| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.980891719745223, | |
| "eval_steps": 500, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 6.475443363189697, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9435, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 6.0149922370910645, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9088, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 6.234716892242432, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.946, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 5.534856796264648, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9178, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 4.338842868804932, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.8717, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 2.4903976917266846, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7575, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 3.8526647090911865, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7954, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 4.179830551147461, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8056, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 3.9865660667419434, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.7168, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 4.04654598236084, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.7573, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 2.9917404651641846, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.7598, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 2.017793655395508, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.7146, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 2.0348641872406006, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.7697, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 1.8617016077041626, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6807, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 1.5839647054672241, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6682, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 1.2628346681594849, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6205, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 1.4339826107025146, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6606, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 1.0776035785675049, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6367, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 1.2484275102615356, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.7081, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 1.166043996810913, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6139, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 1.0878933668136597, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6364, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 0.8449444770812988, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6327, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 0.9466868042945862, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.7027, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 0.9680841565132141, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6549, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 0.7421926259994507, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6129, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 0.8124146461486816, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.609, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 1.6263659000396729, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.9629, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 0.7524719834327698, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.6027, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 0.592254102230072, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5297, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 0.7852796912193298, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.6182, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 0.830742359161377, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.6581, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 0.6159372925758362, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.6236, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 0.7226252555847168, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.618, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2993630573248407, | |
| "grad_norm": 0.6193709373474121, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.542, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3375796178343948, | |
| "grad_norm": 0.7392263412475586, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.53, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3757961783439492, | |
| "grad_norm": 0.6274488568305969, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6762, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4140127388535033, | |
| "grad_norm": 0.5716177821159363, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5749, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4522292993630574, | |
| "grad_norm": 0.7565421462059021, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5814, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.4904458598726116, | |
| "grad_norm": 0.7007409334182739, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.6141, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5286624203821657, | |
| "grad_norm": 0.5515427589416504, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.4922, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5668789808917198, | |
| "grad_norm": 0.701950192451477, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5835, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.605095541401274, | |
| "grad_norm": 0.6372572779655457, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5014, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.643312101910828, | |
| "grad_norm": 0.6354508996009827, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6538, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6815286624203822, | |
| "grad_norm": 0.5548213720321655, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5819, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7197452229299364, | |
| "grad_norm": 0.4872737228870392, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.4779, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7579617834394905, | |
| "grad_norm": 0.5852910876274109, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.6624, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7961783439490446, | |
| "grad_norm": 0.49762701988220215, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5426, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8343949044585988, | |
| "grad_norm": 0.5693338513374329, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5471, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.872611464968153, | |
| "grad_norm": 0.6038720607757568, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6022, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 0.49600663781166077, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.507, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9490445859872612, | |
| "grad_norm": 0.5460913181304932, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5469, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9872611464968153, | |
| "grad_norm": 0.530115008354187, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5327, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0254777070063694, | |
| "grad_norm": 1.14858877658844, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.837, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.0636942675159236, | |
| "grad_norm": 0.5498645901679993, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.5906, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1019108280254777, | |
| "grad_norm": 0.5188542604446411, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.5788, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.140127388535032, | |
| "grad_norm": 0.4956539273262024, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5361, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.178343949044586, | |
| "grad_norm": 0.6037235260009766, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5584, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.21656050955414, | |
| "grad_norm": 0.4257795810699463, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4353, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.254777070063694, | |
| "grad_norm": 0.5109358429908752, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5412, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.2929936305732483, | |
| "grad_norm": 0.4674205780029297, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.5459, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3312101910828025, | |
| "grad_norm": 0.47286278009414673, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4947, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3694267515923566, | |
| "grad_norm": 0.49508610367774963, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.4856, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.4076433121019107, | |
| "grad_norm": 0.42003560066223145, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.4792, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.445859872611465, | |
| "grad_norm": 0.48869258165359497, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.5505, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.484076433121019, | |
| "grad_norm": 0.5102881789207458, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.4618, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.522292993630573, | |
| "grad_norm": 0.5060379505157471, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5975, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5605095541401273, | |
| "grad_norm": 0.4596272110939026, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5184, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.5987261146496814, | |
| "grad_norm": 0.4477904736995697, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.4971, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.6369426751592355, | |
| "grad_norm": 0.4290829002857208, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.4872, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6751592356687897, | |
| "grad_norm": 0.4998835325241089, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4768, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.713375796178344, | |
| "grad_norm": 0.5172039270401001, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5693, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7515923566878984, | |
| "grad_norm": 0.4305104613304138, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5935, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.789808917197452, | |
| "grad_norm": 0.4256463348865509, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.511, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.8280254777070066, | |
| "grad_norm": 0.5269811153411865, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.577, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 0.4299260377883911, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5992, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.904458598726115, | |
| "grad_norm": 0.45451387763023376, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.5475, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9426751592356686, | |
| "grad_norm": 0.4655923843383789, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5231, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.980891719745223, | |
| "grad_norm": 0.4951322376728058, | |
| "learning_rate": 0.0, | |
| "loss": 0.5876, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.980891719745223, | |
| "step": 78, | |
| "total_flos": 47460738555904.0, | |
| "train_loss": 0.6236666376009966, | |
| "train_runtime": 2174.9739, | |
| "train_samples_per_second": 3.448, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 78, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 47460738555904.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |