| { | |
| "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": 6.2321955819532056, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.1299, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0759493670886076, | |
| "grad_norm": 6.397090952708749, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.11, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11392405063291139, | |
| "grad_norm": 6.349163602292759, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.125, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 5.713868391188725, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1546, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.189873417721519, | |
| "grad_norm": 4.287919076630793, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.0565, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22784810126582278, | |
| "grad_norm": 2.396856354263235, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9856, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26582278481012656, | |
| "grad_norm": 3.540150090341803, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.9961, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.759145737857168, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9622, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34177215189873417, | |
| "grad_norm": 3.086911074326093, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.9033, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 2.9618372690716477, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.9452, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4177215189873418, | |
| "grad_norm": 2.4763137661223658, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.9315, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 2.0926770659683953, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.8777, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4936708860759494, | |
| "grad_norm": 1.7505685764271455, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.848, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5316455696202531, | |
| "grad_norm": 1.512796616669028, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.8551, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.569620253164557, | |
| "grad_norm": 1.2026976231592823, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.8485, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.4393320274366892, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.8322, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6455696202531646, | |
| "grad_norm": 1.3566161198109896, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.8491, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6835443037974683, | |
| "grad_norm": 1.055807833762249, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.8419, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7215189873417721, | |
| "grad_norm": 0.996118367943648, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.8152, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.9844936604084962, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.7957, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7974683544303798, | |
| "grad_norm": 1.0450298818669628, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.7961, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8354430379746836, | |
| "grad_norm": 0.8924584541482958, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.8425, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8734177215189873, | |
| "grad_norm": 0.8970048266526586, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.7869, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.9573111409296974, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.8044, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9493670886075949, | |
| "grad_norm": 0.9945894336310853, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.8165, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "grad_norm": 0.8088482972004412, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.7404, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0253164556962024, | |
| "grad_norm": 1.752627743447675, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 1.2652, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0632911392405062, | |
| "grad_norm": 0.8309788977774167, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.6982, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.7812716184877682, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.7055, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 0.904952036923565, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.6881, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1772151898734178, | |
| "grad_norm": 0.7667411597000007, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.7038, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2151898734177216, | |
| "grad_norm": 0.8060826355713245, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.7522, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.8337103519716693, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.7306, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2911392405063291, | |
| "grad_norm": 0.8732782258487949, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.6936, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3291139240506329, | |
| "grad_norm": 0.7237473915756617, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.6713, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3670886075949367, | |
| "grad_norm": 0.9196013808252542, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.7093, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.8430526539448349, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.7263, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4430379746835442, | |
| "grad_norm": 0.7503720904346862, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.6946, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.481012658227848, | |
| "grad_norm": 0.7115951216039298, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.6824, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 0.7384652332779402, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.6891, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.6797422357784153, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.6183, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.5949367088607596, | |
| "grad_norm": 0.8045133198446812, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.7234, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6329113924050633, | |
| "grad_norm": 0.6643896406710053, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6443, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6708860759493671, | |
| "grad_norm": 0.7527969997155343, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.7151, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.7580822381393623, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.7331, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7468354430379747, | |
| "grad_norm": 0.7226411658606853, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.7688, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.7848101265822784, | |
| "grad_norm": 0.7165104681960608, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.6973, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8227848101265822, | |
| "grad_norm": 0.6956013338621195, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.6475, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.6852345882443253, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6481, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 0.7959804905336986, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.6982, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9367088607594938, | |
| "grad_norm": 0.7058897097194478, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.6517, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.9746835443037973, | |
| "grad_norm": 0.687820847173246, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.7307, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0126582278481013, | |
| "grad_norm": 1.2952298279020342, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 1.0404, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.6827146755971158, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.6627, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.088607594936709, | |
| "grad_norm": 0.6800945354242551, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.6245, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1265822784810124, | |
| "grad_norm": 0.6113298819680627, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.6195, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1645569620253164, | |
| "grad_norm": 0.5999594472055836, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.6018, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.7003806513414853, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.6285, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.240506329113924, | |
| "grad_norm": 0.6016362024524727, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5301, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 0.6309273405878884, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.6759, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3164556962025316, | |
| "grad_norm": 0.6855054019550281, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.6941, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.6907968679521843, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.6191, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.392405063291139, | |
| "grad_norm": 0.6518455679940384, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.6541, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.430379746835443, | |
| "grad_norm": 0.5927051903483913, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.5554, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.4683544303797467, | |
| "grad_norm": 0.6141850816209767, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.678, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.600733293987595, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.6257, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5443037974683547, | |
| "grad_norm": 0.5862063021575895, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5732, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.5822784810126582, | |
| "grad_norm": 0.6417002974092123, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.6665, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.620253164556962, | |
| "grad_norm": 0.6156545677126318, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.547, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.6481985817296648, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.6649, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.6962025316455698, | |
| "grad_norm": 0.6004076915534737, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.6024, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.7341772151898733, | |
| "grad_norm": 0.6346585334888803, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5961, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7721518987341773, | |
| "grad_norm": 0.6458655409414998, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.6483, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.6443351296756988, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.6368, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.848101265822785, | |
| "grad_norm": 0.611399715797446, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.6091, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.8860759493670884, | |
| "grad_norm": 0.6085566469234578, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.6389, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9240506329113924, | |
| "grad_norm": 0.6268652042686883, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.6282, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.6139445375041865, | |
| "learning_rate": 0.0, | |
| "loss": 0.6352, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "step": 78, | |
| "total_flos": 34697443155968.0, | |
| "train_loss": 0.7563248490675901, | |
| "train_runtime": 1050.0431, | |
| "train_samples_per_second": 7.143, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "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": 34697443155968.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |