| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 75, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.04145077720207254, |
| "grad_norm": 8.21605396270752, |
| "learning_rate": 0.0, |
| "loss": 0.6757508516311646, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.08290155440414508, |
| "grad_norm": 8.345704078674316, |
| "learning_rate": 1.25e-06, |
| "loss": 0.6867995858192444, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.12435233160621761, |
| "grad_norm": 8.243107795715332, |
| "learning_rate": 2.5e-06, |
| "loss": 0.7058541774749756, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.16580310880829016, |
| "grad_norm": 7.024388313293457, |
| "learning_rate": 3.7500000000000005e-06, |
| "loss": 0.6871891021728516, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.20725388601036268, |
| "grad_norm": 5.133410453796387, |
| "learning_rate": 5e-06, |
| "loss": 0.616441011428833, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.24870466321243523, |
| "grad_norm": 1.9848037958145142, |
| "learning_rate": 6.25e-06, |
| "loss": 0.5559022426605225, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.29015544041450775, |
| "grad_norm": 3.340723752975464, |
| "learning_rate": 7.500000000000001e-06, |
| "loss": 0.597740650177002, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.3316062176165803, |
| "grad_norm": 3.22680401802063, |
| "learning_rate": 8.750000000000001e-06, |
| "loss": 0.5974273085594177, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.37305699481865284, |
| "grad_norm": 2.873014211654663, |
| "learning_rate": 1e-05, |
| "loss": 0.5814194083213806, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.41450777202072536, |
| "grad_norm": 2.6093578338623047, |
| "learning_rate": 9.994504457428557e-06, |
| "loss": 0.5662238597869873, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.45595854922279794, |
| "grad_norm": 2.017469882965088, |
| "learning_rate": 9.978029910109491e-06, |
| "loss": 0.5501809120178223, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.49740932642487046, |
| "grad_norm": 1.4958215951919556, |
| "learning_rate": 9.950612572673255e-06, |
| "loss": 0.5414538383483887, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.538860103626943, |
| "grad_norm": 1.6731749773025513, |
| "learning_rate": 9.91231271437788e-06, |
| "loss": 0.5331971645355225, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.5803108808290155, |
| "grad_norm": 1.8399823904037476, |
| "learning_rate": 9.863214526624065e-06, |
| "loss": 0.5391272902488708, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.6217616580310881, |
| "grad_norm": 1.7738268375396729, |
| "learning_rate": 9.803425937884202e-06, |
| "loss": 0.547919750213623, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.6632124352331606, |
| "grad_norm": 1.5366426706314087, |
| "learning_rate": 9.733078376452172e-06, |
| "loss": 0.5016499161720276, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.7046632124352331, |
| "grad_norm": 1.3754934072494507, |
| "learning_rate": 9.652326481535434e-06, |
| "loss": 0.5027351379394531, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.7461139896373057, |
| "grad_norm": 1.5104612112045288, |
| "learning_rate": 9.561347763324484e-06, |
| "loss": 0.5046743154525757, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.7875647668393783, |
| "grad_norm": 1.3040716648101807, |
| "learning_rate": 9.460342212786933e-06, |
| "loss": 0.4925868511199951, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.8290155440414507, |
| "grad_norm": 1.18463933467865, |
| "learning_rate": 9.349531862043952e-06, |
| "loss": 0.49534356594085693, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.8704663212435233, |
| "grad_norm": 1.2657711505889893, |
| "learning_rate": 9.229160296295488e-06, |
| "loss": 0.46897661685943604, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.9119170984455959, |
| "grad_norm": 1.3168799877166748, |
| "learning_rate": 9.099492118367123e-06, |
| "loss": 0.5024734735488892, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.9533678756476683, |
| "grad_norm": 1.2915098667144775, |
| "learning_rate": 8.960812367055646e-06, |
| "loss": 0.487984299659729, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.9948186528497409, |
| "grad_norm": 1.2028911113739014, |
| "learning_rate": 8.81342589055191e-06, |
| "loss": 0.4934556186199188, |
| "step": 24 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 2.874122381210327, |
| "learning_rate": 8.657656676318346e-06, |
| "loss": 0.4702609181404114, |
| "step": 25 |
| }, |
| { |
| "epoch": 1.0414507772020725, |
| "grad_norm": 1.3279752731323242, |
| "learning_rate": 8.49384713889421e-06, |
| "loss": 0.4079495072364807, |
| "step": 26 |
| }, |
| { |
| "epoch": 1.0829015544041452, |
| "grad_norm": 1.2503715753555298, |
| "learning_rate": 8.32235736719411e-06, |
| "loss": 0.4119013249874115, |
| "step": 27 |
| }, |
| { |
| "epoch": 1.1243523316062176, |
| "grad_norm": 1.137712836265564, |
| "learning_rate": 8.143564332954426e-06, |
| "loss": 0.41211771965026855, |
| "step": 28 |
| }, |
| { |
| "epoch": 1.16580310880829, |
| "grad_norm": 1.2087459564208984, |
| "learning_rate": 7.957861062067614e-06, |
| "loss": 0.394731342792511, |
| "step": 29 |
| }, |
| { |
| "epoch": 1.2072538860103628, |
| "grad_norm": 1.270243763923645, |
| "learning_rate": 7.765655770625997e-06, |
| "loss": 0.38476723432540894, |
| "step": 30 |
| }, |
| { |
| "epoch": 1.2487046632124352, |
| "grad_norm": 1.2289609909057617, |
| "learning_rate": 7.56737096757421e-06, |
| "loss": 0.40923917293548584, |
| "step": 31 |
| }, |
| { |
| "epoch": 1.2901554404145077, |
| "grad_norm": 1.1888467073440552, |
| "learning_rate": 7.363442525942827e-06, |
| "loss": 0.39686059951782227, |
| "step": 32 |
| }, |
| { |
| "epoch": 1.3316062176165804, |
| "grad_norm": 1.3771826028823853, |
| "learning_rate": 7.1543187247048525e-06, |
| "loss": 0.37492647767066956, |
| "step": 33 |
| }, |
| { |
| "epoch": 1.3730569948186528, |
| "grad_norm": 1.2060784101486206, |
| "learning_rate": 6.9404592633612486e-06, |
| "loss": 0.4068436920642853, |
| "step": 34 |
| }, |
| { |
| "epoch": 1.4145077720207253, |
| "grad_norm": 1.1462123394012451, |
| "learning_rate": 6.722334251421665e-06, |
| "loss": 0.3741143047809601, |
| "step": 35 |
| }, |
| { |
| "epoch": 1.455958549222798, |
| "grad_norm": 1.1399158239364624, |
| "learning_rate": 6.500423175001705e-06, |
| "loss": 0.37510427832603455, |
| "step": 36 |
| }, |
| { |
| "epoch": 1.4974093264248705, |
| "grad_norm": 1.118634819984436, |
| "learning_rate": 6.275213842808383e-06, |
| "loss": 0.3838654160499573, |
| "step": 37 |
| }, |
| { |
| "epoch": 1.5388601036269431, |
| "grad_norm": 1.110047459602356, |
| "learning_rate": 6.047201313830724e-06, |
| "loss": 0.38469570875167847, |
| "step": 38 |
| }, |
| { |
| "epoch": 1.5803108808290154, |
| "grad_norm": 1.1466346979141235, |
| "learning_rate": 5.816886809092651e-06, |
| "loss": 0.38225385546684265, |
| "step": 39 |
| }, |
| { |
| "epoch": 1.621761658031088, |
| "grad_norm": 1.1116628646850586, |
| "learning_rate": 5.584776609860414e-06, |
| "loss": 0.37283027172088623, |
| "step": 40 |
| }, |
| { |
| "epoch": 1.6632124352331608, |
| "grad_norm": 1.192334771156311, |
| "learning_rate": 5.351380944726465e-06, |
| "loss": 0.381252646446228, |
| "step": 41 |
| }, |
| { |
| "epoch": 1.704663212435233, |
| "grad_norm": 1.1414068937301636, |
| "learning_rate": 5.117212868016303e-06, |
| "loss": 0.37559133768081665, |
| "step": 42 |
| }, |
| { |
| "epoch": 1.7461139896373057, |
| "grad_norm": 1.0843558311462402, |
| "learning_rate": 4.882787131983698e-06, |
| "loss": 0.37207067012786865, |
| "step": 43 |
| }, |
| { |
| "epoch": 1.7875647668393784, |
| "grad_norm": 1.1973923444747925, |
| "learning_rate": 4.6486190552735375e-06, |
| "loss": 0.3798796236515045, |
| "step": 44 |
| }, |
| { |
| "epoch": 1.8290155440414506, |
| "grad_norm": 1.1059690713882446, |
| "learning_rate": 4.415223390139588e-06, |
| "loss": 0.37983524799346924, |
| "step": 45 |
| }, |
| { |
| "epoch": 1.8704663212435233, |
| "grad_norm": 1.0808876752853394, |
| "learning_rate": 4.183113190907349e-06, |
| "loss": 0.36370715498924255, |
| "step": 46 |
| }, |
| { |
| "epoch": 1.911917098445596, |
| "grad_norm": 1.1558469533920288, |
| "learning_rate": 3.952798686169279e-06, |
| "loss": 0.35979095101356506, |
| "step": 47 |
| }, |
| { |
| "epoch": 1.9533678756476682, |
| "grad_norm": 1.0924503803253174, |
| "learning_rate": 3.7247861571916183e-06, |
| "loss": 0.3729965388774872, |
| "step": 48 |
| }, |
| { |
| "epoch": 1.994818652849741, |
| "grad_norm": 1.1263461112976074, |
| "learning_rate": 3.4995768249982975e-06, |
| "loss": 0.3774448335170746, |
| "step": 49 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 3.2452127933502197, |
| "learning_rate": 3.2776657485783357e-06, |
| "loss": 0.3493170440196991, |
| "step": 50 |
| }, |
| { |
| "epoch": 2.0414507772020727, |
| "grad_norm": 1.1634777784347534, |
| "learning_rate": 3.059540736638751e-06, |
| "loss": 0.30372169613838196, |
| "step": 51 |
| }, |
| { |
| "epoch": 2.082901554404145, |
| "grad_norm": 1.145424246788025, |
| "learning_rate": 2.8456812752951483e-06, |
| "loss": 0.30972692370414734, |
| "step": 52 |
| }, |
| { |
| "epoch": 2.1243523316062176, |
| "grad_norm": 1.1736555099487305, |
| "learning_rate": 2.636557474057173e-06, |
| "loss": 0.2958666980266571, |
| "step": 53 |
| }, |
| { |
| "epoch": 2.1658031088082903, |
| "grad_norm": 1.1114293336868286, |
| "learning_rate": 2.4326290324257896e-06, |
| "loss": 0.30712461471557617, |
| "step": 54 |
| }, |
| { |
| "epoch": 2.2072538860103625, |
| "grad_norm": 1.124864935874939, |
| "learning_rate": 2.234344229374003e-06, |
| "loss": 0.3098776042461395, |
| "step": 55 |
| }, |
| { |
| "epoch": 2.2487046632124352, |
| "grad_norm": 1.0596271753311157, |
| "learning_rate": 2.042138937932388e-06, |
| "loss": 0.2897800803184509, |
| "step": 56 |
| }, |
| { |
| "epoch": 2.290155440414508, |
| "grad_norm": 1.0975596904754639, |
| "learning_rate": 1.856435667045577e-06, |
| "loss": 0.2883821725845337, |
| "step": 57 |
| }, |
| { |
| "epoch": 2.33160621761658, |
| "grad_norm": 1.0175517797470093, |
| "learning_rate": 1.677642632805892e-06, |
| "loss": 0.30240756273269653, |
| "step": 58 |
| }, |
| { |
| "epoch": 2.373056994818653, |
| "grad_norm": 1.0900399684906006, |
| "learning_rate": 1.5061528611057917e-06, |
| "loss": 0.2914489209651947, |
| "step": 59 |
| }, |
| { |
| "epoch": 2.4145077720207255, |
| "grad_norm": 1.058831810951233, |
| "learning_rate": 1.3423433236816563e-06, |
| "loss": 0.2905658483505249, |
| "step": 60 |
| }, |
| { |
| "epoch": 2.4559585492227978, |
| "grad_norm": 1.0601320266723633, |
| "learning_rate": 1.186574109448091e-06, |
| "loss": 0.2968388795852661, |
| "step": 61 |
| }, |
| { |
| "epoch": 2.4974093264248705, |
| "grad_norm": 1.0583627223968506, |
| "learning_rate": 1.0391876329443534e-06, |
| "loss": 0.2974582612514496, |
| "step": 62 |
| }, |
| { |
| "epoch": 2.538860103626943, |
| "grad_norm": 1.0791302919387817, |
| "learning_rate": 9.005078816328772e-07, |
| "loss": 0.2915363907814026, |
| "step": 63 |
| }, |
| { |
| "epoch": 2.5803108808290154, |
| "grad_norm": 1.0637892484664917, |
| "learning_rate": 7.708397037045129e-07, |
| "loss": 0.28974199295043945, |
| "step": 64 |
| }, |
| { |
| "epoch": 2.621761658031088, |
| "grad_norm": 1.0805476903915405, |
| "learning_rate": 6.50468137956049e-07, |
| "loss": 0.2940986156463623, |
| "step": 65 |
| }, |
| { |
| "epoch": 2.6632124352331608, |
| "grad_norm": 1.0829534530639648, |
| "learning_rate": 5.396577872130676e-07, |
| "loss": 0.2951546311378479, |
| "step": 66 |
| }, |
| { |
| "epoch": 2.704663212435233, |
| "grad_norm": 1.0342479944229126, |
| "learning_rate": 4.386522366755169e-07, |
| "loss": 0.28311941027641296, |
| "step": 67 |
| }, |
| { |
| "epoch": 2.7461139896373057, |
| "grad_norm": 1.0548819303512573, |
| "learning_rate": 3.4767351846456744e-07, |
| "loss": 0.2835821509361267, |
| "step": 68 |
| }, |
| { |
| "epoch": 2.7875647668393784, |
| "grad_norm": 1.0738929510116577, |
| "learning_rate": 2.669216235478295e-07, |
| "loss": 0.28978827595710754, |
| "step": 69 |
| }, |
| { |
| "epoch": 2.8290155440414506, |
| "grad_norm": 1.0309244394302368, |
| "learning_rate": 1.9657406211579966e-07, |
| "loss": 0.2983836233615875, |
| "step": 70 |
| }, |
| { |
| "epoch": 2.8704663212435233, |
| "grad_norm": 1.0488348007202148, |
| "learning_rate": 1.3678547337593494e-07, |
| "loss": 0.2910747230052948, |
| "step": 71 |
| }, |
| { |
| "epoch": 2.911917098445596, |
| "grad_norm": 1.0620460510253906, |
| "learning_rate": 8.768728562211948e-08, |
| "loss": 0.2941761910915375, |
| "step": 72 |
| }, |
| { |
| "epoch": 2.9533678756476682, |
| "grad_norm": 1.113092303276062, |
| "learning_rate": 4.9387427326745287e-08, |
| "loss": 0.3007400333881378, |
| "step": 73 |
| }, |
| { |
| "epoch": 2.994818652849741, |
| "grad_norm": 1.084908127784729, |
| "learning_rate": 2.1970089890509527e-08, |
| "loss": 0.29567357897758484, |
| "step": 74 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 2.7071526050567627, |
| "learning_rate": 5.495542571443135e-09, |
| "loss": 0.29400432109832764, |
| "step": 75 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 75, |
| "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": 1.7711341743552594e+18, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
