| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 78, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.038461538461538464, |
| "grad_norm": 6.914577007293701, |
| "learning_rate": 1.25e-06, |
| "loss": 1.0609, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.07692307692307693, |
| "grad_norm": 6.80674409866333, |
| "learning_rate": 2.5e-06, |
| "loss": 1.038, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.11538461538461539, |
| "grad_norm": 6.7133402824401855, |
| "learning_rate": 3.7500000000000005e-06, |
| "loss": 1.0364, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.15384615384615385, |
| "grad_norm": 6.110616207122803, |
| "learning_rate": 5e-06, |
| "loss": 1.0249, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.19230769230769232, |
| "grad_norm": 4.418550968170166, |
| "learning_rate": 6.25e-06, |
| "loss": 0.9886, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.23076923076923078, |
| "grad_norm": 2.7144410610198975, |
| "learning_rate": 7.500000000000001e-06, |
| "loss": 0.9436, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.2692307692307692, |
| "grad_norm": 4.880862712860107, |
| "learning_rate": 8.750000000000001e-06, |
| "loss": 0.9802, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.3076923076923077, |
| "grad_norm": 5.004383563995361, |
| "learning_rate": 1e-05, |
| "loss": 0.9611, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.34615384615384615, |
| "grad_norm": 4.310351371765137, |
| "learning_rate": 9.994965332706574e-06, |
| "loss": 0.9134, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.38461538461538464, |
| "grad_norm": 4.357025623321533, |
| "learning_rate": 9.979871469976197e-06, |
| "loss": 0.8968, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.4230769230769231, |
| "grad_norm": 3.0148415565490723, |
| "learning_rate": 9.954748808839675e-06, |
| "loss": 0.8748, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.46153846153846156, |
| "grad_norm": 1.8528469800949097, |
| "learning_rate": 9.91964794299315e-06, |
| "loss": 0.8389, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 1.6753920316696167, |
| "learning_rate": 9.874639560909118e-06, |
| "loss": 0.8092, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.5384615384615384, |
| "grad_norm": 1.5386508703231812, |
| "learning_rate": 9.819814303479268e-06, |
| "loss": 0.8134, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.5769230769230769, |
| "grad_norm": 1.145003318786621, |
| "learning_rate": 9.755282581475769e-06, |
| "loss": 0.7763, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.6153846153846154, |
| "grad_norm": 1.0045708417892456, |
| "learning_rate": 9.681174353198687e-06, |
| "loss": 0.7616, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.6538461538461539, |
| "grad_norm": 0.9372178316116333, |
| "learning_rate": 9.597638862757255e-06, |
| "loss": 0.7614, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.6923076923076923, |
| "grad_norm": 0.9609572291374207, |
| "learning_rate": 9.504844339512096e-06, |
| "loss": 0.7543, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.7307692307692307, |
| "grad_norm": 0.7558037638664246, |
| "learning_rate": 9.40297765928369e-06, |
| "loss": 0.7372, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.7692307692307693, |
| "grad_norm": 0.9150065183639526, |
| "learning_rate": 9.292243968009332e-06, |
| "loss": 0.722, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.8076923076923077, |
| "grad_norm": 0.8774355053901672, |
| "learning_rate": 9.172866268606514e-06, |
| "loss": 0.7326, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.8461538461538461, |
| "grad_norm": 1.0381125211715698, |
| "learning_rate": 9.045084971874738e-06, |
| "loss": 0.735, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.8846153846153846, |
| "grad_norm": 0.7722716331481934, |
| "learning_rate": 8.90915741234015e-06, |
| "loss": 0.7408, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.9230769230769231, |
| "grad_norm": 0.8549503684043884, |
| "learning_rate": 8.765357330018056e-06, |
| "loss": 0.7248, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.9615384615384616, |
| "grad_norm": 0.6797922253608704, |
| "learning_rate": 8.613974319136959e-06, |
| "loss": 0.7081, |
| "step": 25 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.7838498950004578, |
| "learning_rate": 8.455313244934324e-06, |
| "loss": 0.7124, |
| "step": 26 |
| }, |
| { |
| "epoch": 1.0384615384615385, |
| "grad_norm": 0.6971613764762878, |
| "learning_rate": 8.289693629698564e-06, |
| "loss": 0.6771, |
| "step": 27 |
| }, |
| { |
| "epoch": 1.0769230769230769, |
| "grad_norm": 0.6067571043968201, |
| "learning_rate": 8.117449009293668e-06, |
| "loss": 0.6713, |
| "step": 28 |
| }, |
| { |
| "epoch": 1.1153846153846154, |
| "grad_norm": 0.6506593227386475, |
| "learning_rate": 7.938926261462366e-06, |
| "loss": 0.6942, |
| "step": 29 |
| }, |
| { |
| "epoch": 1.1538461538461537, |
| "grad_norm": 0.6305216550827026, |
| "learning_rate": 7.754484907260513e-06, |
| "loss": 0.6708, |
| "step": 30 |
| }, |
| { |
| "epoch": 1.1923076923076923, |
| "grad_norm": 0.6441857814788818, |
| "learning_rate": 7.564496387029532e-06, |
| "loss": 0.6584, |
| "step": 31 |
| }, |
| { |
| "epoch": 1.2307692307692308, |
| "grad_norm": 0.6354879140853882, |
| "learning_rate": 7.369343312364994e-06, |
| "loss": 0.6745, |
| "step": 32 |
| }, |
| { |
| "epoch": 1.2692307692307692, |
| "grad_norm": 0.6297513842582703, |
| "learning_rate": 7.169418695587791e-06, |
| "loss": 0.6657, |
| "step": 33 |
| }, |
| { |
| "epoch": 1.3076923076923077, |
| "grad_norm": 0.6114018559455872, |
| "learning_rate": 6.965125158269619e-06, |
| "loss": 0.684, |
| "step": 34 |
| }, |
| { |
| "epoch": 1.3461538461538463, |
| "grad_norm": 0.5940524339675903, |
| "learning_rate": 6.7568741204067145e-06, |
| "loss": 0.6631, |
| "step": 35 |
| }, |
| { |
| "epoch": 1.3846153846153846, |
| "grad_norm": 0.5948757529258728, |
| "learning_rate": 6.545084971874738e-06, |
| "loss": 0.6774, |
| "step": 36 |
| }, |
| { |
| "epoch": 1.4230769230769231, |
| "grad_norm": 0.6056201457977295, |
| "learning_rate": 6.330184227833376e-06, |
| "loss": 0.6803, |
| "step": 37 |
| }, |
| { |
| "epoch": 1.4615384615384617, |
| "grad_norm": 0.5731696486473083, |
| "learning_rate": 6.112604669781572e-06, |
| "loss": 0.6704, |
| "step": 38 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 0.6757548451423645, |
| "learning_rate": 5.892784473993184e-06, |
| "loss": 0.6614, |
| "step": 39 |
| }, |
| { |
| "epoch": 1.5384615384615383, |
| "grad_norm": 0.5654463768005371, |
| "learning_rate": 5.671166329088278e-06, |
| "loss": 0.6617, |
| "step": 40 |
| }, |
| { |
| "epoch": 1.5769230769230769, |
| "grad_norm": 0.47416621446609497, |
| "learning_rate": 5.448196544517168e-06, |
| "loss": 0.6663, |
| "step": 41 |
| }, |
| { |
| "epoch": 1.6153846153846154, |
| "grad_norm": 0.558396577835083, |
| "learning_rate": 5.224324151752575e-06, |
| "loss": 0.667, |
| "step": 42 |
| }, |
| { |
| "epoch": 1.6538461538461537, |
| "grad_norm": 0.6663322448730469, |
| "learning_rate": 5e-06, |
| "loss": 0.6759, |
| "step": 43 |
| }, |
| { |
| "epoch": 1.6923076923076923, |
| "grad_norm": 0.46421608328819275, |
| "learning_rate": 4.775675848247427e-06, |
| "loss": 0.6657, |
| "step": 44 |
| }, |
| { |
| "epoch": 1.7307692307692308, |
| "grad_norm": 0.4642084836959839, |
| "learning_rate": 4.551803455482833e-06, |
| "loss": 0.6686, |
| "step": 45 |
| }, |
| { |
| "epoch": 1.7692307692307692, |
| "grad_norm": 0.618902325630188, |
| "learning_rate": 4.3288336709117246e-06, |
| "loss": 0.6416, |
| "step": 46 |
| }, |
| { |
| "epoch": 1.8076923076923077, |
| "grad_norm": 0.45461708307266235, |
| "learning_rate": 4.107215526006818e-06, |
| "loss": 0.6376, |
| "step": 47 |
| }, |
| { |
| "epoch": 1.8461538461538463, |
| "grad_norm": 0.48321276903152466, |
| "learning_rate": 3.887395330218429e-06, |
| "loss": 0.6592, |
| "step": 48 |
| }, |
| { |
| "epoch": 1.8846153846153846, |
| "grad_norm": 0.4695417881011963, |
| "learning_rate": 3.669815772166625e-06, |
| "loss": 0.6527, |
| "step": 49 |
| }, |
| { |
| "epoch": 1.9230769230769231, |
| "grad_norm": 0.4849114418029785, |
| "learning_rate": 3.4549150281252635e-06, |
| "loss": 0.6449, |
| "step": 50 |
| }, |
| { |
| "epoch": 1.9615384615384617, |
| "grad_norm": 0.4203159511089325, |
| "learning_rate": 3.2431258795932863e-06, |
| "loss": 0.6576, |
| "step": 51 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.39121460914611816, |
| "learning_rate": 3.0348748417303826e-06, |
| "loss": 0.646, |
| "step": 52 |
| }, |
| { |
| "epoch": 2.0384615384615383, |
| "grad_norm": 0.4727398157119751, |
| "learning_rate": 2.83058130441221e-06, |
| "loss": 0.6295, |
| "step": 53 |
| }, |
| { |
| "epoch": 2.076923076923077, |
| "grad_norm": 0.43550893664360046, |
| "learning_rate": 2.6306566876350072e-06, |
| "loss": 0.6265, |
| "step": 54 |
| }, |
| { |
| "epoch": 2.1153846153846154, |
| "grad_norm": 0.37577158212661743, |
| "learning_rate": 2.43550361297047e-06, |
| "loss": 0.6404, |
| "step": 55 |
| }, |
| { |
| "epoch": 2.1538461538461537, |
| "grad_norm": 0.39477619528770447, |
| "learning_rate": 2.245515092739488e-06, |
| "loss": 0.6407, |
| "step": 56 |
| }, |
| { |
| "epoch": 2.1923076923076925, |
| "grad_norm": 0.40658116340637207, |
| "learning_rate": 2.061073738537635e-06, |
| "loss": 0.6267, |
| "step": 57 |
| }, |
| { |
| "epoch": 2.230769230769231, |
| "grad_norm": 0.44563761353492737, |
| "learning_rate": 1.8825509907063328e-06, |
| "loss": 0.62, |
| "step": 58 |
| }, |
| { |
| "epoch": 2.269230769230769, |
| "grad_norm": 0.3756290674209595, |
| "learning_rate": 1.7103063703014372e-06, |
| "loss": 0.623, |
| "step": 59 |
| }, |
| { |
| "epoch": 2.3076923076923075, |
| "grad_norm": 0.3645164668560028, |
| "learning_rate": 1.544686755065677e-06, |
| "loss": 0.622, |
| "step": 60 |
| }, |
| { |
| "epoch": 2.3461538461538463, |
| "grad_norm": 0.4305531680583954, |
| "learning_rate": 1.3860256808630429e-06, |
| "loss": 0.6321, |
| "step": 61 |
| }, |
| { |
| "epoch": 2.3846153846153846, |
| "grad_norm": 0.3671000301837921, |
| "learning_rate": 1.234642669981946e-06, |
| "loss": 0.6312, |
| "step": 62 |
| }, |
| { |
| "epoch": 2.423076923076923, |
| "grad_norm": 0.41619616746902466, |
| "learning_rate": 1.0908425876598512e-06, |
| "loss": 0.6336, |
| "step": 63 |
| }, |
| { |
| "epoch": 2.4615384615384617, |
| "grad_norm": 0.41701027750968933, |
| "learning_rate": 9.549150281252633e-07, |
| "loss": 0.6209, |
| "step": 64 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 0.38584837317466736, |
| "learning_rate": 8.271337313934869e-07, |
| "loss": 0.6337, |
| "step": 65 |
| }, |
| { |
| "epoch": 2.5384615384615383, |
| "grad_norm": 0.3533637821674347, |
| "learning_rate": 7.077560319906696e-07, |
| "loss": 0.6415, |
| "step": 66 |
| }, |
| { |
| "epoch": 2.5769230769230766, |
| "grad_norm": 0.32751405239105225, |
| "learning_rate": 5.9702234071631e-07, |
| "loss": 0.6403, |
| "step": 67 |
| }, |
| { |
| "epoch": 2.6153846153846154, |
| "grad_norm": 0.3373560309410095, |
| "learning_rate": 4.951556604879049e-07, |
| "loss": 0.6354, |
| "step": 68 |
| }, |
| { |
| "epoch": 2.6538461538461537, |
| "grad_norm": 0.3308131992816925, |
| "learning_rate": 4.0236113724274716e-07, |
| "loss": 0.6298, |
| "step": 69 |
| }, |
| { |
| "epoch": 2.6923076923076925, |
| "grad_norm": 0.36556267738342285, |
| "learning_rate": 3.18825646801314e-07, |
| "loss": 0.6287, |
| "step": 70 |
| }, |
| { |
| "epoch": 2.730769230769231, |
| "grad_norm": 0.3372323513031006, |
| "learning_rate": 2.447174185242324e-07, |
| "loss": 0.6282, |
| "step": 71 |
| }, |
| { |
| "epoch": 2.769230769230769, |
| "grad_norm": 0.3662268817424774, |
| "learning_rate": 1.801856965207338e-07, |
| "loss": 0.624, |
| "step": 72 |
| }, |
| { |
| "epoch": 2.8076923076923075, |
| "grad_norm": 0.40386688709259033, |
| "learning_rate": 1.253604390908819e-07, |
| "loss": 0.6156, |
| "step": 73 |
| }, |
| { |
| "epoch": 2.8461538461538463, |
| "grad_norm": 0.33463576436042786, |
| "learning_rate": 8.035205700685167e-08, |
| "loss": 0.6317, |
| "step": 74 |
| }, |
| { |
| "epoch": 2.8846153846153846, |
| "grad_norm": 0.36022576689720154, |
| "learning_rate": 4.52511911603265e-08, |
| "loss": 0.6034, |
| "step": 75 |
| }, |
| { |
| "epoch": 2.9230769230769234, |
| "grad_norm": 0.34557822346687317, |
| "learning_rate": 2.012853002380466e-08, |
| "loss": 0.6266, |
| "step": 76 |
| }, |
| { |
| "epoch": 2.9615384615384617, |
| "grad_norm": 0.3375401794910431, |
| "learning_rate": 5.034667293427053e-09, |
| "loss": 0.6245, |
| "step": 77 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 0.3406679630279541, |
| "learning_rate": 0.0, |
| "loss": 0.6309, |
| "step": 78 |
| }, |
| { |
| "epoch": 3.0, |
| "step": 78, |
| "total_flos": 1.1805367575825613e+17, |
| "train_loss": 0.713857365724368, |
| "train_runtime": 2041.9553, |
| "train_samples_per_second": 3.655, |
| "train_steps_per_second": 0.038 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 78, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.1805367575825613e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
