| { | |
| "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": 3.730775833129883, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.5604, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0759493670886076, | |
| "grad_norm": 3.6736209392547607, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.5593, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11392405063291139, | |
| "grad_norm": 3.981900930404663, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.6038, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 3.8179543018341064, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6291, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.189873417721519, | |
| "grad_norm": 2.5255050659179688, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.5591, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22784810126582278, | |
| "grad_norm": 2.027761220932007, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5085, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26582278481012656, | |
| "grad_norm": 3.217236280441284, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.5539, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.350926399230957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5758, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34177215189873417, | |
| "grad_norm": 2.6881141662597656, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.5012, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 2.5854835510253906, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.4834, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4177215189873418, | |
| "grad_norm": 1.539324164390564, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.4387, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.5237408876419067, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.5002, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4936708860759494, | |
| "grad_norm": 1.1909750699996948, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.4044, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5316455696202531, | |
| "grad_norm": 0.9470555782318115, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.387, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.569620253164557, | |
| "grad_norm": 0.8260341882705688, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.4018, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 0.7401832938194275, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.4, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6455696202531646, | |
| "grad_norm": 0.7956098914146423, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.4011, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6835443037974683, | |
| "grad_norm": 0.6847963929176331, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.3731, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7215189873417721, | |
| "grad_norm": 0.6996802687644958, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.3818, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.7493308782577515, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.4179, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7974683544303798, | |
| "grad_norm": 0.5356415510177612, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.3786, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8354430379746836, | |
| "grad_norm": 0.6219912767410278, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.368, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8734177215189873, | |
| "grad_norm": 0.5167115330696106, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.3686, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.5557819604873657, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.3481, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9493670886075949, | |
| "grad_norm": 0.5410380959510803, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.3491, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "grad_norm": 0.46455472707748413, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.3619, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0253164556962024, | |
| "grad_norm": 0.8875452876091003, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.6171, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0632911392405062, | |
| "grad_norm": 0.5331273078918457, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.3656, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.5281752347946167, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.3338, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 0.49164435267448425, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.3502, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1772151898734178, | |
| "grad_norm": 0.4439161419868469, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.3366, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2151898734177216, | |
| "grad_norm": 0.4810440242290497, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.3378, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.41076362133026123, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.3459, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2911392405063291, | |
| "grad_norm": 0.35466068983078003, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.2932, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3291139240506329, | |
| "grad_norm": 0.34635645151138306, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.3038, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3670886075949367, | |
| "grad_norm": 0.3956928253173828, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.3305, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.3470723628997803, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.3243, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4430379746835442, | |
| "grad_norm": 0.3708542585372925, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.3097, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.481012658227848, | |
| "grad_norm": 0.34405726194381714, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.2806, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 0.40255266427993774, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.3565, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.30957698822021484, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.2935, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.5949367088607596, | |
| "grad_norm": 0.34429731965065, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.3026, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6329113924050633, | |
| "grad_norm": 0.37622812390327454, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3221, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6708860759493671, | |
| "grad_norm": 0.3291604220867157, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.3033, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.31401413679122925, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.3059, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7468354430379747, | |
| "grad_norm": 0.3352019786834717, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3011, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7848101265822784, | |
| "grad_norm": 0.3071447014808655, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.3253, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8227848101265822, | |
| "grad_norm": 0.35264137387275696, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.3514, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.33903148770332336, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.308, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 0.32152560353279114, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.3251, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9367088607594938, | |
| "grad_norm": 0.3024255335330963, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.3039, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9746835443037973, | |
| "grad_norm": 0.28586339950561523, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.298, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0126582278481013, | |
| "grad_norm": 0.7231127023696899, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.5808, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.2656592130661011, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.3013, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.088607594936709, | |
| "grad_norm": 0.3404831290245056, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.319, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1265822784810124, | |
| "grad_norm": 0.2838174104690552, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.2785, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1645569620253164, | |
| "grad_norm": 0.3077821433544159, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.2878, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.3221641778945923, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.3001, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.240506329113924, | |
| "grad_norm": 0.29776477813720703, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.3073, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 0.2897549569606781, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.2899, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3164556962025316, | |
| "grad_norm": 0.28489553928375244, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.2983, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.24896368384361267, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.2788, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.392405063291139, | |
| "grad_norm": 0.27976804971694946, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.2916, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.430379746835443, | |
| "grad_norm": 0.26489588618278503, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.2914, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.4683544303797467, | |
| "grad_norm": 0.25014784932136536, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.2879, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.27488312125205994, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.3198, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5443037974683547, | |
| "grad_norm": 0.2766207456588745, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.324, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.5822784810126582, | |
| "grad_norm": 0.2341255247592926, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2938, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.620253164556962, | |
| "grad_norm": 0.23452407121658325, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.2742, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.24318979680538177, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2693, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.6962025316455698, | |
| "grad_norm": 0.258581280708313, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.3057, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7341772151898733, | |
| "grad_norm": 0.25280535221099854, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.2846, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7721518987341773, | |
| "grad_norm": 0.22219274938106537, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.2799, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.24744276702404022, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.298, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.848101265822785, | |
| "grad_norm": 0.22492331266403198, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.2843, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.8860759493670884, | |
| "grad_norm": 0.2421312928199768, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2815, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9240506329113924, | |
| "grad_norm": 0.25611355900764465, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.3065, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.23878204822540283, | |
| "learning_rate": 0.0, | |
| "loss": 0.267, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "step": 78, | |
| "total_flos": 153686659301376.0, | |
| "train_loss": 0.36335091522106755, | |
| "train_runtime": 1907.3671, | |
| "train_samples_per_second": 3.932, | |
| "train_steps_per_second": 0.041 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "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": 153686659301376.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |