| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9904153354632586, | |
| "eval_steps": 500, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 6.96260404586792, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.256, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 6.960289001464844, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.272, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 6.674771308898926, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.2149, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 6.231781482696533, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1908, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 4.5194010734558105, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.1785, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 2.962165355682373, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.1088, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 5.8254714012146, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.1322, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 5.946160793304443, | |
| "learning_rate": 1e-05, | |
| "loss": 1.167, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 5.620875835418701, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 1.111, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 5.018191814422607, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 1.0709, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 3.36198353767395, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 1.0263, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 2.0939364433288574, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 1.0423, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 2.189216375350952, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 1.0537, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 1.9374192953109741, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.9397, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 1.5520998239517212, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.9728, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 1.2040224075317383, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.978, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 1.338582158088684, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.9721, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 1.3925938606262207, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.9182, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 0.9509784579277039, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.9423, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 1.1120100021362305, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.9681, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 1.0758018493652344, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.9246, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.9043078422546387, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.9107, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.9808814525604248, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.9757, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.960244357585907, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.9087, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.7865895628929138, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.9069, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 1.1364617347717285, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.9176, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 2.6127283573150635, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 1.7353, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.9426343441009521, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.8718, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.7166311144828796, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.8767, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.7416899800300598, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.8519, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 1.011359691619873, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.8712, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.7813844084739685, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.8017, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.719464898109436, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.9768, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.7463328838348389, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.8451, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.7480574250221252, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.8488, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.650842010974884, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.8201, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.6185774803161621, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.9023, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.6526139378547668, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.903, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.6962924599647522, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.8806, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.6151214241981506, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.8545, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.6207312941551208, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.8064, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.6180050373077393, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.9111, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.6145247220993042, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8251, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.6231905221939087, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.8202, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.603053629398346, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.8841, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.5454174280166626, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.8646, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.6569262742996216, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.9252, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.5513457655906677, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.7664, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.6261488795280457, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.9086, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.5807673931121826, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.8572, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.6229838728904724, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.83, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.5468594431877136, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.7902, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 1.7930537462234497, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 1.6136, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.6046034693717957, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.8294, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.4906443655490875, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.7905, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.5799528956413269, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.7976, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.5433791279792786, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.845, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.5426772832870483, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.7831, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.54535973072052, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.8578, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.5305079221725464, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.7821, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.5088701844215393, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.8223, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.5119186043739319, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.8712, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.43340709805488586, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.7531, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.49028053879737854, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.7817, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.47262850403785706, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.8451, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.46987396478652954, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.8427, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.4778902232646942, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.8761, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.4655120074748993, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.8348, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.4640553593635559, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.821, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.4409966468811035, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.7615, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.5229158401489258, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.8042, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.4807683825492859, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.8352, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.42318591475486755, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.7691, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.5227220058441162, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.8709, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.45626455545425415, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.7912, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.4580370783805847, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.8722, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.46287021040916443, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.8025, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.5080867409706116, | |
| "learning_rate": 0.0, | |
| "loss": 0.7607, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 78, | |
| "total_flos": 73989648588800.0, | |
| "train_loss": 0.9269619354834924, | |
| "train_runtime": 7429.9065, | |
| "train_samples_per_second": 1.009, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "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": 73989648588800.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |