| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.37677304964539005, | |
| "eval_steps": 500, | |
| "global_step": 4250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004432624113475178, | |
| "grad_norm": 0.009513450788341307, | |
| "learning_rate": 5.066502241635949e-06, | |
| "loss": 0.0353, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008865248226950355, | |
| "grad_norm": 0.12959164414811705, | |
| "learning_rate": 5.964204463534375e-06, | |
| "loss": 0.0209, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.013297872340425532, | |
| "grad_norm": 0.03079899259156006, | |
| "learning_rate": 6.489326600159014e-06, | |
| "loss": 0.0114, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01773049645390071, | |
| "grad_norm": 1.904334998115993, | |
| "learning_rate": 6.855414889666249e-06, | |
| "loss": 0.0426, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.022163120567375887, | |
| "grad_norm": 0.00029657300723797526, | |
| "learning_rate": 7.14571142073327e-06, | |
| "loss": 0.0092, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.026595744680851064, | |
| "grad_norm": 0.5635451567630495, | |
| "learning_rate": 7.382704575824005e-06, | |
| "loss": 0.0207, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03102836879432624, | |
| "grad_norm": 17.174480153841266, | |
| "learning_rate": 7.582965380729319e-06, | |
| "loss": 0.0128, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.03546099290780142, | |
| "grad_norm": 0.0029856714585610065, | |
| "learning_rate": 7.756367076993413e-06, | |
| "loss": 0.0087, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0398936170212766, | |
| "grad_norm": 46.59324697917607, | |
| "learning_rate": 7.909269732702365e-06, | |
| "loss": 0.0358, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.044326241134751775, | |
| "grad_norm": 0.011038539916815544, | |
| "learning_rate": 8.046011658635775e-06, | |
| "loss": 0.0099, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04875886524822695, | |
| "grad_norm": 0.06122707900561642, | |
| "learning_rate": 8.169684802472282e-06, | |
| "loss": 0.0257, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.05319148936170213, | |
| "grad_norm": 11.17528630173658, | |
| "learning_rate": 8.28257072561721e-06, | |
| "loss": 0.0399, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.057624113475177305, | |
| "grad_norm": 0.004434274436946811, | |
| "learning_rate": 8.386401186603816e-06, | |
| "loss": 0.0106, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.06205673758865248, | |
| "grad_norm": 0.00022486652288303773, | |
| "learning_rate": 8.482521734002021e-06, | |
| "loss": 0.0175, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.06648936170212766, | |
| "grad_norm": 0.08252216933057055, | |
| "learning_rate": 8.571998643735778e-06, | |
| "loss": 0.0122, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.07092198581560284, | |
| "grad_norm": 38.71346721678791, | |
| "learning_rate": 8.655691228400912e-06, | |
| "loss": 0.0107, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.07535460992907801, | |
| "grad_norm": 0.016399532174579123, | |
| "learning_rate": 8.734302159716567e-06, | |
| "loss": 0.0272, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.0797872340425532, | |
| "grad_norm": 13.96210364288869, | |
| "learning_rate": 8.808413368821617e-06, | |
| "loss": 0.0162, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.08421985815602837, | |
| "grad_norm": 20.616657581291367, | |
| "learning_rate": 8.878512216644468e-06, | |
| "loss": 0.018, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.08865248226950355, | |
| "grad_norm": 8.554096191036654, | |
| "learning_rate": 8.945010936770532e-06, | |
| "loss": 0.0122, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09308510638297872, | |
| "grad_norm": 9.002995310481221, | |
| "learning_rate": 9.008261325009621e-06, | |
| "loss": 0.004, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.0975177304964539, | |
| "grad_norm": 0.04332552525052799, | |
| "learning_rate": 9.068566005374956e-06, | |
| "loss": 0.0049, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.10195035460992907, | |
| "grad_norm": 0.566818534495066, | |
| "learning_rate": 9.126187187473155e-06, | |
| "loss": 0.0348, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.10638297872340426, | |
| "grad_norm": 2.2102759352121244, | |
| "learning_rate": 9.181353557126703e-06, | |
| "loss": 0.0157, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.11081560283687943, | |
| "grad_norm": 0.0007883033335850071, | |
| "learning_rate": 9.233228424991629e-06, | |
| "loss": 0.0153, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.11524822695035461, | |
| "grad_norm": 8.962657174789879, | |
| "learning_rate": 9.28410340850224e-06, | |
| "loss": 0.011, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.1196808510638298, | |
| "grad_norm": 0.08478258585114734, | |
| "learning_rate": 9.333055212266742e-06, | |
| "loss": 0.0193, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.12411347517730496, | |
| "grad_norm": 5.126269299634291, | |
| "learning_rate": 9.380223955900445e-06, | |
| "loss": 0.0224, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.12854609929078015, | |
| "grad_norm": 0.001630326745394445, | |
| "learning_rate": 9.425734982425631e-06, | |
| "loss": 0.0198, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.13297872340425532, | |
| "grad_norm": 31.3251284414765, | |
| "learning_rate": 9.469700865634203e-06, | |
| "loss": 0.0201, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1374113475177305, | |
| "grad_norm": 0.003720201610220148, | |
| "learning_rate": 9.512223087748166e-06, | |
| "loss": 0.0386, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.14184397163120568, | |
| "grad_norm": 1.4908536371789232, | |
| "learning_rate": 9.553393450299339e-06, | |
| "loss": 0.0155, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.14627659574468085, | |
| "grad_norm": 0.18557111682138577, | |
| "learning_rate": 9.593295267571064e-06, | |
| "loss": 0.0132, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.15070921985815602, | |
| "grad_norm": 0.0012351164899243579, | |
| "learning_rate": 9.632004381614992e-06, | |
| "loss": 0.0169, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1551418439716312, | |
| "grad_norm": 4.390950581164127, | |
| "learning_rate": 9.669590029926173e-06, | |
| "loss": 0.0188, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.1595744680851064, | |
| "grad_norm": 1.5823201233659507, | |
| "learning_rate": 9.706115590720043e-06, | |
| "loss": 0.0176, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.16400709219858156, | |
| "grad_norm": 0.6041784257950227, | |
| "learning_rate": 9.741639225963296e-06, | |
| "loss": 0.0076, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.16843971631205673, | |
| "grad_norm": 8.096633927567137, | |
| "learning_rate": 9.776214438542894e-06, | |
| "loss": 0.0254, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.17287234042553193, | |
| "grad_norm": 0.001381384244099819, | |
| "learning_rate": 9.80989055697396e-06, | |
| "loss": 0.021, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.1773049645390071, | |
| "grad_norm": 0.6173202999510908, | |
| "learning_rate": 9.842713158668956e-06, | |
| "loss": 0.0153, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.18173758865248227, | |
| "grad_norm": 0.4702304667158159, | |
| "learning_rate": 9.874724440882673e-06, | |
| "loss": 0.0322, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.18617021276595744, | |
| "grad_norm": 0.007418267836546937, | |
| "learning_rate": 9.905963546908046e-06, | |
| "loss": 0.0189, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.19060283687943264, | |
| "grad_norm": 0.040611583128721894, | |
| "learning_rate": 9.936466853848526e-06, | |
| "loss": 0.004, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.1950354609929078, | |
| "grad_norm": 35.80087991837194, | |
| "learning_rate": 9.96626822727338e-06, | |
| "loss": 0.022, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.19946808510638298, | |
| "grad_norm": 0.44489231698014525, | |
| "learning_rate": 9.995399247226628e-06, | |
| "loss": 0.0165, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.20390070921985815, | |
| "grad_norm": 0.0002357955980587179, | |
| "learning_rate": 9.980299448384557e-06, | |
| "loss": 0.0091, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 0.011156109382490509, | |
| "learning_rate": 9.955673758865249e-06, | |
| "loss": 0.0293, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 0.010237311021590336, | |
| "learning_rate": 9.931048069345942e-06, | |
| "loss": 0.0267, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.21719858156028368, | |
| "grad_norm": 0.1907181340319684, | |
| "learning_rate": 9.906422379826635e-06, | |
| "loss": 0.0035, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.22163120567375885, | |
| "grad_norm": 7.244482111036572, | |
| "learning_rate": 9.881796690307329e-06, | |
| "loss": 0.0129, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.22606382978723405, | |
| "grad_norm": 2.614333142067009, | |
| "learning_rate": 9.857171000788024e-06, | |
| "loss": 0.0176, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.23049645390070922, | |
| "grad_norm": 0.19026758190049964, | |
| "learning_rate": 9.832545311268717e-06, | |
| "loss": 0.0118, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.2349290780141844, | |
| "grad_norm": 1.926354870277716, | |
| "learning_rate": 9.80791962174941e-06, | |
| "loss": 0.0113, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.2393617021276596, | |
| "grad_norm": 1.3390029353713434e-05, | |
| "learning_rate": 9.783293932230104e-06, | |
| "loss": 0.0095, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.24379432624113476, | |
| "grad_norm": 0.25506782394102356, | |
| "learning_rate": 9.758668242710797e-06, | |
| "loss": 0.0235, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.24822695035460993, | |
| "grad_norm": 0.0011667263533171556, | |
| "learning_rate": 9.73404255319149e-06, | |
| "loss": 0.01, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.2526595744680851, | |
| "grad_norm": 2.676783820787547, | |
| "learning_rate": 9.709416863672184e-06, | |
| "loss": 0.0305, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2570921985815603, | |
| "grad_norm": 0.0019099837220493063, | |
| "learning_rate": 9.684791174152877e-06, | |
| "loss": 0.0246, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.26152482269503546, | |
| "grad_norm": 0.6898702380926142, | |
| "learning_rate": 9.66016548463357e-06, | |
| "loss": 0.0211, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.26595744680851063, | |
| "grad_norm": 0.0006438124789484562, | |
| "learning_rate": 9.635539795114264e-06, | |
| "loss": 0.008, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.2703900709219858, | |
| "grad_norm": 3.819700570925564e-05, | |
| "learning_rate": 9.610914105594957e-06, | |
| "loss": 0.0199, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.274822695035461, | |
| "grad_norm": 0.019487722143147885, | |
| "learning_rate": 9.58628841607565e-06, | |
| "loss": 0.0126, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.27925531914893614, | |
| "grad_norm": 0.49075883570668455, | |
| "learning_rate": 9.561662726556344e-06, | |
| "loss": 0.0114, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.28368794326241137, | |
| "grad_norm": 0.661703227107662, | |
| "learning_rate": 9.537037037037037e-06, | |
| "loss": 0.0081, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.28812056737588654, | |
| "grad_norm": 0.00012406562558187035, | |
| "learning_rate": 9.512411347517732e-06, | |
| "loss": 0.0154, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.2925531914893617, | |
| "grad_norm": 10.456271349514031, | |
| "learning_rate": 9.487785657998426e-06, | |
| "loss": 0.0056, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.2969858156028369, | |
| "grad_norm": 34.88954914653437, | |
| "learning_rate": 9.463159968479119e-06, | |
| "loss": 0.0215, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.30141843971631205, | |
| "grad_norm": 0.0037401951660654617, | |
| "learning_rate": 9.43853427895981e-06, | |
| "loss": 0.0073, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3058510638297872, | |
| "grad_norm": 1.9770145427070573, | |
| "learning_rate": 9.413908589440504e-06, | |
| "loss": 0.026, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.3102836879432624, | |
| "grad_norm": 0.0005171843885202088, | |
| "learning_rate": 9.389282899921197e-06, | |
| "loss": 0.0054, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3147163120567376, | |
| "grad_norm": 6.671945039975605, | |
| "learning_rate": 9.364657210401892e-06, | |
| "loss": 0.0193, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.3191489361702128, | |
| "grad_norm": 24.199369765973504, | |
| "learning_rate": 9.340031520882586e-06, | |
| "loss": 0.0253, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.32358156028368795, | |
| "grad_norm": 0.03237450123162466, | |
| "learning_rate": 9.31540583136328e-06, | |
| "loss": 0.0108, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.3280141843971631, | |
| "grad_norm": 0.06582091922650166, | |
| "learning_rate": 9.290780141843973e-06, | |
| "loss": 0.0052, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.3324468085106383, | |
| "grad_norm": 1.1046639468259907, | |
| "learning_rate": 9.266154452324666e-06, | |
| "loss": 0.0077, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.33687943262411346, | |
| "grad_norm": 0.12685978465854558, | |
| "learning_rate": 9.24152876280536e-06, | |
| "loss": 0.0055, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.34131205673758863, | |
| "grad_norm": 0.001176875548815465, | |
| "learning_rate": 9.216903073286053e-06, | |
| "loss": 0.0116, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.34574468085106386, | |
| "grad_norm": 19.93923722594081, | |
| "learning_rate": 9.192277383766746e-06, | |
| "loss": 0.0339, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.350177304964539, | |
| "grad_norm": 0.5136555426144171, | |
| "learning_rate": 9.16765169424744e-06, | |
| "loss": 0.0109, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.3546099290780142, | |
| "grad_norm": 29.577148937276096, | |
| "learning_rate": 9.143026004728133e-06, | |
| "loss": 0.0195, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.35904255319148937, | |
| "grad_norm": 5.471652062023114, | |
| "learning_rate": 9.118400315208826e-06, | |
| "loss": 0.0075, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.36347517730496454, | |
| "grad_norm": 0.0039179745195827426, | |
| "learning_rate": 9.09377462568952e-06, | |
| "loss": 0.0203, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.3679078014184397, | |
| "grad_norm": 30.214203138409303, | |
| "learning_rate": 9.069148936170213e-06, | |
| "loss": 0.0323, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.3723404255319149, | |
| "grad_norm": 1.0344365911211837, | |
| "learning_rate": 9.044523246650908e-06, | |
| "loss": 0.0252, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.37677304964539005, | |
| "grad_norm": 0.6080095692433664, | |
| "learning_rate": 9.019897557131601e-06, | |
| "loss": 0.0065, | |
| "step": 4250 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 22560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 4250, | |
| "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": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |