83.04338989257812, - "step": 2660 - }, - { - "epoch": 0.6733069188776124, - "grad_norm": 71.83133697509766, - "learning_rate": 4.528324473754486e-05, - "loss": 83.29258422851562, - "step": 2680 - }, - { - "epoch": 0.6783315973766991, - "grad_norm": 58.1335334777832, - "learning_rate": 4.518670330381479e-05, - "loss": 81.46260986328124, - "step": 2700 - }, - { - "epoch": 0.6833562758757857, - "grad_norm": 73.3138656616211, - "learning_rate": 4.5089288974716634e-05, - "loss": 81.26986694335938, - "step": 2720 - }, - { - "epoch": 0.6883809543748725, - "grad_norm": 64.87886810302734, - "learning_rate": 4.49910059625537e-05, - "loss": 82.78674926757813, - "step": 2740 - }, - { - "epoch": 0.6934056328739591, - "grad_norm": 57.293861389160156, - "learning_rate": 4.489185851719208e-05, - "loss": 82.71456909179688, - "step": 2760 - }, - { - "epoch": 0.6984303113730458, - "grad_norm": 53.40017318725586, - "learning_rate": 4.4791850925876926e-05, - "loss": 84.79779052734375, - "step": 2780 - }, - { - "epoch": 0.7034549898721324, - "grad_norm": 48.285736083984375, - "learning_rate": 4.4690987513047044e-05, - "loss": 81.11317138671875, - "step": 2800 - }, - { - "epoch": 0.708479668371219, - "grad_norm": 56.2747802734375, - "learning_rate": 4.4589272640147915e-05, - "loss": 72.64208374023437, - "step": 2820 - }, - { - "epoch": 0.7135043468703057, - "grad_norm": 54.770931243896484, - "learning_rate": 4.44867107054431e-05, - "loss": 75.95457153320312, - "step": 2840 - }, - { - "epoch": 0.7185290253693923, - "grad_norm": 49.85646057128906, - "learning_rate": 4.4383306143824035e-05, - "loss": 74.01212158203126, - "step": 2860 - }, - { - "epoch": 0.7235537038684791, - "grad_norm": 49.286746978759766, - "learning_rate": 4.427906342661828e-05, - "loss": 67.97941284179687, - "step": 2880 - }, - { - "epoch": 0.7285783823675657, - "grad_norm": 56.60735321044922, - "learning_rate": 4.4173987061396186e-05, - "loss": 70.84263916015625, - "step": 2900 - }, - { - "epoch": 0.7336030608666524, - "grad_norm": 57.23542404174805, - "learning_rate": 4.406808159177593e-05, - "loss": 65.6290283203125, - "step": 2920 - }, - { - "epoch": 0.738627739365739, - "grad_norm": 83.55549621582031, - "learning_rate": 4.396135159722711e-05, - "loss": 65.52483520507812, - "step": 2940 - }, - { - "epoch": 0.7436524178648257, - "grad_norm": 53.960811614990234, - "learning_rate": 4.3853801692872686e-05, - "loss": 67.09555053710938, - "step": 2960 - }, - { - "epoch": 0.7486770963639123, - "grad_norm": 45.474632263183594, - "learning_rate": 4.374543652928943e-05, - "loss": 67.3612548828125, - "step": 2980 - }, - { - "epoch": 0.753701774862999, - "grad_norm": 56.11079025268555, - "learning_rate": 4.363626079230681e-05, - "loss": 62.279473876953126, - "step": 3000 - }, - { - "epoch": 0.7587264533620857, - "grad_norm": 49.5059928894043, - "learning_rate": 4.3526279202804387e-05, - "loss": 73.84429321289062, - "step": 3020 - }, - { - "epoch": 0.7637511318611723, - "grad_norm": 53.92646789550781, - "learning_rate": 4.3415496516507695e-05, - "loss": 61.59212646484375, - "step": 3040 - }, - { - "epoch": 0.768775810360259, - "grad_norm": 57.92325210571289, - "learning_rate": 4.3303917523782546e-05, - "loss": 65.36521606445312, - "step": 3060 - }, - { - "epoch": 0.7738004888593456, - "grad_norm": 47.250648498535156, - "learning_rate": 4.3191547049427954e-05, - "loss": 66.87857055664062, - "step": 3080 - }, - { - "epoch": 0.7788251673584323, - "grad_norm": 91.52741241455078, - "learning_rate": 4.307838995246745e-05, - "loss": 62.80021362304687, - "step": 3100 - }, - { - "epoch": 0.7838498458575189, - "grad_norm": 46.733856201171875, - "learning_rate": 4.296445112593902e-05, - "loss": 55.04381103515625, - "step": 3120 - }, - { - "epoch": 0.7888745243566057, - "grad_norm": 44.601497650146484, - "learning_rate": 4.284973549668347e-05, - "loss": 55.08817749023437, - "step": 3140 - }, - { - "epoch": 0.7938992028556923, - "grad_norm": 50.94633102416992, - "learning_rate": 4.273424802513145e-05, - "loss": 52.93193359375, - "step": 3160 - }, - { - "epoch": 0.798923881354779, - "grad_norm": 43.45872497558594, - "learning_rate": 4.2617993705088886e-05, - "loss": 54.57942504882813, - "step": 3180 - }, - { - "epoch": 0.8039485598538656, - "grad_norm": 46.4177131652832, - "learning_rate": 4.2500977563521115e-05, - "loss": 54.19765014648438, - "step": 3200 - }, - { - "epoch": 0.8089732383529522, - "grad_norm": 48.374786376953125, - "learning_rate": 4.238320466033546e-05, - "loss": 51.5730712890625, - "step": 3220 - }, - { - "epoch": 0.8139979168520389, - "grad_norm": 51.431087493896484, - "learning_rate": 4.2264680088162456e-05, - "loss": 52.94944458007812, - "step": 3240 - }, - { - "epoch": 0.8190225953511255, - "grad_norm": 40.70262145996094, - "learning_rate": 4.214540897213565e-05, - "loss": 53.7551513671875, - "step": 3260 - }, - { - "epoch": 0.8240472738502123, - "grad_norm": 43.26173782348633, - "learning_rate": 4.202539646966993e-05, - "loss": 55.24525756835938, - "step": 3280 - }, - { - "epoch": 0.8290719523492989, - "grad_norm": 47.32640838623047, - "learning_rate": 4.190464777023861e-05, - "loss": 48.86954040527344, - "step": 3300 - }, - { - "epoch": 0.8340966308483856, - "grad_norm": 47.30034255981445, - "learning_rate": 4.178316809514893e-05, - "loss": 46.55088806152344, - "step": 3320 - }, - { - "epoch": 0.8391213093474722, - "grad_norm": 51.20639419555664, - "learning_rate": 4.1660962697316346e-05, - "loss": 48.73444519042969, - "step": 3340 - }, - { - "epoch": 0.8441459878465589, - "grad_norm": 51.74330139160156, - "learning_rate": 4.153803686103737e-05, - "loss": 45.40309448242188, - "step": 3360 - }, - { - "epoch": 0.8491706663456455, - "grad_norm": 41.944496154785156, - "learning_rate": 4.141439590176105e-05, - "loss": 46.365499877929686, - "step": 3380 - }, - { - "epoch": 0.8541953448447323, - "grad_norm": 41.049713134765625, - "learning_rate": 4.129004516585916e-05, - "loss": 47.59626770019531, - "step": 3400 - }, - { - "epoch": 0.8592200233438189, - "grad_norm": 44.36544418334961, - "learning_rate": 4.116499003039499e-05, - "loss": 42.41243286132813, - "step": 3420 - }, - { - "epoch": 0.8642447018429055, - "grad_norm": 46.1484489440918, - "learning_rate": 4.1039235902890875e-05, - "loss": 44.016815185546875, - "step": 3440 - }, - { - "epoch": 0.8692693803419922, - "grad_norm": 38.971961975097656, - "learning_rate": 4.091278822109428e-05, - "loss": 41.796511840820315, - "step": 3460 - }, - { - "epoch": 0.8742940588410788, - "grad_norm": 54.43899917602539, - "learning_rate": 4.0785652452742776e-05, - "loss": 43.20156555175781, - "step": 3480 - }, - { - "epoch": 0.8793187373401655, - "grad_norm": 36.39228820800781, - "learning_rate": 4.065783409532755e-05, - "loss": 41.3662353515625, - "step": 3500 - }, - { - "epoch": 0.8843434158392521, - "grad_norm": 43.41612243652344, - "learning_rate": 4.052933867585567e-05, - "loss": 41.19517211914062, - "step": 3520 - }, - { - "epoch": 0.8893680943383389, - "grad_norm": 50.1053352355957, - "learning_rate": 4.040017175061113e-05, - "loss": 45.48277893066406, - "step": 3540 - }, - { - "epoch": 0.8943927728374255, - "grad_norm": 68.26251983642578, - "learning_rate": 4.027033890491458e-05, - "loss": 40.87652587890625, - "step": 3560 - }, - { - "epoch": 0.8994174513365122, - "grad_norm": 46.13456344604492, - "learning_rate": 4.0139845752881814e-05, - "loss": 39.160336303710935, - "step": 3580 - }, - { - "epoch": 0.9044421298355988, - "grad_norm": 44.120697021484375, - "learning_rate": 4.0008697937180984e-05, - "loss": 41.72444458007813, - "step": 3600 - }, - { - "epoch": 0.9094668083346854, - "grad_norm": 36.38677215576172, - "learning_rate": 3.987690112878862e-05, - "loss": 37.204827880859376, - "step": 3620 - }, - { - "epoch": 0.9144914868337721, - "grad_norm": 49.02151870727539, - "learning_rate": 3.974446102674444e-05, - "loss": 37.10683898925781, - "step": 3640 - }, - { - "epoch": 0.9195161653328587, - "grad_norm": 36.17109298706055, - "learning_rate": 3.9611383357904866e-05, - "loss": 38.15001525878906, - "step": 3660 - }, - { - "epoch": 0.9245408438319455, - "grad_norm": 40.26628112792969, - "learning_rate": 3.947767387669542e-05, - "loss": 44.486627197265626, - "step": 3680 - }, - { - "epoch": 0.9295655223310321, - "grad_norm": 38.918766021728516, - "learning_rate": 3.934333836486189e-05, - "loss": 35.51749267578125, - "step": 3700 - }, - { - "epoch": 0.9345902008301188, - "grad_norm": 36.1772346496582, - "learning_rate": 3.920838263122033e-05, - "loss": 32.570498657226565, - "step": 3720 - }, - { - "epoch": 0.9396148793292054, - "grad_norm": 39.06047058105469, - "learning_rate": 3.9072812511405855e-05, - "loss": 34.915673828125, - "step": 3740 - }, - { - "epoch": 0.9446395578282921, - "grad_norm": 41.863502502441406, - "learning_rate": 3.893663386762032e-05, - "loss": 34.646026611328125, - "step": 3760 - }, - { - "epoch": 0.9496642363273787, - "grad_norm": 32.75388717651367, - "learning_rate": 3.8799852588378836e-05, - "loss": 33.01146545410156, - "step": 3780 - }, - { - "epoch": 0.9546889148264655, - "grad_norm": 45.868865966796875, - "learning_rate": 3.8662474588255146e-05, - "loss": 30.23636474609375, - "step": 3800 - }, - { - "epoch": 0.9597135933255521, - "grad_norm": 37.765052795410156, - "learning_rate": 3.852450580762584e-05, - "loss": 31.944390869140626, - "step": 3820 - }, - { - "epoch": 0.9647382718246387, - "grad_norm": 33.98917007446289, - "learning_rate": 3.8385952212413525e-05, - "loss": 31.06385498046875, - "step": 3840 - }, - { - "epoch": 0.9697629503237254, - "grad_norm": 32.69452667236328, - "learning_rate": 3.824681979382883e-05, - "loss": 30.422335815429687, - "step": 3860 - }, - { - "epoch": 0.974787628822812, - "grad_norm": 47.07701110839844, - "learning_rate": 3.810711456811136e-05, - "loss": 30.911959838867187, - "step": 3880 - }, - { - "epoch": 0.9798123073218987, - "grad_norm": 37.96084976196289, - "learning_rate": 3.796684257626952e-05, - "loss": 29.244085693359374, - "step": 3900 - }, - { - "epoch": 0.9848369858209853, - "grad_norm": 32.45835494995117, - "learning_rate": 3.7826009883819315e-05, - "loss": 27.361874389648438, - "step": 3920 - }, - { - "epoch": 0.989861664320072, - "grad_norm": 35.04416275024414, - "learning_rate": 3.7684622580522055e-05, - "loss": 27.51587829589844, - "step": 3940 - }, - { - "epoch": 0.9948863428191587, - "grad_norm": 45.533992767333984, - "learning_rate": 3.754268678012106e-05, - "loss": 27.282977294921874, - "step": 3960 - }, - { - "epoch": 0.9999110213182454, - "grad_norm": 37.77592086791992, - "learning_rate": 3.740020862007724e-05, - "loss": 27.77158203125, - "step": 3980 - }, - { - "epoch": 1.0047734445741323, - "grad_norm": 29.9664306640625, - "learning_rate": 3.7257194261303765e-05, - "loss": 24.126179504394532, - "step": 4000 - }, - { - "epoch": 1.009798123073219, - "grad_norm": 36.00257873535156, - "learning_rate": 3.71136498878996e-05, - "loss": 25.783734130859376, - "step": 4020 - }, - { - "epoch": 1.0148228015723058, - "grad_norm": 71.59026336669922, - "learning_rate": 3.696958170688215e-05, - "loss": 24.485565185546875, - "step": 4040 - }, - { - "epoch": 1.0198474800713924, - "grad_norm": 31.764368057250977, - "learning_rate": 3.682499594791885e-05, - "loss": 23.261607360839843, - "step": 4060 - }, - { - "epoch": 1.024872158570479, - "grad_norm": 31.287395477294922, - "learning_rate": 3.667989886305774e-05, - "loss": 23.564991760253907, - "step": 4080 - }, - { - "epoch": 1.0298968370695656, - "grad_norm": 29.753856658935547, - "learning_rate": 3.653429672645721e-05, - "loss": 25.142503356933595, - "step": 4100 - }, - { - "epoch": 1.0349215155686522, - "grad_norm": 29.53838539123535, - "learning_rate": 3.638819583411458e-05, - "loss": 23.48755340576172, - "step": 4120 - }, - { - "epoch": 1.039946194067739, - "grad_norm": 29.025772094726562, - "learning_rate": 3.624160250359398e-05, - "loss": 21.697846984863283, - "step": 4140 - }, - { - "epoch": 1.0449708725668256, - "grad_norm": 31.875118255615234, - "learning_rate": 3.609452307375308e-05, - "loss": 24.43586883544922, - "step": 4160 - }, - { - "epoch": 1.0499955510659122, - "grad_norm": 34.18278121948242, - "learning_rate": 3.5946963904469024e-05, - "loss": 21.297456359863283, - "step": 4180 - }, - { - "epoch": 1.0550202295649989, - "grad_norm": 31.082683563232422, - "learning_rate": 3.579893137636343e-05, - "loss": 22.34284362792969, - "step": 4200 - }, - { - "epoch": 1.0600449080640857, - "grad_norm": 31.194740295410156, - "learning_rate": 3.565043189052647e-05, - "loss": 23.093136596679688, - "step": 4220 - }, - { - "epoch": 1.0650695865631723, - "grad_norm": 32.549827575683594, - "learning_rate": 3.550147186824008e-05, - "loss": 21.879423522949217, - "step": 4240 - }, - { - "epoch": 1.070094265062259, - "grad_norm": 32.4059944152832, - "learning_rate": 3.53520577507003e-05, - "loss": 20.31920623779297, - "step": 4260 - }, - { - "epoch": 1.0751189435613455, - "grad_norm": 29.14031219482422, - "learning_rate": 3.520219599873877e-05, - "loss": 21.056622314453126, - "step": 4280 - }, - { - "epoch": 1.0801436220604321, - "grad_norm": 23.570371627807617, - "learning_rate": 3.5051893092543306e-05, - "loss": 19.163606262207033, - "step": 4300 - }, - { - "epoch": 1.085168300559519, - "grad_norm": 30.20454216003418, - "learning_rate": 3.490115553137776e-05, - "loss": 21.274189758300782, - "step": 4320 - }, - { - "epoch": 1.0901929790586056, - "grad_norm": 27.927400588989258, - "learning_rate": 3.47499898333009e-05, - "loss": 18.568623352050782, - "step": 4340 - }, - { - "epoch": 1.0952176575576922, - "grad_norm": 31.376602172851562, - "learning_rate": 3.459840253488465e-05, - "loss": 19.727203369140625, - "step": 4360 - }, - { - "epoch": 1.1002423360567788, - "grad_norm": 36.069026947021484, - "learning_rate": 3.444640019093136e-05, - "loss": 18.880654907226564, - "step": 4380 - }, - { - "epoch": 1.1052670145558656, - "grad_norm": 33.837711334228516, - "learning_rate": 3.429398937419044e-05, - "loss": 18.870826721191406, - "step": 4400 - }, - { - "epoch": 1.1102916930549522, - "grad_norm": 29.549623489379883, - "learning_rate": 3.4141176675074104e-05, - "loss": 18.073291015625, - "step": 4420 - }, - { - "epoch": 1.1153163715540388, - "grad_norm": 24.568265914916992, - "learning_rate": 3.39879687013724e-05, - "loss": 19.973286437988282, - "step": 4440 - }, - { - "epoch": 1.1203410500531255, - "grad_norm": 26.699716567993164, - "learning_rate": 3.383437207796748e-05, - "loss": 17.782139587402344, - "step": 4460 - }, - { - "epoch": 1.1253657285522123, - "grad_norm": 28.5200252532959, - "learning_rate": 3.3680393446547155e-05, - "loss": 17.14067687988281, - "step": 4480 - }, - { - "epoch": 1.130390407051299, - "grad_norm": 38.18730163574219, - "learning_rate": 3.3526039465317675e-05, - "loss": 18.262770080566405, - "step": 4500 - }, - { - "epoch": 1.1354150855503855, - "grad_norm": 22.557958602905273, - "learning_rate": 3.337131680871584e-05, - "loss": 17.867935180664062, - "step": 4520 - }, - { - "epoch": 1.1404397640494721, - "grad_norm": 22.901031494140625, - "learning_rate": 3.3216232167120356e-05, - "loss": 17.154512023925783, - "step": 4540 - }, - { - "epoch": 1.1454644425485587, - "grad_norm": 27.419776916503906, - "learning_rate": 3.30607922465626e-05, - "loss": 16.06898193359375, - "step": 4560 - }, - { - "epoch": 1.1504891210476456, - "grad_norm": 25.54627799987793, - "learning_rate": 3.2905003768436576e-05, - "loss": 15.820521545410156, - "step": 4580 - }, - { - "epoch": 1.1555137995467322, - "grad_norm": 23.961387634277344, - "learning_rate": 3.274887346920829e-05, - "loss": 14.808900451660156, - "step": 4600 - }, - { - "epoch": 1.1605384780458188, - "grad_norm": 23.095775604248047, - "learning_rate": 3.259240810012452e-05, - "loss": 14.714529418945313, - "step": 4620 - }, - { - "epoch": 1.1655631565449054, - "grad_norm": 22.404417037963867, - "learning_rate": 3.243561442692078e-05, - "loss": 14.217619323730469, - "step": 4640 - }, - { - "epoch": 1.170587835043992, - "grad_norm": 28.005577087402344, - "learning_rate": 3.2278499229528836e-05, - "loss": 15.66644287109375, - "step": 4660 - }, - { - "epoch": 1.1756125135430788, - "grad_norm": 22.640522003173828, - "learning_rate": 3.212106930178354e-05, - "loss": 14.664166259765626, - "step": 4680 - }, - { - "epoch": 1.1806371920421654, - "grad_norm": 24.132041931152344, - "learning_rate": 3.1963331451129e-05, - "loss": 16.09429473876953, - "step": 4700 - }, - { - "epoch": 1.185661870541252, - "grad_norm": 25.478002548217773, - "learning_rate": 3.180529249832428e-05, - "loss": 14.240249633789062, - "step": 4720 - }, - { - "epoch": 1.1906865490403387, - "grad_norm": 24.66731834411621, - "learning_rate": 3.1646959277148406e-05, - "loss": 14.484941101074218, - "step": 4740 - }, - { - "epoch": 1.1957112275394255, - "grad_norm": 23.402143478393555, - "learning_rate": 3.14883386341049e-05, - "loss": 13.931251525878906, - "step": 4760 - }, - { - "epoch": 1.200735906038512, - "grad_norm": 21.010387420654297, - "learning_rate": 3.132943742812574e-05, - "loss": 14.163758850097656, - "step": 4780 - }, - { - "epoch": 1.2057605845375987, - "grad_norm": 20.641809463500977, - "learning_rate": 3.1170262530274736e-05, - "loss": 12.913584899902343, - "step": 4800 - }, - { - "epoch": 1.2107852630366853, - "grad_norm": 23.429895401000977, - "learning_rate": 3.101082082345045e-05, - "loss": 12.281029510498048, - "step": 4820 - }, - { - "epoch": 1.2158099415357722, - "grad_norm": 28.63577651977539, - "learning_rate": 3.0851119202088555e-05, - "loss": 12.448915100097656, - "step": 4840 - }, - { - "epoch": 1.2208346200348588, - "grad_norm": 23.663249969482422, - "learning_rate": 3.069116457186371e-05, - "loss": 13.750531005859376, - "step": 4860 - }, - { - "epoch": 1.2258592985339454, - "grad_norm": 22.31790542602539, - "learning_rate": 3.053096384939097e-05, - "loss": 12.410870361328126, - "step": 4880 - }, - { - "epoch": 1.230883977033032, - "grad_norm": 21.297197341918945, - "learning_rate": 3.037052396192667e-05, - "loss": 12.693349456787109, - "step": 4900 - }, - { - "epoch": 1.2359086555321186, - "grad_norm": 23.489055633544922, - "learning_rate": 3.0209851847068932e-05, - "loss": 12.602223968505859, - "step": 4920 - }, - { - "epoch": 1.2409333340312054, - "grad_norm": 23.18564224243164, - "learning_rate": 3.004895445245762e-05, - "loss": 11.585443878173828, - "step": 4940 - }, - { - "epoch": 1.245958012530292, - "grad_norm": 22.05788230895996, - "learning_rate": 2.988783873547397e-05, - "loss": 11.15335922241211, - "step": 4960 - }, - { - "epoch": 1.2509826910293786, - "grad_norm": 21.006710052490234, - "learning_rate": 2.9726511662939695e-05, - "loss": 11.558931732177735, - "step": 4980 - }, - { - "epoch": 1.2560073695284653, - "grad_norm": 33.19025421142578, - "learning_rate": 2.956498021081575e-05, - "loss": 11.444783020019532, - "step": 5000 - }, - { - "epoch": 1.2610320480275519, - "grad_norm": 21.366567611694336, - "learning_rate": 2.9403251363900736e-05, - "loss": 10.560770416259766, - "step": 5020 - }, - { - "epoch": 1.2660567265266387, - "grad_norm": 22.778701782226562, - "learning_rate": 2.9241332115528763e-05, - "loss": 10.666181945800782, - "step": 5040 - }, - { - "epoch": 1.2710814050257253, - "grad_norm": 20.653833389282227, - "learning_rate": 2.907922946726716e-05, - "loss": 11.136424255371093, - "step": 5060 - }, - { - "epoch": 1.276106083524812, - "grad_norm": 21.236427307128906, - "learning_rate": 2.891695042861363e-05, - "loss": 11.04461212158203, - "step": 5080 - }, - { - "epoch": 1.2811307620238988, - "grad_norm": 39.907684326171875, - "learning_rate": 2.875450201669323e-05, - "loss": 10.980261993408202, - "step": 5100 - }, - { - "epoch": 1.2861554405229854, - "grad_norm": 19.791812896728516, - "learning_rate": 2.8591891255954873e-05, - "loss": 10.691975402832032, - "step": 5120 - }, - { - "epoch": 1.291180119022072, - "grad_norm": 16.12239646911621, - "learning_rate": 2.8429125177867622e-05, - "loss": 11.203057098388673, - "step": 5140 - }, - { - "epoch": 1.2962047975211586, - "grad_norm": 22.192853927612305, - "learning_rate": 2.8266210820616644e-05, - "loss": 9.194258880615234, - "step": 5160 - }, - { - "epoch": 1.3012294760202452, - "grad_norm": 16.881107330322266, - "learning_rate": 2.8103155228798855e-05, - "loss": 9.5519775390625, - "step": 5180 - }, - { - "epoch": 1.306254154519332, - "grad_norm": 36.23583221435547, - "learning_rate": 2.7939965453118318e-05, - "loss": 9.040764617919923, - "step": 5200 - }, - { - "epoch": 1.3112788330184186, - "grad_norm": 17.388961791992188, - "learning_rate": 2.777664855008133e-05, - "loss": 9.911444854736327, - "step": 5220 - }, - { - "epoch": 1.3163035115175052, - "grad_norm": 109.7496566772461, - "learning_rate": 2.761321158169134e-05, - "loss": 9.183967590332031, - "step": 5240 - }, - { - "epoch": 1.3213281900165919, - "grad_norm": 19.465408325195312, - "learning_rate": 2.7449661615143545e-05, - "loss": 9.167644500732422, - "step": 5260 - }, - { - "epoch": 1.3263528685156785, - "grad_norm": 23.866313934326172, - "learning_rate": 2.7286005722519314e-05, - "loss": 10.72784652709961, - "step": 5280 - }, - { - "epoch": 1.3313775470147653, - "grad_norm": 16.75422477722168, - "learning_rate": 2.712225098048037e-05, - "loss": 8.88067855834961, - "step": 5300 - }, - { - "epoch": 1.336402225513852, - "grad_norm": 21.11101531982422, - "learning_rate": 2.6958404469962794e-05, - "loss": 9.210741424560547, - "step": 5320 - }, - { - "epoch": 1.3414269040129385, - "grad_norm": 18.632394790649414, - "learning_rate": 2.6794473275870847e-05, - "loss": 8.116027069091796, - "step": 5340 - }, - { - "epoch": 1.3464515825120251, - "grad_norm": 16.196258544921875, - "learning_rate": 2.6630464486770584e-05, - "loss": 8.565786743164063, - "step": 5360 - }, - { - "epoch": 1.3514762610111117, - "grad_norm": 15.409602165222168, - "learning_rate": 2.6466385194583367e-05, - "loss": 7.9316253662109375, - "step": 5380 - }, - { - "epoch": 1.3565009395101986, - "grad_norm": 19.422887802124023, - "learning_rate": 2.6302242494279194e-05, - "loss": 7.932023620605468, - "step": 5400 - }, - { - "epoch": 1.3615256180092852, - "grad_norm": 17.550283432006836, - "learning_rate": 2.613804348356988e-05, - "loss": 8.40458984375, - "step": 5420 - }, - { - "epoch": 1.3665502965083718, - "grad_norm": 21.085948944091797, - "learning_rate": 2.597379526260219e-05, - "loss": 7.8362586975097654, - "step": 5440 - }, - { - "epoch": 1.3715749750074586, - "grad_norm": 16.09673309326172, - "learning_rate": 2.5809504933650775e-05, - "loss": 7.722866058349609, - "step": 5460 - }, - { - "epoch": 1.3765996535065452, - "grad_norm": 13.360176086425781, - "learning_rate": 2.564517960081108e-05, - "loss": 8.046268463134766, - "step": 5480 - }, - { - "epoch": 1.3816243320056318, - "grad_norm": 17.158672332763672, - "learning_rate": 2.5480826369692178e-05, - "loss": 8.267652893066407, - "step": 5500 - }, - { - "epoch": 1.3866490105047184, - "grad_norm": 16.403295516967773, - "learning_rate": 2.5316452347109486e-05, - "loss": 7.891246795654297, - "step": 5520 - }, - { - "epoch": 1.391673689003805, - "grad_norm": 18.296703338623047, - "learning_rate": 2.5152064640777444e-05, - "loss": 8.161449432373047, - "step": 5540 - }, - { - "epoch": 1.396698367502892, - "grad_norm": 13.22531509399414, - "learning_rate": 2.4987670359002225e-05, - "loss": 7.458624267578125, - "step": 5560 - }, - { - "epoch": 1.4017230460019785, - "grad_norm": 15.091835975646973, - "learning_rate": 2.482327661037431e-05, - "loss": 6.726652526855469, - "step": 5580 - }, - { - "epoch": 1.406747724501065, - "grad_norm": 24.141721725463867, - "learning_rate": 2.4658890503461142e-05, - "loss": 7.5736839294433596, - "step": 5600 - }, - { - "epoch": 1.4117724030001517, - "grad_norm": 15.211458206176758, - "learning_rate": 2.4494519146499718e-05, - "loss": 6.948004150390625, - "step": 5620 - }, - { - "epoch": 1.4167970814992383, - "grad_norm": 18.808673858642578, - "learning_rate": 2.4330169647089222e-05, - "loss": 7.693637084960938, - "step": 5640 - }, - { - "epoch": 1.4218217599983252, - "grad_norm": 12.74593734741211, - "learning_rate": 2.416584911188371e-05, - "loss": 6.476108551025391, - "step": 5660 - }, - { - "epoch": 1.4268464384974118, - "grad_norm": 14.523116111755371, - "learning_rate": 2.400156464628479e-05, - "loss": 6.905941772460937, - "step": 5680 - }, - { - "epoch": 1.4318711169964984, - "grad_norm": 13.235814094543457, - "learning_rate": 2.3837323354134374e-05, - "loss": 6.449698638916016, - "step": 5700 - }, - { - "epoch": 1.4368957954955852, - "grad_norm": 12.135284423828125, - "learning_rate": 2.3673132337407514e-05, - "loss": 6.168265151977539, - "step": 5720 - }, - { - "epoch": 1.4419204739946718, - "grad_norm": 14.88333511352539, - "learning_rate": 2.350899869590529e-05, - "loss": 6.634694671630859, - "step": 5740 - }, - { - "epoch": 1.4469451524937584, - "grad_norm": 14.54021167755127, - "learning_rate": 2.3344929526947816e-05, - "loss": 6.686222076416016, - "step": 5760 - }, - { - "epoch": 1.451969830992845, - "grad_norm": 14.360895156860352, - "learning_rate": 2.3180931925067337e-05, - "loss": 5.889413833618164, - "step": 5780 - }, - { - "epoch": 1.4569945094919317, - "grad_norm": 21.827463150024414, - "learning_rate": 2.3017012981701482e-05, - "loss": 6.283362960815429, - "step": 5800 - }, - { - "epoch": 1.4620191879910185, - "grad_norm": 14.690330505371094, - "learning_rate": 2.2853179784886566e-05, - "loss": 5.913167190551758, - "step": 5820 - }, - { - "epoch": 1.467043866490105, - "grad_norm": 14.252549171447754, - "learning_rate": 2.2689439418951152e-05, - "loss": 6.030026245117187, - "step": 5840 - }, - { - "epoch": 1.4720685449891917, - "grad_norm": 11.477004051208496, - "learning_rate": 2.252579896420968e-05, - "loss": 6.0997673034667965, - "step": 5860 - }, - { - "epoch": 1.4770932234882783, - "grad_norm": 13.692665100097656, - "learning_rate": 2.2362265496656314e-05, - "loss": 5.615577316284179, - "step": 5880 - }, - { - "epoch": 1.482117901987365, - "grad_norm": 13.419648170471191, - "learning_rate": 2.2198846087659e-05, - "loss": 6.581723022460937, - "step": 5900 - }, - { - "epoch": 1.4871425804864518, - "grad_norm": 13.096921920776367, - "learning_rate": 2.203554780365365e-05, - "loss": 6.183859252929688, - "step": 5920 - }, - { - "epoch": 1.4921672589855384, - "grad_norm": 15.762218475341797, - "learning_rate": 2.187237770583859e-05, - "loss": 5.984473419189453, - "step": 5940 - }, - { - "epoch": 1.497191937484625, - "grad_norm": 14.980542182922363, - "learning_rate": 2.170934284986924e-05, - "loss": 5.577951431274414, - "step": 5960 - }, - { - "epoch": 1.5022166159837118, - "grad_norm": 18.4476318359375, - "learning_rate": 2.1546450285553012e-05, - "loss": 5.640556716918946, - "step": 5980 - }, - { - "epoch": 1.5072412944827982, - "grad_norm": 11.552315711975098, - "learning_rate": 2.1383707056544476e-05, - "loss": 5.833734130859375, - "step": 6000 - }, - { - "epoch": 1.512265972981885, - "grad_norm": 11.801403045654297, - "learning_rate": 2.122112020004078e-05, - "loss": 5.025530624389648, - "step": 6020 - }, - { - "epoch": 1.5172906514809716, - "grad_norm": 13.236927032470703, - "learning_rate": 2.1058696746477337e-05, - "loss": 5.505057907104492, - "step": 6040 - }, - { - "epoch": 1.5223153299800583, - "grad_norm": 11.760169982910156, - "learning_rate": 2.0896443719223834e-05, - "loss": 5.180996322631836, - "step": 6060 - }, - { - "epoch": 1.527340008479145, - "grad_norm": 13.339624404907227, - "learning_rate": 2.0734368134280572e-05, - "loss": 5.530018615722656, - "step": 6080 - }, - { - "epoch": 1.5323646869782315, - "grad_norm": 15.223504066467285, - "learning_rate": 2.057247699997501e-05, - "loss": 5.101480865478516, - "step": 6100 - }, - { - "epoch": 1.5373893654773183, - "grad_norm": 11.12869644165039, - "learning_rate": 2.0410777316658776e-05, - "loss": 5.387827301025391, - "step": 6120 - }, - { - "epoch": 1.542414043976405, - "grad_norm": 11.168490409851074, - "learning_rate": 2.0249276076404937e-05, - "loss": 5.365410232543946, - "step": 6140 - }, - { - "epoch": 1.5474387224754915, - "grad_norm": 11.577016830444336, - "learning_rate": 2.0087980262705663e-05, - "loss": 5.477450561523438, - "step": 6160 - }, - { - "epoch": 1.5524634009745784, - "grad_norm": 14.104944229125977, - "learning_rate": 1.992689685017028e-05, - "loss": 4.753940200805664, - "step": 6180 - }, - { - "epoch": 1.557488079473665, - "grad_norm": 10.449115753173828, - "learning_rate": 1.9766032804223612e-05, - "loss": 4.47644157409668, - "step": 6200 - }, - { - "epoch": 1.5625127579727516, - "grad_norm": 14.98228931427002, - "learning_rate": 1.960539508080485e-05, - "loss": 4.596777725219726, - "step": 6220 - }, - { - "epoch": 1.5675374364718384, - "grad_norm": 16.68485450744629, - "learning_rate": 1.944499062606675e-05, - "loss": 4.914170455932617, - "step": 6240 - }, - { - "epoch": 1.5725621149709248, - "grad_norm": 14.826934814453125, - "learning_rate": 1.9284826376075256e-05, - "loss": 4.500260925292968, - "step": 6260 - }, - { - "epoch": 1.5775867934700116, - "grad_norm": 15.89423656463623, - "learning_rate": 1.9124909256509622e-05, - "loss": 4.9693656921386715, - "step": 6280 - }, - { - "epoch": 1.5826114719690982, - "grad_norm": 25.843351364135742, - "learning_rate": 1.896524618236289e-05, - "loss": 4.7220813751220705, - "step": 6300 - }, - { - "epoch": 2.6456694974026145, - "grad_norm": 15.981147766113281, - "learning_rate": 4.999985708624277e-05, - "loss": 6.141958999633789, - "step": 6320 - }, - { - "epoch": 2.6540439396516753, - "grad_norm": 12.432202339172363, - "learning_rate": 4.999851618030373e-05, - "loss": 6.936581420898437, - "step": 6340 - }, - { - "epoch": 2.6624183819007365, - "grad_norm": 14.735859870910645, - "learning_rate": 4.9995763869538124e-05, - "loss": 7.491865539550782, - "step": 6360 - }, - { - "epoch": 2.670792824149798, - "grad_norm": 12.628633499145508, - "learning_rate": 4.999160030934015e-05, - "loss": 7.847127532958984, - "step": 6380 - }, - { - "epoch": 2.679167266398859, - "grad_norm": 14.50905990600586, - "learning_rate": 4.9986025734782515e-05, - "loss": 6.550693511962891, - "step": 6400 - }, - { - "epoch": 2.6875417086479203, - "grad_norm": 18.310335159301758, - "learning_rate": 4.997904046060316e-05, - "loss": 6.068325042724609, - "step": 6420 - }, - { - "epoch": 2.695916150896981, - "grad_norm": 12.4039888381958, - "learning_rate": 4.997064488118748e-05, - "loss": 5.966919708251953, - "step": 6440 - }, - { - "epoch": 2.7042905931460424, - "grad_norm": 14.189772605895996, - "learning_rate": 4.996083947054607e-05, - "loss": 5.816884613037109, - "step": 6460 - }, - { - "epoch": 2.7126650353951036, - "grad_norm": 11.678384780883789, - "learning_rate": 4.9949624782287944e-05, - "loss": 6.372280120849609, - "step": 6480 - }, - { - "epoch": 2.721039477644165, - "grad_norm": 13.517382621765137, - "learning_rate": 4.993700144958927e-05, - "loss": 5.818735504150391, - "step": 6500 - }, - { - "epoch": 2.7294139198932257, - "grad_norm": 11.332883834838867, - "learning_rate": 4.992297018515767e-05, - "loss": 5.053961181640625, - "step": 6520 - }, - { - "epoch": 2.737788362142287, - "grad_norm": 29.601367950439453, - "learning_rate": 4.990753178119193e-05, - "loss": 5.2946617126464846, - "step": 6540 - }, - { - "epoch": 2.746162804391348, - "grad_norm": 10.445046424865723, - "learning_rate": 4.989068710933732e-05, - "loss": 4.721809005737304, - "step": 6560 - }, - { - "epoch": 2.7545372466404094, - "grad_norm": 14.08784294128418, - "learning_rate": 4.9872437120636304e-05, - "loss": 4.68652572631836, - "step": 6580 - }, - { - "epoch": 2.7629116888894707, - "grad_norm": 10.382508277893066, - "learning_rate": 4.985278284547495e-05, - "loss": 4.90959358215332, - "step": 6600 - }, - { - "epoch": 2.7712861311385315, - "grad_norm": 9.340018272399902, - "learning_rate": 4.983172539352465e-05, - "loss": 4.94355354309082, - "step": 6620 - }, - { - "epoch": 2.7796605733875928, - "grad_norm": 11.943198204040527, - "learning_rate": 4.980926595367956e-05, - "loss": 4.442295074462891, - "step": 6640 - }, - { - "epoch": 2.788035015636654, - "grad_norm": 12.623554229736328, - "learning_rate": 4.978540579398941e-05, - "loss": 4.63629150390625, - "step": 6660 - }, - { - "epoch": 2.796409457885715, - "grad_norm": 9.113373756408691, - "learning_rate": 4.9760146261587936e-05, - "loss": 4.220952987670898, - "step": 6680 - }, - { - "epoch": 2.804783900134776, - "grad_norm": 9.733516693115234, - "learning_rate": 4.973348878261681e-05, - "loss": 4.143902206420899, - "step": 6700 - }, - { - "epoch": 2.8131583423838373, - "grad_norm": 10.420555114746094, - "learning_rate": 4.970543486214516e-05, - "loss": 4.141384506225586, - "step": 6720 - }, - { - "epoch": 2.8215327846328986, - "grad_norm": 7.3338212966918945, - "learning_rate": 4.967598608408454e-05, - "loss": 3.72052001953125, - "step": 6740 - }, - { - "epoch": 2.82990722688196, - "grad_norm": 10.468330383300781, - "learning_rate": 4.9645144111099515e-05, - "loss": 3.4560214996337892, - "step": 6760 - }, - { - "epoch": 2.8382816691310206, - "grad_norm": 10.967416763305664, - "learning_rate": 4.961291068451385e-05, - "loss": 3.2705413818359377, - "step": 6780 - }, - { - "epoch": 2.846656111380082, - "grad_norm": 12.951521873474121, - "learning_rate": 4.9579287624212093e-05, - "loss": 3.3012428283691406, - "step": 6800 - }, - { - "epoch": 2.855030553629143, - "grad_norm": 9.387336730957031, - "learning_rate": 4.9544276828536906e-05, - "loss": 3.2594974517822264, - "step": 6820 - }, - { - "epoch": 2.863404995878204, - "grad_norm": 8.693846702575684, - "learning_rate": 4.9507880274181854e-05, - "loss": 2.9371822357177733, - "step": 6840 - }, - { - "epoch": 2.871779438127265, - "grad_norm": 11.925010681152344, - "learning_rate": 4.94701000160798e-05, - "loss": 2.991471290588379, - "step": 6860 - }, - { - "epoch": 2.8801538803763265, - "grad_norm": 9.22618579864502, - "learning_rate": 4.943093818728689e-05, - "loss": 3.6338191986083985, - "step": 6880 - }, - { - "epoch": 2.8885283226253877, - "grad_norm": 9.440751075744629, - "learning_rate": 4.939039699886213e-05, - "loss": 2.745065689086914, - "step": 6900 - }, - { - "epoch": 2.896902764874449, - "grad_norm": 12.722249984741211, - "learning_rate": 4.9348478739742526e-05, - "loss": 2.652340126037598, - "step": 6920 - }, - { - "epoch": 2.90527720712351, - "grad_norm": 10.900734901428223, - "learning_rate": 4.930518577661388e-05, - "loss": 2.8322593688964846, - "step": 6940 - }, - { - "epoch": 2.913651649372571, - "grad_norm": 6.965799808502197, - "learning_rate": 4.926052055377716e-05, - "loss": 2.7872573852539064, - "step": 6960 - }, - { - "epoch": 2.9220260916216323, - "grad_norm": 6.74927282333374, - "learning_rate": 4.921448559301048e-05, - "loss": 2.9263822555541994, - "step": 6980 - }, - { - "epoch": 2.930400533870693, - "grad_norm": 8.862627029418945, - "learning_rate": 4.916708349342674e-05, - "loss": 2.499791717529297, - "step": 7000 - }, - { - "epoch": 2.9387749761197544, - "grad_norm": 10.037351608276367, - "learning_rate": 4.911831693132686e-05, - "loss": 2.275549125671387, - "step": 7020 - }, - { - "epoch": 2.9471494183688156, - "grad_norm": 6.2945075035095215, - "learning_rate": 4.906818866004871e-05, - "loss": 2.404999542236328, - "step": 7040 - }, - { - "epoch": 2.955523860617877, - "grad_norm": 4.414289951324463, - "learning_rate": 4.901670150981164e-05, - "loss": 2.1930471420288087, - "step": 7060 - }, - { - "epoch": 2.963898302866938, - "grad_norm": 5.771992206573486, - "learning_rate": 4.8963858387556665e-05, - "loss": 2.3143028259277343, - "step": 7080 - }, - { - "epoch": 2.9722727451159994, - "grad_norm": 6.430806636810303, - "learning_rate": 4.890966227678239e-05, - "loss": 2.1435062408447267, - "step": 7100 - }, - { - "epoch": 2.98064718736506, - "grad_norm": 4.818403244018555, - "learning_rate": 4.885411623737651e-05, - "loss": 1.8850128173828125, - "step": 7120 - }, - { - "epoch": 2.9890216296141214, - "grad_norm": 6.249518394470215, - "learning_rate": 4.87972234054431e-05, - "loss": 2.0096624374389647, - "step": 7140 - }, - { - "epoch": 2.9973960718631827, - "grad_norm": 7.708676815032959, - "learning_rate": 4.87389869931255e-05, - "loss": 1.9306636810302735, - "step": 7160 - }, - { - "epoch": 3.0054433874618898, - "grad_norm": 10.92924690246582, - "learning_rate": 4.8679410288425e-05, - "loss": 1.60665283203125, - "step": 7180 - }, - { - "epoch": 3.013817829710951, - "grad_norm": 8.802605628967285, - "learning_rate": 4.86184966550152e-05, - "loss": 1.6958501815795899, - "step": 7200 - }, - { - "epoch": 3.022192271960012, - "grad_norm": 11.568385124206543, - "learning_rate": 4.8556249532052045e-05, - "loss": 1.7447893142700195, - "step": 7220 - }, - { - "epoch": 3.030566714209073, - "grad_norm": 4.640422344207764, - "learning_rate": 4.8492672433979744e-05, - "loss": 1.7216526031494142, - "step": 7240 - }, - { - "epoch": 3.0389411564581343, - "grad_norm": 5.1430816650390625, - "learning_rate": 4.8427768950332264e-05, - "loss": 1.9419631958007812, - "step": 7260 - }, - { - "epoch": 3.0473155987071956, - "grad_norm": 4.505365371704102, - "learning_rate": 4.836154274553069e-05, - "loss": 1.4452963829040528, - "step": 7280 - }, - { - "epoch": 3.0556900409562564, - "grad_norm": 4.836411952972412, - "learning_rate": 4.829399755867636e-05, - "loss": 1.5338068962097169, - "step": 7300 - }, - { - "epoch": 3.0640644832053177, - "grad_norm": 6.093108177185059, - "learning_rate": 4.8225137203339734e-05, - "loss": 1.553403663635254, - "step": 7320 - }, - { - "epoch": 3.072438925454379, - "grad_norm": 6.984614372253418, - "learning_rate": 4.8154965567345073e-05, - "loss": 1.5232481956481934, - "step": 7340 - }, - { - "epoch": 3.08081336770344, - "grad_norm": 5.966179370880127, - "learning_rate": 4.8083486612550954e-05, - "loss": 1.9450897216796874, - "step": 7360 - }, - { - "epoch": 3.0891878099525014, - "grad_norm": 8.660058975219727, - "learning_rate": 4.801070437462659e-05, - "loss": 1.4149410247802734, - "step": 7380 - }, - { - "epoch": 3.0975622522015622, - "grad_norm": 4.126588821411133, - "learning_rate": 4.793662296282393e-05, - "loss": 1.0894225120544434, - "step": 7400 - }, - { - "epoch": 3.1059366944506235, - "grad_norm": 9.536477088928223, - "learning_rate": 4.7861246559745725e-05, - "loss": 1.519902515411377, - "step": 7420 - }, - { - "epoch": 3.1143111366996847, - "grad_norm": 5.378171443939209, - "learning_rate": 4.7784579421109334e-05, - "loss": 1.1999958038330079, - "step": 7440 - }, - { - "epoch": 3.122685578948746, - "grad_norm": 4.443706512451172, - "learning_rate": 4.770662587550643e-05, - "loss": 1.2191826820373535, - "step": 7460 - }, - { - "epoch": 3.131060021197807, - "grad_norm": 4.879043102264404, - "learning_rate": 4.762739032415868e-05, - "loss": 1.2340889930725099, - "step": 7480 - }, - { - "epoch": 3.139434463446868, - "grad_norm": 4.470728874206543, - "learning_rate": 4.754687724066917e-05, - "loss": 1.423619270324707, - "step": 7500 - }, - { - "epoch": 3.1478089056959293, - "grad_norm": 5.47915506362915, - "learning_rate": 4.746509117076991e-05, - "loss": 1.251408863067627, - "step": 7520 - }, - { - "epoch": 3.1561833479449906, - "grad_norm": 4.566497325897217, - "learning_rate": 4.7382036732065104e-05, - "loss": 1.0826839447021483, - "step": 7540 - }, - { - "epoch": 3.1645577901940514, - "grad_norm": 6.905686855316162, - "learning_rate": 4.72977186137705e-05, - "loss": 1.069216251373291, - "step": 7560 - }, - { - "epoch": 3.1729322324431126, - "grad_norm": 2.964932680130005, - "learning_rate": 4.721214157644861e-05, - "loss": 1.1391191482543945, - "step": 7580 - }, - { - "epoch": 3.181306674692174, - "grad_norm": 6.019253253936768, - "learning_rate": 4.712531045173996e-05, - "loss": 0.9639101028442383, - "step": 7600 - }, - { - "epoch": 3.189681116941235, - "grad_norm": 3.0169851779937744, - "learning_rate": 4.703723014209025e-05, - "loss": 0.934177017211914, - "step": 7620 - }, - { - "epoch": 3.198055559190296, - "grad_norm": 5.5374369621276855, - "learning_rate": 4.694790562047361e-05, - "loss": 0.921787166595459, - "step": 7640 - }, - { - "epoch": 3.206430001439357, - "grad_norm": 3.7313361167907715, - "learning_rate": 4.685734193011182e-05, - "loss": 0.9222763061523438, - "step": 7660 - }, - { - "epoch": 3.2148044436884184, - "grad_norm": 3.172354221343994, - "learning_rate": 4.676554418418953e-05, - "loss": 0.9454127311706543, - "step": 7680 - }, - { - "epoch": 3.2231788859374797, - "grad_norm": 2.3172996044158936, - "learning_rate": 4.667251756556563e-05, - "loss": 1.050184154510498, - "step": 7700 - }, - { - "epoch": 3.2315533281865405, - "grad_norm": 5.589139938354492, - "learning_rate": 4.657826732648061e-05, - "loss": 0.7619458675384522, - "step": 7720 - }, - { - "epoch": 3.2399277704356018, - "grad_norm": 4.144975662231445, - "learning_rate": 4.648279878825998e-05, - "loss": 0.8281337738037109, - "step": 7740 - }, - { - "epoch": 3.248302212684663, - "grad_norm": 3.205533266067505, - "learning_rate": 4.63861173410139e-05, - "loss": 0.8384788513183594, - "step": 7760 - }, - { - "epoch": 3.2566766549337243, - "grad_norm": 4.673986911773682, - "learning_rate": 4.628822844333278e-05, - "loss": 0.7749058723449707, - "step": 7780 - } - ], - "logging_steps": 20, - "max_steps": 8362, - "num_input_tokens_seen": 0, - "num_train_epochs": 4, - "save_steps": 1000000000, - "stateful_callbacks": { - "TrainerControl": { - "args": { - "should_epoch_stop": false, - "should_evaluate": false, - "should_log": false, - "should_save": true, - "should_training_stop": false - }, - "attributes": {} - } - }, - "total_flos": 1.825226422111411e+16, - "train_batch_size": 1, - "trial_name": null, - "trial_params": null -} diff --git a/slots/9/latest.json b/slots/9/latest.json deleted file mode 100644 index 6167f617114ce41b4b0bbfa88af2b44a4c235939..0000000000000000000000000000000000000000 --- a/slots/9/latest.json +++ /dev/null @@ -1 +0,0 @@ -{"worker_id": "slot:9", "checkpoint": "slots/9/checkpoint-7793", "step": 7793, "updated_at": 1776416983}