| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.962025316455696, | |
| "eval_steps": 500, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0379746835443038, | |
| "grad_norm": 6.322867828766394, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.1853, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0759493670886076, | |
| "grad_norm": 6.419631229377136, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.1894, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11392405063291139, | |
| "grad_norm": 6.123859837139322, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.1819, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 5.730496688797018, | |
| "learning_rate": 5e-06, | |
| "loss": 1.185, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.189873417721519, | |
| "grad_norm": 4.263174722983781, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.1505, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22784810126582278, | |
| "grad_norm": 2.683123108816851, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.0975, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26582278481012656, | |
| "grad_norm": 4.8764045153049205, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.1437, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 4.7747202313088435, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1449, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34177215189873417, | |
| "grad_norm": 4.090362793015124, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 1.0421, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 3.867556871616575, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 1.0195, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4177215189873418, | |
| "grad_norm": 2.8611506917628042, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 1.0317, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 2.055543563023435, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.9858, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4936708860759494, | |
| "grad_norm": 1.8665101708862306, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.9341, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5316455696202531, | |
| "grad_norm": 1.7907078513357393, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.9768, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.569620253164557, | |
| "grad_norm": 1.2956772422882297, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.9264, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.08733731828189, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.9285, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6455696202531646, | |
| "grad_norm": 1.4807598553358707, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.9173, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6835443037974683, | |
| "grad_norm": 1.2675456491692723, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.9608, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7215189873417721, | |
| "grad_norm": 1.135845455230578, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.8877, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.9906625674370299, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.9001, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7974683544303798, | |
| "grad_norm": 0.988852353015704, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.891, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8354430379746836, | |
| "grad_norm": 0.9352575141488356, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.8795, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8734177215189873, | |
| "grad_norm": 1.0810158792621125, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.8951, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.9388467231306077, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.8726, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9493670886075949, | |
| "grad_norm": 0.8905205504959225, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.8657, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "grad_norm": 0.9059008606597845, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.8881, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0253164556962024, | |
| "grad_norm": 1.50914294933104, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 1.4054, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0632911392405062, | |
| "grad_norm": 0.8989969653837433, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.828, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.7365504391952206, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.7757, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 0.7771353651547056, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.8128, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1772151898734178, | |
| "grad_norm": 0.7329648407385149, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.8271, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2151898734177216, | |
| "grad_norm": 0.8708541012074602, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.7949, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.8219398900265773, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.8218, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2911392405063291, | |
| "grad_norm": 0.6836647905831809, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.76, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3291139240506329, | |
| "grad_norm": 0.7740164049545867, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.7939, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3670886075949367, | |
| "grad_norm": 0.7112407820299882, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.7473, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.944362419762214, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.8813, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4430379746835442, | |
| "grad_norm": 0.6578453724602004, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.831, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.481012658227848, | |
| "grad_norm": 0.7308183225469851, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.8186, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 0.7606811132589464, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.844, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.6709329345610502, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.8697, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.5949367088607596, | |
| "grad_norm": 0.671829426021201, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.8231, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6329113924050633, | |
| "grad_norm": 0.6258079452625654, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8021, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6708860759493671, | |
| "grad_norm": 0.5987765696750919, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.8431, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.5497792910562198, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.7526, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7468354430379747, | |
| "grad_norm": 0.6859801502170986, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.848, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7848101265822784, | |
| "grad_norm": 0.6259828677236124, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.8349, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8227848101265822, | |
| "grad_norm": 0.6429380679492415, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.776, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.5800699210464617, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.846, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 0.589176986601615, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.764, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9367088607594938, | |
| "grad_norm": 0.6136580690668652, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.7651, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9746835443037973, | |
| "grad_norm": 0.6677819073651702, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.8124, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0126582278481013, | |
| "grad_norm": 1.2086976350637177, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 1.2474, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.5186346707538071, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.7472, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.088607594936709, | |
| "grad_norm": 0.5902793966699749, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.7427, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1265822784810124, | |
| "grad_norm": 0.6013248632197061, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.7223, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1645569620253164, | |
| "grad_norm": 0.5206928883036986, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.7785, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.545725395907017, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.8173, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.240506329113924, | |
| "grad_norm": 0.5131612572298808, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.6874, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 0.4972635814485538, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.7127, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3164556962025316, | |
| "grad_norm": 0.595555626926528, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.8729, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.49132587331935235, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.6866, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.392405063291139, | |
| "grad_norm": 0.5800188999639047, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.8314, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.430379746835443, | |
| "grad_norm": 0.5075531476820602, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.767, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.4683544303797467, | |
| "grad_norm": 0.49282082341101324, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.6697, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.522571820796664, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.7851, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5443037974683547, | |
| "grad_norm": 0.508786398558625, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.7497, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.5822784810126582, | |
| "grad_norm": 0.48335927152289127, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.8061, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.620253164556962, | |
| "grad_norm": 0.4857675842060652, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.725, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.550727121800424, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.8015, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.6962025316455698, | |
| "grad_norm": 0.5142894719489793, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.7381, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7341772151898733, | |
| "grad_norm": 0.5076349944627733, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.7199, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7721518987341773, | |
| "grad_norm": 0.4974431134522027, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.771, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.5106869627411758, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.7902, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.848101265822785, | |
| "grad_norm": 0.5204117806778187, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.7437, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.8860759493670884, | |
| "grad_norm": 0.5267960128365246, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.7574, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9240506329113924, | |
| "grad_norm": 0.4977662918975598, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.7778, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.5091785177405566, | |
| "learning_rate": 0.0, | |
| "loss": 0.7852, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "step": 78, | |
| "total_flos": 61080682725376.0, | |
| "train_loss": 0.8717104089565766, | |
| "train_runtime": 1444.7973, | |
| "train_samples_per_second": 5.191, | |
| "train_steps_per_second": 0.054 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "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": 61080682725376.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |