| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9850746268656714, | |
| "eval_steps": 500, | |
| "global_step": 75, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03980099502487562, | |
| "grad_norm": 0.49672797322273254, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.634, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07960199004975124, | |
| "grad_norm": 0.4738655984401703, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.6666, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11940298507462686, | |
| "grad_norm": 0.4305928945541382, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.6673, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15920398009950248, | |
| "grad_norm": 0.2917488217353821, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5859, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19900497512437812, | |
| "grad_norm": 0.3740411102771759, | |
| "learning_rate": 6.25e-05, | |
| "loss": 1.641, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23880597014925373, | |
| "grad_norm": 0.3605235815048218, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.5957, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.27860696517412936, | |
| "grad_norm": 0.2877049744129181, | |
| "learning_rate": 8.75e-05, | |
| "loss": 1.5118, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.31840796019900497, | |
| "grad_norm": 0.2710678279399872, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5322, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.3582089552238806, | |
| "grad_norm": 0.26977717876434326, | |
| "learning_rate": 9.994504457428558e-05, | |
| "loss": 1.5521, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.39800995024875624, | |
| "grad_norm": 0.2279333472251892, | |
| "learning_rate": 9.978029910109491e-05, | |
| "loss": 1.5501, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.43781094527363185, | |
| "grad_norm": 0.22376185655593872, | |
| "learning_rate": 9.950612572673255e-05, | |
| "loss": 1.5476, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.47761194029850745, | |
| "grad_norm": 0.22192299365997314, | |
| "learning_rate": 9.91231271437788e-05, | |
| "loss": 1.5191, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5174129353233831, | |
| "grad_norm": 0.20543566346168518, | |
| "learning_rate": 9.863214526624065e-05, | |
| "loss": 1.5528, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5572139303482587, | |
| "grad_norm": 0.20152153074741364, | |
| "learning_rate": 9.8034259378842e-05, | |
| "loss": 1.4816, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.204048290848732, | |
| "learning_rate": 9.733078376452171e-05, | |
| "loss": 1.5519, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6368159203980099, | |
| "grad_norm": 0.19486752152442932, | |
| "learning_rate": 9.652326481535435e-05, | |
| "loss": 1.5183, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6766169154228856, | |
| "grad_norm": 0.1839868277311325, | |
| "learning_rate": 9.561347763324484e-05, | |
| "loss": 1.5272, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7164179104477612, | |
| "grad_norm": 0.1841403841972351, | |
| "learning_rate": 9.460342212786932e-05, | |
| "loss": 1.523, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7562189054726368, | |
| "grad_norm": 0.17461948096752167, | |
| "learning_rate": 9.349531862043952e-05, | |
| "loss": 1.5316, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7960199004975125, | |
| "grad_norm": 0.1808539479970932, | |
| "learning_rate": 9.229160296295488e-05, | |
| "loss": 1.4991, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.835820895522388, | |
| "grad_norm": 0.1901710480451584, | |
| "learning_rate": 9.099492118367123e-05, | |
| "loss": 1.5086, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8756218905472637, | |
| "grad_norm": 0.183896005153656, | |
| "learning_rate": 8.960812367055646e-05, | |
| "loss": 1.5435, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.9154228855721394, | |
| "grad_norm": 0.17992867529392242, | |
| "learning_rate": 8.81342589055191e-05, | |
| "loss": 1.5201, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9552238805970149, | |
| "grad_norm": 0.17739859223365784, | |
| "learning_rate": 8.657656676318346e-05, | |
| "loss": 1.5059, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9950248756218906, | |
| "grad_norm": 0.17343272268772125, | |
| "learning_rate": 8.493847138894209e-05, | |
| "loss": 1.5023, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0348258706467661, | |
| "grad_norm": 0.181735098361969, | |
| "learning_rate": 8.322357367194109e-05, | |
| "loss": 1.4237, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0746268656716418, | |
| "grad_norm": 0.1831899732351303, | |
| "learning_rate": 8.143564332954425e-05, | |
| "loss": 1.4127, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.1144278606965174, | |
| "grad_norm": 0.17571744322776794, | |
| "learning_rate": 7.957861062067614e-05, | |
| "loss": 1.415, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.154228855721393, | |
| "grad_norm": 0.17378666996955872, | |
| "learning_rate": 7.765655770625997e-05, | |
| "loss": 1.3883, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 0.18887346982955933, | |
| "learning_rate": 7.56737096757421e-05, | |
| "loss": 1.3688, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2338308457711442, | |
| "grad_norm": 0.1842719465494156, | |
| "learning_rate": 7.363442525942826e-05, | |
| "loss": 1.3973, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2736318407960199, | |
| "grad_norm": 0.180495485663414, | |
| "learning_rate": 7.154318724704853e-05, | |
| "loss": 1.4196, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3134328358208955, | |
| "grad_norm": 0.18591800332069397, | |
| "learning_rate": 6.940459263361249e-05, | |
| "loss": 1.3993, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.3532338308457712, | |
| "grad_norm": 0.1974848359823227, | |
| "learning_rate": 6.722334251421665e-05, | |
| "loss": 1.4004, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3930348258706466, | |
| "grad_norm": 0.18464985489845276, | |
| "learning_rate": 6.500423175001705e-05, | |
| "loss": 1.4223, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4328358208955223, | |
| "grad_norm": 0.19280724227428436, | |
| "learning_rate": 6.275213842808383e-05, | |
| "loss": 1.4227, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.472636815920398, | |
| "grad_norm": 0.2069428563117981, | |
| "learning_rate": 6.0472013138307235e-05, | |
| "loss": 1.4064, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5124378109452736, | |
| "grad_norm": 0.19798892736434937, | |
| "learning_rate": 5.816886809092651e-05, | |
| "loss": 1.4589, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5522388059701493, | |
| "grad_norm": 0.19987910985946655, | |
| "learning_rate": 5.584776609860414e-05, | |
| "loss": 1.3616, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5920398009950247, | |
| "grad_norm": 0.19689033925533295, | |
| "learning_rate": 5.351380944726465e-05, | |
| "loss": 1.402, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6318407960199006, | |
| "grad_norm": 0.20285791158676147, | |
| "learning_rate": 5.117212868016303e-05, | |
| "loss": 1.4233, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.671641791044776, | |
| "grad_norm": 0.20456896722316742, | |
| "learning_rate": 4.882787131983698e-05, | |
| "loss": 1.3909, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.7114427860696517, | |
| "grad_norm": 0.1989758014678955, | |
| "learning_rate": 4.648619055273537e-05, | |
| "loss": 1.4099, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7512437810945274, | |
| "grad_norm": 0.20214244723320007, | |
| "learning_rate": 4.415223390139588e-05, | |
| "loss": 1.4281, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.20304188132286072, | |
| "learning_rate": 4.183113190907349e-05, | |
| "loss": 1.3892, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.8308457711442787, | |
| "grad_norm": 0.20430393517017365, | |
| "learning_rate": 3.952798686169279e-05, | |
| "loss": 1.3681, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8706467661691542, | |
| "grad_norm": 0.216043621301651, | |
| "learning_rate": 3.7247861571916185e-05, | |
| "loss": 1.4124, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.9104477611940298, | |
| "grad_norm": 0.20856983959674835, | |
| "learning_rate": 3.499576824998298e-05, | |
| "loss": 1.3772, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9502487562189055, | |
| "grad_norm": 0.21441003680229187, | |
| "learning_rate": 3.277665748578336e-05, | |
| "loss": 1.3912, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9900497512437811, | |
| "grad_norm": 0.2081148624420166, | |
| "learning_rate": 3.0595407366387504e-05, | |
| "loss": 1.4066, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.029850746268657, | |
| "grad_norm": 0.21571974456310272, | |
| "learning_rate": 2.8456812752951485e-05, | |
| "loss": 1.2869, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.0696517412935322, | |
| "grad_norm": 0.22047464549541473, | |
| "learning_rate": 2.636557474057173e-05, | |
| "loss": 1.3308, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.109452736318408, | |
| "grad_norm": 0.21658696234226227, | |
| "learning_rate": 2.4326290324257894e-05, | |
| "loss": 1.3483, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1492537313432836, | |
| "grad_norm": 0.22115129232406616, | |
| "learning_rate": 2.234344229374003e-05, | |
| "loss": 1.3131, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.189054726368159, | |
| "grad_norm": 0.21616297960281372, | |
| "learning_rate": 2.042138937932388e-05, | |
| "loss": 1.32, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.228855721393035, | |
| "grad_norm": 0.2277287244796753, | |
| "learning_rate": 1.8564356670455767e-05, | |
| "loss": 1.3285, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.2686567164179103, | |
| "grad_norm": 0.21879006922245026, | |
| "learning_rate": 1.677642632805892e-05, | |
| "loss": 1.3478, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.308457711442786, | |
| "grad_norm": 0.2215149849653244, | |
| "learning_rate": 1.5061528611057918e-05, | |
| "loss": 1.3001, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.3482587064676617, | |
| "grad_norm": 0.2294432371854782, | |
| "learning_rate": 1.3423433236816563e-05, | |
| "loss": 1.288, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 0.22794978320598602, | |
| "learning_rate": 1.1865741094480909e-05, | |
| "loss": 1.2953, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.427860696517413, | |
| "grad_norm": 0.22905172407627106, | |
| "learning_rate": 1.0391876329443533e-05, | |
| "loss": 1.3366, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.4676616915422884, | |
| "grad_norm": 0.2197260707616806, | |
| "learning_rate": 9.005078816328771e-06, | |
| "loss": 1.3045, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.5074626865671643, | |
| "grad_norm": 0.23517443239688873, | |
| "learning_rate": 7.708397037045129e-06, | |
| "loss": 1.2732, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.5472636815920398, | |
| "grad_norm": 0.22419849038124084, | |
| "learning_rate": 6.50468137956049e-06, | |
| "loss": 1.3236, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.587064676616915, | |
| "grad_norm": 0.22183862328529358, | |
| "learning_rate": 5.3965778721306755e-06, | |
| "loss": 1.3171, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.626865671641791, | |
| "grad_norm": 0.2313942015171051, | |
| "learning_rate": 4.386522366755169e-06, | |
| "loss": 1.3622, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.22842709720134735, | |
| "learning_rate": 3.476735184645674e-06, | |
| "loss": 1.3202, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.7064676616915424, | |
| "grad_norm": 0.21870732307434082, | |
| "learning_rate": 2.6692162354782944e-06, | |
| "loss": 1.3243, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.746268656716418, | |
| "grad_norm": 0.2380901575088501, | |
| "learning_rate": 1.9657406211579966e-06, | |
| "loss": 1.2638, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.7860696517412933, | |
| "grad_norm": 0.2310485988855362, | |
| "learning_rate": 1.3678547337593494e-06, | |
| "loss": 1.3266, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.825870646766169, | |
| "grad_norm": 0.22601790726184845, | |
| "learning_rate": 8.768728562211947e-07, | |
| "loss": 1.3183, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8656716417910446, | |
| "grad_norm": 0.22665373980998993, | |
| "learning_rate": 4.938742732674529e-07, | |
| "loss": 1.3247, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.9054726368159205, | |
| "grad_norm": 0.22332097589969635, | |
| "learning_rate": 2.1970089890509527e-07, | |
| "loss": 1.3518, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.945273631840796, | |
| "grad_norm": 0.2267157882452011, | |
| "learning_rate": 5.4955425714431353e-08, | |
| "loss": 1.2951, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.23764650523662567, | |
| "learning_rate": 0.0, | |
| "loss": 1.3156, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "step": 75, | |
| "total_flos": 1.7993258830774927e+18, | |
| "train_loss": 1.423751606941223, | |
| "train_runtime": 7978.1931, | |
| "train_samples_per_second": 0.604, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 75, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7993258830774927e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |