| { | |
| "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": 6.485745548246275, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.1878, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0759493670886076, | |
| "grad_norm": 6.465822217568478, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.1735, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11392405063291139, | |
| "grad_norm": 6.195277506391659, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.1647, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 5.799769823526993, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1608, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.189873417721519, | |
| "grad_norm": 4.3450177180199, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.1421, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22784810126582278, | |
| "grad_norm": 2.463610967994665, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.0396, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26582278481012656, | |
| "grad_norm": 4.577891696246731, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.1231, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 4.503384198538751, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1075, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34177215189873417, | |
| "grad_norm": 3.8310498612257686, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.9965, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 3.5638344423615647, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.9821, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4177215189873418, | |
| "grad_norm": 2.881786623869597, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 1.0366, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.9488810814561746, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.9426, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4936708860759494, | |
| "grad_norm": 1.8755855672690618, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.8988, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5316455696202531, | |
| "grad_norm": 1.8486508629321972, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.9482, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.569620253164557, | |
| "grad_norm": 1.3806887060387854, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.9246, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.2204733734134552, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.9037, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6455696202531646, | |
| "grad_norm": 1.2853760258931963, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.8985, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6835443037974683, | |
| "grad_norm": 1.2437773881234602, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.9168, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7215189873417721, | |
| "grad_norm": 1.146833630595906, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.8647, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 1.044550835994215, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.8802, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7974683544303798, | |
| "grad_norm": 1.1571835712465777, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.8567, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8354430379746836, | |
| "grad_norm": 1.0549337977158728, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.8573, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8734177215189873, | |
| "grad_norm": 0.9935987376050021, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.8795, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.9568003714663316, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.85, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9493670886075949, | |
| "grad_norm": 1.084326509880812, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.8239, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "grad_norm": 1.0213309692673835, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.8843, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0253164556962024, | |
| "grad_norm": 1.5805051692575611, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 1.366, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0632911392405062, | |
| "grad_norm": 0.8244324641459649, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.784, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.7560188526972407, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.7476, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 0.9004810483037317, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.7522, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1772151898734178, | |
| "grad_norm": 0.6657243853335917, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.7875, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2151898734177216, | |
| "grad_norm": 0.7897561995668748, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.7827, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.8490295067257841, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.7995, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2911392405063291, | |
| "grad_norm": 0.7054475226304026, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.7471, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3291139240506329, | |
| "grad_norm": 0.6714021426026477, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.7856, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3670886075949367, | |
| "grad_norm": 0.7079475124768483, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6909, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.8283603594737133, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.8468, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4430379746835442, | |
| "grad_norm": 0.762781447849204, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.7929, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.481012658227848, | |
| "grad_norm": 0.737050531054483, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.7777, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 0.7681887147382744, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.8305, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.6735546654884726, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.8163, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.5949367088607596, | |
| "grad_norm": 0.6528315419344399, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.7768, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6329113924050633, | |
| "grad_norm": 0.6443387987382299, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7422, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6708860759493671, | |
| "grad_norm": 0.6935358728809672, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.8107, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.6346001503190914, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.7577, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7468354430379747, | |
| "grad_norm": 0.7389755097450126, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.808, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7848101265822784, | |
| "grad_norm": 0.6701646868155358, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.7899, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8227848101265822, | |
| "grad_norm": 0.6061610450983355, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.7518, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.6113450822488097, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.8113, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 0.6942091005029863, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.765, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9367088607594938, | |
| "grad_norm": 0.6050143971344853, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.7156, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9746835443037973, | |
| "grad_norm": 0.7105617556204705, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.7184, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0126582278481013, | |
| "grad_norm": 1.3246392741326032, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 1.214, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.6417792043489915, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.7168, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.088607594936709, | |
| "grad_norm": 0.5964202254877964, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.6855, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1265822784810124, | |
| "grad_norm": 0.5915140999717906, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.6483, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1645569620253164, | |
| "grad_norm": 0.6027347895777584, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.7646, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.6171953791580085, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.7752, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.240506329113924, | |
| "grad_norm": 0.5884028276946883, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.6389, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 0.5954292676999796, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.7054, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3164556962025316, | |
| "grad_norm": 0.6614306100294592, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.8011, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.5442238857296332, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.6654, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.392405063291139, | |
| "grad_norm": 0.5398911342214016, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.7555, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.430379746835443, | |
| "grad_norm": 0.5530982801705516, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.7548, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.4683544303797467, | |
| "grad_norm": 0.5638744668606761, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.6397, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.5564528408103611, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.7229, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5443037974683547, | |
| "grad_norm": 0.5636248726836497, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.7062, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.5822784810126582, | |
| "grad_norm": 0.5866165147488809, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.747, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.620253164556962, | |
| "grad_norm": 0.5143692009437217, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.6617, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.5633167411619424, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.7664, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.6962025316455698, | |
| "grad_norm": 0.5897836640010781, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.7053, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7341772151898733, | |
| "grad_norm": 0.5866463117478233, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.7124, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7721518987341773, | |
| "grad_norm": 0.5214543757917679, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.7179, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.5198883021466474, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.7522, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.848101265822785, | |
| "grad_norm": 0.5664317042404885, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.6989, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.8860759493670884, | |
| "grad_norm": 0.5580909347600271, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.729, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9240506329113924, | |
| "grad_norm": 0.57492441797178, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.7252, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.5767576246482665, | |
| "learning_rate": 0.0, | |
| "loss": 0.7423, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "step": 78, | |
| "total_flos": 48341369913344.0, | |
| "train_loss": 0.8378327083893311, | |
| "train_runtime": 1191.9993, | |
| "train_samples_per_second": 6.292, | |
| "train_steps_per_second": 0.065 | |
| } | |
| ], | |
| "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": 48341369913344.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |