{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6076154469349176, "eval_steps": 500, "global_step": 49500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00613752976701937, "grad_norm": 0.7345873117446899, "learning_rate": 4.9897707837216344e-05, "loss": 1.0473, "step": 500 }, { "epoch": 0.01227505953403874, "grad_norm": 1.5779420137405396, "learning_rate": 4.9795415674432687e-05, "loss": 1.0447, "step": 1000 }, { "epoch": 0.01841258930105811, "grad_norm": 1.1132322549819946, "learning_rate": 4.969312351164903e-05, "loss": 1.0383, "step": 1500 }, { "epoch": 0.02455011906807748, "grad_norm": 1.4636728763580322, "learning_rate": 4.959083134886538e-05, "loss": 1.0395, "step": 2000 }, { "epoch": 0.03068764883509685, "grad_norm": 2.259823799133301, "learning_rate": 4.948853918608172e-05, "loss": 1.0436, "step": 2500 }, { "epoch": 0.03682517860211622, "grad_norm": 3.276669979095459, "learning_rate": 4.938624702329807e-05, "loss": 1.0375, "step": 3000 }, { "epoch": 0.04296270836913559, "grad_norm": 2.432560443878174, "learning_rate": 4.928395486051441e-05, "loss": 1.0189, "step": 3500 }, { "epoch": 0.04910023813615496, "grad_norm": 2.106665849685669, "learning_rate": 4.918166269773075e-05, "loss": 1.0282, "step": 4000 }, { "epoch": 0.05523776790317433, "grad_norm": 10.379436492919922, "learning_rate": 4.9079370534947094e-05, "loss": 1.0355, "step": 4500 }, { "epoch": 0.0613752976701937, "grad_norm": 1.7805677652359009, "learning_rate": 4.897707837216344e-05, "loss": 1.0286, "step": 5000 }, { "epoch": 0.06751282743721307, "grad_norm": 4.699483394622803, "learning_rate": 4.8874786209379785e-05, "loss": 1.0264, "step": 5500 }, { "epoch": 0.07365035720423244, "grad_norm": 3.2066714763641357, "learning_rate": 4.877249404659613e-05, "loss": 1.0174, "step": 6000 }, { "epoch": 0.07978788697125182, "grad_norm": 4.304272651672363, "learning_rate": 4.867020188381247e-05, "loss": 1.0238, "step": 6500 }, { "epoch": 0.08592541673827118, "grad_norm": 1.9387307167053223, "learning_rate": 4.856790972102881e-05, "loss": 1.0217, "step": 7000 }, { "epoch": 0.09206294650529055, "grad_norm": 7.0195488929748535, "learning_rate": 4.846561755824516e-05, "loss": 1.0203, "step": 7500 }, { "epoch": 0.09820047627230992, "grad_norm": 5.841785907745361, "learning_rate": 4.836332539546151e-05, "loss": 1.0201, "step": 8000 }, { "epoch": 0.10433800603932929, "grad_norm": 4.427718162536621, "learning_rate": 4.826103323267785e-05, "loss": 1.0266, "step": 8500 }, { "epoch": 0.11047553580634865, "grad_norm": 2.067218780517578, "learning_rate": 4.815874106989419e-05, "loss": 1.0379, "step": 9000 }, { "epoch": 0.11661306557336804, "grad_norm": 4.303891181945801, "learning_rate": 4.8056448907110535e-05, "loss": 1.0367, "step": 9500 }, { "epoch": 0.1227505953403874, "grad_norm": 5.363577842712402, "learning_rate": 4.795415674432688e-05, "loss": 1.0384, "step": 10000 }, { "epoch": 0.12888812510740677, "grad_norm": 7.8037028312683105, "learning_rate": 4.785186458154322e-05, "loss": 1.0291, "step": 10500 }, { "epoch": 0.13502565487442614, "grad_norm": 2.8838307857513428, "learning_rate": 4.774957241875956e-05, "loss": 1.0362, "step": 11000 }, { "epoch": 0.1411631846414455, "grad_norm": 18.838226318359375, "learning_rate": 4.764728025597591e-05, "loss": 1.0354, "step": 11500 }, { "epoch": 0.14730071440846487, "grad_norm": 1.8900924921035767, "learning_rate": 4.754498809319225e-05, "loss": 1.0204, "step": 12000 }, { "epoch": 0.15343824417548424, "grad_norm": 8.636752128601074, "learning_rate": 4.74426959304086e-05, "loss": 1.0401, "step": 12500 }, { "epoch": 0.15957577394250363, "grad_norm": 3.1361780166625977, "learning_rate": 4.734040376762494e-05, "loss": 1.0237, "step": 13000 }, { "epoch": 0.165713303709523, "grad_norm": 6.990635871887207, "learning_rate": 4.7238111604841285e-05, "loss": 1.0271, "step": 13500 }, { "epoch": 0.17185083347654237, "grad_norm": 2.6320559978485107, "learning_rate": 4.713581944205763e-05, "loss": 1.0235, "step": 14000 }, { "epoch": 0.17798836324356174, "grad_norm": 5.751638889312744, "learning_rate": 4.7033527279273976e-05, "loss": 1.0329, "step": 14500 }, { "epoch": 0.1841258930105811, "grad_norm": 10.333562850952148, "learning_rate": 4.693123511649032e-05, "loss": 1.0474, "step": 15000 }, { "epoch": 0.19026342277760047, "grad_norm": 9.57418155670166, "learning_rate": 4.682894295370666e-05, "loss": 1.0547, "step": 15500 }, { "epoch": 0.19640095254461984, "grad_norm": 6.002443313598633, "learning_rate": 4.6726650790923e-05, "loss": 1.0557, "step": 16000 }, { "epoch": 0.2025384823116392, "grad_norm": 17.77951431274414, "learning_rate": 4.6624358628139344e-05, "loss": 1.0626, "step": 16500 }, { "epoch": 0.20867601207865857, "grad_norm": 10.126110076904297, "learning_rate": 4.652206646535569e-05, "loss": 1.0551, "step": 17000 }, { "epoch": 0.21481354184567794, "grad_norm": 3.7376019954681396, "learning_rate": 4.641977430257204e-05, "loss": 1.0526, "step": 17500 }, { "epoch": 0.2209510716126973, "grad_norm": 19.0529842376709, "learning_rate": 4.6317482139788384e-05, "loss": 1.125, "step": 18000 }, { "epoch": 0.2270886013797167, "grad_norm": 1.9759446382522583, "learning_rate": 4.6215189977004726e-05, "loss": 1.2996, "step": 18500 }, { "epoch": 0.23322613114673607, "grad_norm": 2.7964680194854736, "learning_rate": 4.611289781422107e-05, "loss": 1.0703, "step": 19000 }, { "epoch": 0.23936366091375544, "grad_norm": 18.77906036376953, "learning_rate": 4.601060565143741e-05, "loss": 1.0601, "step": 19500 }, { "epoch": 0.2455011906807748, "grad_norm": 3.129147529602051, "learning_rate": 4.590831348865375e-05, "loss": 1.0835, "step": 20000 }, { "epoch": 0.2516387204477942, "grad_norm": 3.018688917160034, "learning_rate": 4.58060213258701e-05, "loss": 1.0583, "step": 20500 }, { "epoch": 0.25777625021481354, "grad_norm": 32.16483688354492, "learning_rate": 4.570372916308644e-05, "loss": 1.141, "step": 21000 }, { "epoch": 0.2639137799818329, "grad_norm": 8.431964874267578, "learning_rate": 4.5601437000302785e-05, "loss": 1.0802, "step": 21500 }, { "epoch": 0.2700513097488523, "grad_norm": 14.990534782409668, "learning_rate": 4.5499144837519134e-05, "loss": 1.0589, "step": 22000 }, { "epoch": 0.27618883951587164, "grad_norm": 6.134013652801514, "learning_rate": 4.5396852674735476e-05, "loss": 1.086, "step": 22500 }, { "epoch": 0.282326369282891, "grad_norm": 15.587114334106445, "learning_rate": 4.529456051195182e-05, "loss": 1.0743, "step": 23000 }, { "epoch": 0.2884638990499104, "grad_norm": 20.36590003967285, "learning_rate": 4.519226834916816e-05, "loss": 1.0605, "step": 23500 }, { "epoch": 0.29460142881692974, "grad_norm": 6.840169906616211, "learning_rate": 4.508997618638451e-05, "loss": 1.0602, "step": 24000 }, { "epoch": 0.3007389585839491, "grad_norm": 22.10498046875, "learning_rate": 4.498768402360085e-05, "loss": 1.0732, "step": 24500 }, { "epoch": 0.3068764883509685, "grad_norm": 2.3044190406799316, "learning_rate": 4.488539186081719e-05, "loss": 1.0614, "step": 25000 }, { "epoch": 0.31301401811798785, "grad_norm": 19.358158111572266, "learning_rate": 4.4783099698033535e-05, "loss": 1.0666, "step": 25500 }, { "epoch": 0.31915154788500727, "grad_norm": 47.77769088745117, "learning_rate": 4.468080753524988e-05, "loss": 1.0669, "step": 26000 }, { "epoch": 0.32528907765202664, "grad_norm": 31.26322364807129, "learning_rate": 4.4578515372466226e-05, "loss": 1.0627, "step": 26500 }, { "epoch": 0.331426607419046, "grad_norm": 31.237651824951172, "learning_rate": 4.4476223209682575e-05, "loss": 1.0659, "step": 27000 }, { "epoch": 0.33756413718606537, "grad_norm": 10.651726722717285, "learning_rate": 4.437393104689892e-05, "loss": 1.07, "step": 27500 }, { "epoch": 0.34370166695308474, "grad_norm": 10.767593383789062, "learning_rate": 4.427163888411526e-05, "loss": 1.0783, "step": 28000 }, { "epoch": 0.3498391967201041, "grad_norm": 13.177048683166504, "learning_rate": 4.41693467213316e-05, "loss": 1.078, "step": 28500 }, { "epoch": 0.3559767264871235, "grad_norm": 41.34538269042969, "learning_rate": 4.406705455854794e-05, "loss": 1.0832, "step": 29000 }, { "epoch": 0.36211425625414284, "grad_norm": 2.813678026199341, "learning_rate": 4.3964762395764285e-05, "loss": 1.0695, "step": 29500 }, { "epoch": 0.3682517860211622, "grad_norm": 32.63833236694336, "learning_rate": 4.3862470232980634e-05, "loss": 1.0714, "step": 30000 }, { "epoch": 0.3743893157881816, "grad_norm": 6.144928932189941, "learning_rate": 4.3760178070196976e-05, "loss": 1.3794, "step": 30500 }, { "epoch": 0.38052684555520094, "grad_norm": 21.691205978393555, "learning_rate": 4.365788590741332e-05, "loss": 1.0838, "step": 31000 }, { "epoch": 0.3866643753222203, "grad_norm": 5.830935955047607, "learning_rate": 4.355559374462967e-05, "loss": 1.0541, "step": 31500 }, { "epoch": 0.3928019050892397, "grad_norm": 5.59116792678833, "learning_rate": 4.345330158184601e-05, "loss": 1.0799, "step": 32000 }, { "epoch": 0.39893943485625905, "grad_norm": 5.144099712371826, "learning_rate": 4.335100941906235e-05, "loss": 1.0622, "step": 32500 }, { "epoch": 0.4050769646232784, "grad_norm": 30.603145599365234, "learning_rate": 4.324871725627869e-05, "loss": 1.0597, "step": 33000 }, { "epoch": 0.4112144943902978, "grad_norm": 14.602336883544922, "learning_rate": 4.314642509349504e-05, "loss": 1.0731, "step": 33500 }, { "epoch": 0.41735202415731715, "grad_norm": 11.165698051452637, "learning_rate": 4.3044132930711384e-05, "loss": 1.0871, "step": 34000 }, { "epoch": 0.4234895539243365, "grad_norm": 7.897899627685547, "learning_rate": 4.2941840767927726e-05, "loss": 1.0535, "step": 34500 }, { "epoch": 0.4296270836913559, "grad_norm": 34.061134338378906, "learning_rate": 4.283954860514407e-05, "loss": 1.3215, "step": 35000 }, { "epoch": 0.43576461345837525, "grad_norm": 5.050889015197754, "learning_rate": 4.273725644236041e-05, "loss": 1.1501, "step": 35500 }, { "epoch": 0.4419021432253946, "grad_norm": 11.23651123046875, "learning_rate": 4.263496427957676e-05, "loss": 1.1088, "step": 36000 }, { "epoch": 0.44803967299241404, "grad_norm": 11.963942527770996, "learning_rate": 4.25326721167931e-05, "loss": 1.0844, "step": 36500 }, { "epoch": 0.4541772027594334, "grad_norm": 19.694747924804688, "learning_rate": 4.243037995400945e-05, "loss": 1.0981, "step": 37000 }, { "epoch": 0.4603147325264528, "grad_norm": 32.396446228027344, "learning_rate": 4.232808779122579e-05, "loss": 1.0856, "step": 37500 }, { "epoch": 0.46645226229347214, "grad_norm": 30.794092178344727, "learning_rate": 4.2225795628442134e-05, "loss": 1.0956, "step": 38000 }, { "epoch": 0.4725897920604915, "grad_norm": 25.20745086669922, "learning_rate": 4.2123503465658476e-05, "loss": 1.0813, "step": 38500 }, { "epoch": 0.4787273218275109, "grad_norm": 13.46246337890625, "learning_rate": 4.202121130287482e-05, "loss": 1.0843, "step": 39000 }, { "epoch": 0.48486485159453024, "grad_norm": 9.27540397644043, "learning_rate": 4.191891914009117e-05, "loss": 1.1053, "step": 39500 }, { "epoch": 0.4910023813615496, "grad_norm": 3.5075111389160156, "learning_rate": 4.181662697730751e-05, "loss": 1.092, "step": 40000 }, { "epoch": 0.497139911128569, "grad_norm": 6.219689846038818, "learning_rate": 4.171433481452385e-05, "loss": 1.081, "step": 40500 }, { "epoch": 0.5032774408955883, "grad_norm": 16.255443572998047, "learning_rate": 4.161204265174019e-05, "loss": 1.0919, "step": 41000 }, { "epoch": 0.5094149706626078, "grad_norm": 32.27680206298828, "learning_rate": 4.150975048895654e-05, "loss": 1.0681, "step": 41500 }, { "epoch": 0.5155525004296271, "grad_norm": 107.78765869140625, "learning_rate": 4.1407458326172884e-05, "loss": 1.0826, "step": 42000 }, { "epoch": 0.5216900301966465, "grad_norm": 8.029512405395508, "learning_rate": 4.1305166163389226e-05, "loss": 1.0809, "step": 42500 }, { "epoch": 0.5278275599636658, "grad_norm": 44.91604232788086, "learning_rate": 4.1202874000605575e-05, "loss": 1.093, "step": 43000 }, { "epoch": 0.5339650897306852, "grad_norm": 11.860628128051758, "learning_rate": 4.110058183782192e-05, "loss": 1.0709, "step": 43500 }, { "epoch": 0.5401026194977046, "grad_norm": 30.72344398498535, "learning_rate": 4.099828967503826e-05, "loss": 1.0623, "step": 44000 }, { "epoch": 0.546240149264724, "grad_norm": 19.160911560058594, "learning_rate": 4.08959975122546e-05, "loss": 1.0801, "step": 44500 }, { "epoch": 0.5523776790317433, "grad_norm": 12.648110389709473, "learning_rate": 4.079370534947094e-05, "loss": 1.0833, "step": 45000 }, { "epoch": 0.5585152087987627, "grad_norm": 25.488170623779297, "learning_rate": 4.0691413186687285e-05, "loss": 1.0816, "step": 45500 }, { "epoch": 0.564652738565782, "grad_norm": 8.982975006103516, "learning_rate": 4.0589121023903634e-05, "loss": 1.0697, "step": 46000 }, { "epoch": 0.5707902683328014, "grad_norm": 12.776642799377441, "learning_rate": 4.048682886111998e-05, "loss": 1.0864, "step": 46500 }, { "epoch": 0.5769277980998208, "grad_norm": 70.14471435546875, "learning_rate": 4.0384536698336325e-05, "loss": 1.0744, "step": 47000 }, { "epoch": 0.5830653278668402, "grad_norm": 26.72732162475586, "learning_rate": 4.028224453555267e-05, "loss": 1.0897, "step": 47500 }, { "epoch": 0.5892028576338595, "grad_norm": 22.69961166381836, "learning_rate": 4.017995237276901e-05, "loss": 1.0727, "step": 48000 }, { "epoch": 0.5953403874008789, "grad_norm": 6.76914119720459, "learning_rate": 4.007766020998535e-05, "loss": 1.0621, "step": 48500 }, { "epoch": 0.6014779171678982, "grad_norm": 41.32302474975586, "learning_rate": 3.99753680472017e-05, "loss": 1.0582, "step": 49000 }, { "epoch": 0.6076154469349176, "grad_norm": 47.626644134521484, "learning_rate": 3.987307588441804e-05, "loss": 1.0681, "step": 49500 } ], "logging_steps": 500, "max_steps": 244398, "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }