| { | |
| "best_metric": 1.2524751424789429, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-250", | |
| "epoch": 3.005586592178771, | |
| "eval_steps": 50, | |
| "global_step": 269, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0111731843575419, | |
| "grad_norm": 162.40016174316406, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 7.5612, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0111731843575419, | |
| "eval_loss": 10.110641479492188, | |
| "eval_runtime": 3.5559, | |
| "eval_samples_per_second": 42.465, | |
| "eval_steps_per_second": 21.373, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0223463687150838, | |
| "grad_norm": 201.2489776611328, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 8.9985, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0335195530726257, | |
| "grad_norm": 181.89503479003906, | |
| "learning_rate": 1e-05, | |
| "loss": 9.1837, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0446927374301676, | |
| "grad_norm": 153.2060546875, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 8.9975, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.055865921787709494, | |
| "grad_norm": 104.02288818359375, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 7.9499, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0670391061452514, | |
| "grad_norm": 92.24451446533203, | |
| "learning_rate": 2e-05, | |
| "loss": 7.5897, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0782122905027933, | |
| "grad_norm": 65.57331085205078, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 6.5642, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0893854748603352, | |
| "grad_norm": 28.319116592407227, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 5.9872, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1005586592178771, | |
| "grad_norm": 84.72834777832031, | |
| "learning_rate": 3e-05, | |
| "loss": 5.6777, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.11173184357541899, | |
| "grad_norm": 62.04957962036133, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 5.452, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12290502793296089, | |
| "grad_norm": 23.43031883239746, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 5.0966, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1340782122905028, | |
| "grad_norm": 11.170083045959473, | |
| "learning_rate": 4e-05, | |
| "loss": 4.9892, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1452513966480447, | |
| "grad_norm": 9.359170913696289, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 4.8074, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1564245810055866, | |
| "grad_norm": 7.227917194366455, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 4.5527, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16759776536312848, | |
| "grad_norm": 5.746039867401123, | |
| "learning_rate": 5e-05, | |
| "loss": 4.4023, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1787709497206704, | |
| "grad_norm": 5.6375274658203125, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 4.2542, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18994413407821228, | |
| "grad_norm": 5.626579761505127, | |
| "learning_rate": 5.666666666666667e-05, | |
| "loss": 4.205, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2011173184357542, | |
| "grad_norm": 6.251049041748047, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1443, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2122905027932961, | |
| "grad_norm": 5.178546905517578, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 3.7947, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.22346368715083798, | |
| "grad_norm": 5.792481899261475, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 3.6496, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2346368715083799, | |
| "grad_norm": 8.204561233520508, | |
| "learning_rate": 7e-05, | |
| "loss": 3.4046, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.24581005586592178, | |
| "grad_norm": 6.862538814544678, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 3.2466, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2569832402234637, | |
| "grad_norm": 6.1007490158081055, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 3.1117, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2681564245810056, | |
| "grad_norm": 4.2520365715026855, | |
| "learning_rate": 8e-05, | |
| "loss": 3.1683, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.27932960893854747, | |
| "grad_norm": 16.54338264465332, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 3.0407, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2905027932960894, | |
| "grad_norm": 5.3217034339904785, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 2.8435, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3016759776536313, | |
| "grad_norm": 5.066921234130859, | |
| "learning_rate": 9e-05, | |
| "loss": 2.8639, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3128491620111732, | |
| "grad_norm": 4.023467063903809, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 2.7216, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3240223463687151, | |
| "grad_norm": 5.351987361907959, | |
| "learning_rate": 9.666666666666667e-05, | |
| "loss": 2.503, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.33519553072625696, | |
| "grad_norm": 10.897500038146973, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5283, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3463687150837989, | |
| "grad_norm": 3.5467631816864014, | |
| "learning_rate": 9.999568045802217e-05, | |
| "loss": 2.4514, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3575418994413408, | |
| "grad_norm": 4.198688507080078, | |
| "learning_rate": 9.998272257842641e-05, | |
| "loss": 2.6802, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3687150837988827, | |
| "grad_norm": 2.829094409942627, | |
| "learning_rate": 9.996112860009688e-05, | |
| "loss": 2.359, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.37988826815642457, | |
| "grad_norm": 2.836956262588501, | |
| "learning_rate": 9.993090225407743e-05, | |
| "loss": 2.1674, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.39106145251396646, | |
| "grad_norm": 3.0385210514068604, | |
| "learning_rate": 9.989204876292688e-05, | |
| "loss": 2.1289, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4022346368715084, | |
| "grad_norm": 3.1165988445281982, | |
| "learning_rate": 9.984457483981669e-05, | |
| "loss": 2.0466, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4134078212290503, | |
| "grad_norm": 2.5134077072143555, | |
| "learning_rate": 9.978848868737098e-05, | |
| "loss": 2.0756, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4245810055865922, | |
| "grad_norm": 2.617610454559326, | |
| "learning_rate": 9.972379999624936e-05, | |
| "loss": 1.8898, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.43575418994413406, | |
| "grad_norm": 2.49375581741333, | |
| "learning_rate": 9.96505199434725e-05, | |
| "loss": 1.8914, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.44692737430167595, | |
| "grad_norm": 2.5499541759490967, | |
| "learning_rate": 9.956866119049095e-05, | |
| "loss": 1.74, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4581005586592179, | |
| "grad_norm": 2.702899217605591, | |
| "learning_rate": 9.947823788099753e-05, | |
| "loss": 1.6642, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4692737430167598, | |
| "grad_norm": 2.9417200088500977, | |
| "learning_rate": 9.937926563848346e-05, | |
| "loss": 1.7848, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.48044692737430167, | |
| "grad_norm": 2.8817129135131836, | |
| "learning_rate": 9.927176156353899e-05, | |
| "loss": 1.7686, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.49162011173184356, | |
| "grad_norm": 2.706786632537842, | |
| "learning_rate": 9.91557442308987e-05, | |
| "loss": 1.5918, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5027932960893855, | |
| "grad_norm": 4.0628342628479, | |
| "learning_rate": 9.903123368623216e-05, | |
| "loss": 2.0585, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5139664804469274, | |
| "grad_norm": 2.9532899856567383, | |
| "learning_rate": 9.889825144268029e-05, | |
| "loss": 1.9343, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5251396648044693, | |
| "grad_norm": 2.189305305480957, | |
| "learning_rate": 9.875682047713846e-05, | |
| "loss": 1.9791, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5363128491620112, | |
| "grad_norm": 2.71520733833313, | |
| "learning_rate": 9.860696522628639e-05, | |
| "loss": 2.0729, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.547486033519553, | |
| "grad_norm": 2.677870988845825, | |
| "learning_rate": 9.844871158236591e-05, | |
| "loss": 1.8758, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5586592178770949, | |
| "grad_norm": 2.1260600090026855, | |
| "learning_rate": 9.828208688870735e-05, | |
| "loss": 1.8366, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5586592178770949, | |
| "eval_loss": 1.7909045219421387, | |
| "eval_runtime": 3.541, | |
| "eval_samples_per_second": 42.643, | |
| "eval_steps_per_second": 21.463, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5698324022346368, | |
| "grad_norm": 1.9885830879211426, | |
| "learning_rate": 9.810711993500507e-05, | |
| "loss": 1.7649, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5810055865921788, | |
| "grad_norm": 1.8383235931396484, | |
| "learning_rate": 9.792384095234313e-05, | |
| "loss": 1.6706, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5921787709497207, | |
| "grad_norm": 2.144374370574951, | |
| "learning_rate": 9.773228160797188e-05, | |
| "loss": 1.5816, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6033519553072626, | |
| "grad_norm": 1.8016228675842285, | |
| "learning_rate": 9.753247499983649e-05, | |
| "loss": 1.7056, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.6145251396648045, | |
| "grad_norm": 1.934396505355835, | |
| "learning_rate": 9.732445565085824e-05, | |
| "loss": 1.8101, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6256983240223464, | |
| "grad_norm": 1.7126082181930542, | |
| "learning_rate": 9.71082595029695e-05, | |
| "loss": 1.7412, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6368715083798883, | |
| "grad_norm": 2.053506851196289, | |
| "learning_rate": 9.688392391090373e-05, | |
| "loss": 1.6484, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6480446927374302, | |
| "grad_norm": 1.8453326225280762, | |
| "learning_rate": 9.665148763574123e-05, | |
| "loss": 1.7627, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.659217877094972, | |
| "grad_norm": 1.7451444864273071, | |
| "learning_rate": 9.64109908382119e-05, | |
| "loss": 1.4411, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6703910614525139, | |
| "grad_norm": 1.7417210340499878, | |
| "learning_rate": 9.616247507175623e-05, | |
| "loss": 1.546, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6815642458100558, | |
| "grad_norm": 2.221048355102539, | |
| "learning_rate": 9.590598327534564e-05, | |
| "loss": 1.7249, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6927374301675978, | |
| "grad_norm": 2.038527727127075, | |
| "learning_rate": 9.564155976606339e-05, | |
| "loss": 1.8133, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.7039106145251397, | |
| "grad_norm": 2.2194125652313232, | |
| "learning_rate": 9.536925023144742e-05, | |
| "loss": 1.7806, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.7150837988826816, | |
| "grad_norm": 1.9220144748687744, | |
| "learning_rate": 9.508910172159635e-05, | |
| "loss": 1.5446, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7262569832402235, | |
| "grad_norm": 1.841299057006836, | |
| "learning_rate": 9.480116264104011e-05, | |
| "loss": 1.4487, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7374301675977654, | |
| "grad_norm": 2.3878183364868164, | |
| "learning_rate": 9.450548274037653e-05, | |
| "loss": 1.4077, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7486033519553073, | |
| "grad_norm": 2.5445780754089355, | |
| "learning_rate": 9.420211310767533e-05, | |
| "loss": 1.8375, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7597765363128491, | |
| "grad_norm": 1.9894568920135498, | |
| "learning_rate": 9.389110615965102e-05, | |
| "loss": 1.5472, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.770949720670391, | |
| "grad_norm": 1.5543500185012817, | |
| "learning_rate": 9.35725156326063e-05, | |
| "loss": 1.5146, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7821229050279329, | |
| "grad_norm": 1.6577750444412231, | |
| "learning_rate": 9.324639657314742e-05, | |
| "loss": 1.5323, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7932960893854749, | |
| "grad_norm": 1.6345881223678589, | |
| "learning_rate": 9.291280532867302e-05, | |
| "loss": 1.6763, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.8044692737430168, | |
| "grad_norm": 1.576483130455017, | |
| "learning_rate": 9.257179953763845e-05, | |
| "loss": 1.6414, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8156424581005587, | |
| "grad_norm": 1.661361813545227, | |
| "learning_rate": 9.222343811959693e-05, | |
| "loss": 1.6203, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8268156424581006, | |
| "grad_norm": 1.7419025897979736, | |
| "learning_rate": 9.186778126501916e-05, | |
| "loss": 1.4892, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8379888268156425, | |
| "grad_norm": 1.5462729930877686, | |
| "learning_rate": 9.150489042489367e-05, | |
| "loss": 1.6116, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8491620111731844, | |
| "grad_norm": 1.5897197723388672, | |
| "learning_rate": 9.113482830010918e-05, | |
| "loss": 1.5627, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8603351955307262, | |
| "grad_norm": 1.6474636793136597, | |
| "learning_rate": 9.075765883062093e-05, | |
| "loss": 1.5695, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8715083798882681, | |
| "grad_norm": 1.5242871046066284, | |
| "learning_rate": 9.037344718440322e-05, | |
| "loss": 1.5686, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.88268156424581, | |
| "grad_norm": 1.4273405075073242, | |
| "learning_rate": 8.99822597461894e-05, | |
| "loss": 1.2552, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8938547486033519, | |
| "grad_norm": 1.63188636302948, | |
| "learning_rate": 8.958416410600187e-05, | |
| "loss": 1.379, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9050279329608939, | |
| "grad_norm": 1.5915851593017578, | |
| "learning_rate": 8.917922904747384e-05, | |
| "loss": 1.5448, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.9162011173184358, | |
| "grad_norm": 1.4518282413482666, | |
| "learning_rate": 8.876752453596462e-05, | |
| "loss": 1.3964, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9273743016759777, | |
| "grad_norm": 1.5592926740646362, | |
| "learning_rate": 8.834912170647101e-05, | |
| "loss": 1.3952, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9385474860335196, | |
| "grad_norm": 1.836943507194519, | |
| "learning_rate": 8.792409285133642e-05, | |
| "loss": 1.542, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9497206703910615, | |
| "grad_norm": 1.6657090187072754, | |
| "learning_rate": 8.749251140776016e-05, | |
| "loss": 1.4903, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9608938547486033, | |
| "grad_norm": 1.8024177551269531, | |
| "learning_rate": 8.705445194510868e-05, | |
| "loss": 1.4487, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9720670391061452, | |
| "grad_norm": 1.9545961618423462, | |
| "learning_rate": 8.66099901520315e-05, | |
| "loss": 1.5293, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9832402234636871, | |
| "grad_norm": 2.1410531997680664, | |
| "learning_rate": 8.615920282338355e-05, | |
| "loss": 1.4279, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.994413407821229, | |
| "grad_norm": 1.8293300867080688, | |
| "learning_rate": 8.570216784695637e-05, | |
| "loss": 1.5349, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.005586592178771, | |
| "grad_norm": 2.8246774673461914, | |
| "learning_rate": 8.52389641900206e-05, | |
| "loss": 2.6159, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0167597765363128, | |
| "grad_norm": 1.4292484521865845, | |
| "learning_rate": 8.476967188568188e-05, | |
| "loss": 1.6759, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0279329608938548, | |
| "grad_norm": 1.2621123790740967, | |
| "learning_rate": 8.429437201905254e-05, | |
| "loss": 1.4449, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0391061452513966, | |
| "grad_norm": 1.4738688468933105, | |
| "learning_rate": 8.381314671324159e-05, | |
| "loss": 1.4352, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0502793296089385, | |
| "grad_norm": 1.4630380868911743, | |
| "learning_rate": 8.332607911516545e-05, | |
| "loss": 1.4337, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0614525139664805, | |
| "grad_norm": 1.4351942539215088, | |
| "learning_rate": 8.283325338118153e-05, | |
| "loss": 1.4363, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0726256983240223, | |
| "grad_norm": 1.1830869913101196, | |
| "learning_rate": 8.233475466254765e-05, | |
| "loss": 1.4274, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0837988826815643, | |
| "grad_norm": 1.3410512208938599, | |
| "learning_rate": 8.183066909070947e-05, | |
| "loss": 1.3677, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.094972067039106, | |
| "grad_norm": 1.4120603799819946, | |
| "learning_rate": 8.132108376241849e-05, | |
| "loss": 1.3283, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.106145251396648, | |
| "grad_norm": 1.2132238149642944, | |
| "learning_rate": 8.08060867246834e-05, | |
| "loss": 1.3961, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.1173184357541899, | |
| "grad_norm": 1.466185450553894, | |
| "learning_rate": 8.028576695955711e-05, | |
| "loss": 1.3326, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1173184357541899, | |
| "eval_loss": 1.4376939535140991, | |
| "eval_runtime": 3.541, | |
| "eval_samples_per_second": 42.644, | |
| "eval_steps_per_second": 21.463, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1284916201117319, | |
| "grad_norm": 1.4509429931640625, | |
| "learning_rate": 7.97602143687623e-05, | |
| "loss": 1.3226, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1396648044692737, | |
| "grad_norm": 1.453986406326294, | |
| "learning_rate": 7.922951975815811e-05, | |
| "loss": 1.2928, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1508379888268156, | |
| "grad_norm": 1.3562222719192505, | |
| "learning_rate": 7.869377482205042e-05, | |
| "loss": 1.2992, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1620111731843576, | |
| "grad_norm": 1.3154160976409912, | |
| "learning_rate": 7.815307212734888e-05, | |
| "loss": 1.2855, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1731843575418994, | |
| "grad_norm": 1.438166856765747, | |
| "learning_rate": 7.760750509757298e-05, | |
| "loss": 1.1333, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1843575418994414, | |
| "grad_norm": 1.4575860500335693, | |
| "learning_rate": 7.705716799671019e-05, | |
| "loss": 1.2916, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1955307262569832, | |
| "grad_norm": 1.525395154953003, | |
| "learning_rate": 7.650215591292888e-05, | |
| "loss": 1.2428, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.2067039106145252, | |
| "grad_norm": 1.7797048091888428, | |
| "learning_rate": 7.594256474214882e-05, | |
| "loss": 1.3893, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.217877094972067, | |
| "grad_norm": 1.5947117805480957, | |
| "learning_rate": 7.537849117147212e-05, | |
| "loss": 1.3861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.229050279329609, | |
| "grad_norm": 1.948651671409607, | |
| "learning_rate": 7.481003266247744e-05, | |
| "loss": 1.3417, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2402234636871508, | |
| "grad_norm": 1.8325766324996948, | |
| "learning_rate": 7.423728743438048e-05, | |
| "loss": 1.361, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2513966480446927, | |
| "grad_norm": 1.443632960319519, | |
| "learning_rate": 7.366035444706347e-05, | |
| "loss": 1.3431, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2625698324022347, | |
| "grad_norm": 1.3943829536437988, | |
| "learning_rate": 7.307933338397667e-05, | |
| "loss": 1.3545, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2737430167597765, | |
| "grad_norm": 1.4311888217926025, | |
| "learning_rate": 7.249432463491498e-05, | |
| "loss": 1.3241, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2849162011173183, | |
| "grad_norm": 1.2513173818588257, | |
| "learning_rate": 7.190542927867234e-05, | |
| "loss": 1.4081, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2960893854748603, | |
| "grad_norm": 1.2229642868041992, | |
| "learning_rate": 7.131274906557725e-05, | |
| "loss": 1.5218, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.3072625698324023, | |
| "grad_norm": 1.3140783309936523, | |
| "learning_rate": 7.071638639991207e-05, | |
| "loss": 1.4195, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.318435754189944, | |
| "grad_norm": 1.3129864931106567, | |
| "learning_rate": 7.011644432221958e-05, | |
| "loss": 1.4635, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.329608938547486, | |
| "grad_norm": 1.3215774297714233, | |
| "learning_rate": 6.95130264914993e-05, | |
| "loss": 1.447, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3407821229050279, | |
| "grad_norm": 1.2216203212738037, | |
| "learning_rate": 6.890623716729724e-05, | |
| "loss": 1.1941, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3519553072625698, | |
| "grad_norm": 1.3087091445922852, | |
| "learning_rate": 6.82961811916917e-05, | |
| "loss": 1.4038, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3631284916201118, | |
| "grad_norm": 1.1932587623596191, | |
| "learning_rate": 6.768296397117848e-05, | |
| "loss": 1.3638, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3743016759776536, | |
| "grad_norm": 1.3358490467071533, | |
| "learning_rate": 6.706669145845863e-05, | |
| "loss": 1.3817, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3854748603351954, | |
| "grad_norm": 1.1948127746582031, | |
| "learning_rate": 6.644747013413168e-05, | |
| "loss": 1.2415, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3966480446927374, | |
| "grad_norm": 1.1789873838424683, | |
| "learning_rate": 6.582540698829781e-05, | |
| "loss": 1.0901, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.4078212290502794, | |
| "grad_norm": 1.2974153757095337, | |
| "learning_rate": 6.520060950207185e-05, | |
| "loss": 1.2798, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.4189944134078212, | |
| "grad_norm": 1.408875584602356, | |
| "learning_rate": 6.457318562901256e-05, | |
| "loss": 1.3745, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4301675977653632, | |
| "grad_norm": 1.4454888105392456, | |
| "learning_rate": 6.394324377647028e-05, | |
| "loss": 1.3028, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.441340782122905, | |
| "grad_norm": 1.549099087715149, | |
| "learning_rate": 6.331089278685599e-05, | |
| "loss": 1.3293, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.452513966480447, | |
| "grad_norm": 1.4389153718948364, | |
| "learning_rate": 6.26762419188355e-05, | |
| "loss": 1.1637, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.463687150837989, | |
| "grad_norm": 1.408833622932434, | |
| "learning_rate": 6.203940082845144e-05, | |
| "loss": 1.1842, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4748603351955307, | |
| "grad_norm": 1.5742685794830322, | |
| "learning_rate": 6.140047955017671e-05, | |
| "loss": 1.2201, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4860335195530725, | |
| "grad_norm": 1.8778325319290161, | |
| "learning_rate": 6.075958847790262e-05, | |
| "loss": 1.1914, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4972067039106145, | |
| "grad_norm": 1.5008203983306885, | |
| "learning_rate": 6.011683834586473e-05, | |
| "loss": 1.2746, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.5083798882681565, | |
| "grad_norm": 1.3712409734725952, | |
| "learning_rate": 5.947234020951015e-05, | |
| "loss": 1.244, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5195530726256983, | |
| "grad_norm": 1.298775315284729, | |
| "learning_rate": 5.882620542630901e-05, | |
| "loss": 1.213, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.5307262569832403, | |
| "grad_norm": 1.2305636405944824, | |
| "learning_rate": 5.8178545636514145e-05, | |
| "loss": 1.3607, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.541899441340782, | |
| "grad_norm": 1.1469392776489258, | |
| "learning_rate": 5.752947274387147e-05, | |
| "loss": 1.349, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.553072625698324, | |
| "grad_norm": 1.1323195695877075, | |
| "learning_rate": 5.687909889628529e-05, | |
| "loss": 1.3132, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.564245810055866, | |
| "grad_norm": 1.0949897766113281, | |
| "learning_rate": 5.622753646644102e-05, | |
| "loss": 1.2518, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5754189944134078, | |
| "grad_norm": 1.1683642864227295, | |
| "learning_rate": 5.557489803238933e-05, | |
| "loss": 1.3319, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5865921787709496, | |
| "grad_norm": 1.2218724489212036, | |
| "learning_rate": 5.492129635809473e-05, | |
| "loss": 1.3427, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5977653631284916, | |
| "grad_norm": 1.133186936378479, | |
| "learning_rate": 5.426684437395196e-05, | |
| "loss": 1.2725, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.6089385474860336, | |
| "grad_norm": 1.2256780862808228, | |
| "learning_rate": 5.361165515727374e-05, | |
| "loss": 1.2708, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.6201117318435754, | |
| "grad_norm": 1.2726577520370483, | |
| "learning_rate": 5.295584191275308e-05, | |
| "loss": 1.3795, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6312849162011172, | |
| "grad_norm": 1.2547286748886108, | |
| "learning_rate": 5.229951795290353e-05, | |
| "loss": 1.3061, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.6424581005586592, | |
| "grad_norm": 1.2261368036270142, | |
| "learning_rate": 5.164279667848094e-05, | |
| "loss": 1.245, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6536312849162011, | |
| "grad_norm": 1.2689229249954224, | |
| "learning_rate": 5.0985791558889785e-05, | |
| "loss": 1.3172, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6648044692737431, | |
| "grad_norm": 1.2573802471160889, | |
| "learning_rate": 5.032861611257783e-05, | |
| "loss": 1.2967, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.675977653631285, | |
| "grad_norm": 1.4018381834030151, | |
| "learning_rate": 4.967138388742218e-05, | |
| "loss": 1.2735, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.675977653631285, | |
| "eval_loss": 1.326945424079895, | |
| "eval_runtime": 3.5376, | |
| "eval_samples_per_second": 42.685, | |
| "eval_steps_per_second": 21.484, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6871508379888267, | |
| "grad_norm": 1.5333133935928345, | |
| "learning_rate": 4.901420844111021e-05, | |
| "loss": 1.3852, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6983240223463687, | |
| "grad_norm": 1.3856674432754517, | |
| "learning_rate": 4.835720332151907e-05, | |
| "loss": 1.1126, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.7094972067039107, | |
| "grad_norm": 1.6033830642700195, | |
| "learning_rate": 4.770048204709648e-05, | |
| "loss": 1.3011, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.7206703910614525, | |
| "grad_norm": 1.6162670850753784, | |
| "learning_rate": 4.7044158087246926e-05, | |
| "loss": 1.2179, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.7318435754189943, | |
| "grad_norm": 1.744236946105957, | |
| "learning_rate": 4.6388344842726264e-05, | |
| "loss": 1.1941, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7430167597765363, | |
| "grad_norm": 1.5789906978607178, | |
| "learning_rate": 4.5733155626048036e-05, | |
| "loss": 1.3045, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7541899441340782, | |
| "grad_norm": 1.226691722869873, | |
| "learning_rate": 4.507870364190527e-05, | |
| "loss": 1.3675, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7653631284916202, | |
| "grad_norm": 1.3142743110656738, | |
| "learning_rate": 4.4425101967610674e-05, | |
| "loss": 1.4387, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.776536312849162, | |
| "grad_norm": 1.0727818012237549, | |
| "learning_rate": 4.377246353355899e-05, | |
| "loss": 1.2175, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7877094972067038, | |
| "grad_norm": 1.1169428825378418, | |
| "learning_rate": 4.312090110371473e-05, | |
| "loss": 1.365, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.7988826815642458, | |
| "grad_norm": 1.0804024934768677, | |
| "learning_rate": 4.247052725612852e-05, | |
| "loss": 1.3312, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.8100558659217878, | |
| "grad_norm": 1.2123122215270996, | |
| "learning_rate": 4.1821454363485866e-05, | |
| "loss": 1.3837, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.8212290502793296, | |
| "grad_norm": 1.1932836771011353, | |
| "learning_rate": 4.1173794573690996e-05, | |
| "loss": 1.3426, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.8324022346368714, | |
| "grad_norm": 1.0667084455490112, | |
| "learning_rate": 4.052765979048986e-05, | |
| "loss": 1.223, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.8435754189944134, | |
| "grad_norm": 1.1312137842178345, | |
| "learning_rate": 3.988316165413528e-05, | |
| "loss": 1.2159, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8547486033519553, | |
| "grad_norm": 1.135401725769043, | |
| "learning_rate": 3.924041152209739e-05, | |
| "loss": 1.1931, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8659217877094973, | |
| "grad_norm": 1.2866289615631104, | |
| "learning_rate": 3.859952044982329e-05, | |
| "loss": 1.3266, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8770949720670391, | |
| "grad_norm": 1.3166481256484985, | |
| "learning_rate": 3.7960599171548574e-05, | |
| "loss": 1.2229, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.888268156424581, | |
| "grad_norm": 1.2265543937683105, | |
| "learning_rate": 3.732375808116451e-05, | |
| "loss": 1.2899, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.899441340782123, | |
| "grad_norm": 1.2559617757797241, | |
| "learning_rate": 3.668910721314402e-05, | |
| "loss": 1.13, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.910614525139665, | |
| "grad_norm": 1.2486090660095215, | |
| "learning_rate": 3.605675622352973e-05, | |
| "loss": 1.2083, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.9217877094972067, | |
| "grad_norm": 1.406841516494751, | |
| "learning_rate": 3.542681437098745e-05, | |
| "loss": 1.2352, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.9329608938547485, | |
| "grad_norm": 1.4469358921051025, | |
| "learning_rate": 3.479939049792817e-05, | |
| "loss": 1.2662, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.9441340782122905, | |
| "grad_norm": 1.4758639335632324, | |
| "learning_rate": 3.417459301170219e-05, | |
| "loss": 1.2432, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9553072625698324, | |
| "grad_norm": 1.3574011325836182, | |
| "learning_rate": 3.355252986586832e-05, | |
| "loss": 1.1044, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9664804469273744, | |
| "grad_norm": 1.8598283529281616, | |
| "learning_rate": 3.293330854154136e-05, | |
| "loss": 1.3518, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9776536312849162, | |
| "grad_norm": 1.7371652126312256, | |
| "learning_rate": 3.2317036028821523e-05, | |
| "loss": 1.2855, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.988826815642458, | |
| "grad_norm": 1.4092578887939453, | |
| "learning_rate": 3.1703818808308324e-05, | |
| "loss": 1.3558, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.122119188308716, | |
| "learning_rate": 3.109376283270277e-05, | |
| "loss": 1.6615, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.011173184357542, | |
| "grad_norm": 1.219295859336853, | |
| "learning_rate": 3.0486973508500727e-05, | |
| "loss": 1.4017, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.022346368715084, | |
| "grad_norm": 1.1538641452789307, | |
| "learning_rate": 2.988355567778043e-05, | |
| "loss": 1.2616, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.0335195530726256, | |
| "grad_norm": 1.0700401067733765, | |
| "learning_rate": 2.9283613600087933e-05, | |
| "loss": 1.2648, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.0446927374301676, | |
| "grad_norm": 0.9880838394165039, | |
| "learning_rate": 2.8687250934422772e-05, | |
| "loss": 1.1885, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.0558659217877095, | |
| "grad_norm": 1.051222801208496, | |
| "learning_rate": 2.8094570721327662e-05, | |
| "loss": 1.1746, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.0670391061452515, | |
| "grad_norm": 1.0639920234680176, | |
| "learning_rate": 2.750567536508504e-05, | |
| "loss": 1.2354, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.078212290502793, | |
| "grad_norm": 1.0196352005004883, | |
| "learning_rate": 2.6920666616023327e-05, | |
| "loss": 1.2534, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.089385474860335, | |
| "grad_norm": 1.0681345462799072, | |
| "learning_rate": 2.6339645552936536e-05, | |
| "loss": 1.1096, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.100558659217877, | |
| "grad_norm": 1.196368932723999, | |
| "learning_rate": 2.5762712565619528e-05, | |
| "loss": 1.1021, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.111731843575419, | |
| "grad_norm": 1.0771485567092896, | |
| "learning_rate": 2.5189967337522573e-05, | |
| "loss": 1.0798, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.122905027932961, | |
| "grad_norm": 1.11617910861969, | |
| "learning_rate": 2.46215088285279e-05, | |
| "loss": 1.1074, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.1340782122905027, | |
| "grad_norm": 1.1497374773025513, | |
| "learning_rate": 2.4057435257851175e-05, | |
| "loss": 1.1183, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.1452513966480447, | |
| "grad_norm": 1.1393492221832275, | |
| "learning_rate": 2.349784408707112e-05, | |
| "loss": 1.0711, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.1564245810055866, | |
| "grad_norm": 1.0959515571594238, | |
| "learning_rate": 2.2942832003289823e-05, | |
| "loss": 1.0437, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1675977653631286, | |
| "grad_norm": 1.155680775642395, | |
| "learning_rate": 2.2392494902427025e-05, | |
| "loss": 1.047, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.17877094972067, | |
| "grad_norm": 1.157484531402588, | |
| "learning_rate": 2.1846927872651137e-05, | |
| "loss": 1.0649, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.189944134078212, | |
| "grad_norm": 1.2944828271865845, | |
| "learning_rate": 2.1306225177949585e-05, | |
| "loss": 1.1103, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.201117318435754, | |
| "grad_norm": 1.5556319952011108, | |
| "learning_rate": 2.07704802418419e-05, | |
| "loss": 1.2251, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.212290502793296, | |
| "grad_norm": 1.4562413692474365, | |
| "learning_rate": 2.0239785631237705e-05, | |
| "loss": 1.0627, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.223463687150838, | |
| "grad_norm": 1.393690824508667, | |
| "learning_rate": 1.9714233040442915e-05, | |
| "loss": 0.9858, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.2346368715083798, | |
| "grad_norm": 1.6887412071228027, | |
| "learning_rate": 1.9193913275316626e-05, | |
| "loss": 1.1138, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2346368715083798, | |
| "eval_loss": 1.2823649644851685, | |
| "eval_runtime": 3.5435, | |
| "eval_samples_per_second": 42.613, | |
| "eval_steps_per_second": 21.448, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2458100558659218, | |
| "grad_norm": 2.0551750659942627, | |
| "learning_rate": 1.8678916237581522e-05, | |
| "loss": 1.0606, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.2569832402234637, | |
| "grad_norm": 1.3848985433578491, | |
| "learning_rate": 1.816933090929055e-05, | |
| "loss": 1.2715, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.2681564245810057, | |
| "grad_norm": 1.1580418348312378, | |
| "learning_rate": 1.7665245337452368e-05, | |
| "loss": 1.1072, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.2793296089385473, | |
| "grad_norm": 1.1287815570831299, | |
| "learning_rate": 1.716674661881848e-05, | |
| "loss": 1.1187, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.2905027932960893, | |
| "grad_norm": 1.1857539415359497, | |
| "learning_rate": 1.667392088483456e-05, | |
| "loss": 1.3102, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.3016759776536313, | |
| "grad_norm": 1.0973467826843262, | |
| "learning_rate": 1.6186853286758397e-05, | |
| "loss": 1.2696, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.3128491620111733, | |
| "grad_norm": 1.0390197038650513, | |
| "learning_rate": 1.570562798094747e-05, | |
| "loss": 1.2218, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.3240223463687153, | |
| "grad_norm": 1.2470310926437378, | |
| "learning_rate": 1.5230328114318127e-05, | |
| "loss": 1.3041, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.335195530726257, | |
| "grad_norm": 1.1841509342193604, | |
| "learning_rate": 1.4761035809979395e-05, | |
| "loss": 1.2008, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.346368715083799, | |
| "grad_norm": 1.027030110359192, | |
| "learning_rate": 1.4297832153043656e-05, | |
| "loss": 1.0652, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.357541899441341, | |
| "grad_norm": 1.1887884140014648, | |
| "learning_rate": 1.3840797176616466e-05, | |
| "loss": 1.1927, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.368715083798883, | |
| "grad_norm": 1.1091079711914062, | |
| "learning_rate": 1.3390009847968504e-05, | |
| "loss": 1.157, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3798882681564244, | |
| "grad_norm": 1.1894493103027344, | |
| "learning_rate": 1.2945548054891321e-05, | |
| "loss": 1.1287, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3910614525139664, | |
| "grad_norm": 1.2934393882751465, | |
| "learning_rate": 1.2507488592239847e-05, | |
| "loss": 1.127, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.4022346368715084, | |
| "grad_norm": 1.1570240259170532, | |
| "learning_rate": 1.2075907148663579e-05, | |
| "loss": 1.0553, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.4134078212290504, | |
| "grad_norm": 1.172692060470581, | |
| "learning_rate": 1.1650878293528994e-05, | |
| "loss": 1.0155, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.4245810055865924, | |
| "grad_norm": 1.197474718093872, | |
| "learning_rate": 1.1232475464035385e-05, | |
| "loss": 1.135, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.435754189944134, | |
| "grad_norm": 1.1875431537628174, | |
| "learning_rate": 1.0820770952526155e-05, | |
| "loss": 1.1129, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.446927374301676, | |
| "grad_norm": 1.4438897371292114, | |
| "learning_rate": 1.0415835893998116e-05, | |
| "loss": 1.2491, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.458100558659218, | |
| "grad_norm": 1.3826245069503784, | |
| "learning_rate": 1.0017740253810609e-05, | |
| "loss": 1.1198, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.46927374301676, | |
| "grad_norm": 1.3730559349060059, | |
| "learning_rate": 9.62655281559679e-06, | |
| "loss": 1.0396, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4804469273743015, | |
| "grad_norm": 1.6272450685501099, | |
| "learning_rate": 9.242341169379076e-06, | |
| "loss": 1.1171, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.4916201117318435, | |
| "grad_norm": 1.9734584093093872, | |
| "learning_rate": 8.865171699890834e-06, | |
| "loss": 1.1374, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.5027932960893855, | |
| "grad_norm": 1.3264243602752686, | |
| "learning_rate": 8.49510957510633e-06, | |
| "loss": 1.3275, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.5139664804469275, | |
| "grad_norm": 1.3713715076446533, | |
| "learning_rate": 8.132218734980852e-06, | |
| "loss": 1.2514, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.5251396648044695, | |
| "grad_norm": 1.2266292572021484, | |
| "learning_rate": 7.776561880403072e-06, | |
| "loss": 1.1443, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.536312849162011, | |
| "grad_norm": 1.2601666450500488, | |
| "learning_rate": 7.4282004623615396e-06, | |
| "loss": 1.2684, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.547486033519553, | |
| "grad_norm": 1.222358226776123, | |
| "learning_rate": 7.0871946713269856e-06, | |
| "loss": 1.1953, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.558659217877095, | |
| "grad_norm": 1.1266566514968872, | |
| "learning_rate": 6.753603426852589e-06, | |
| "loss": 1.1542, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.5698324022346366, | |
| "grad_norm": 1.035483479499817, | |
| "learning_rate": 6.427484367393699e-06, | |
| "loss": 1.1903, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.5810055865921786, | |
| "grad_norm": 1.0014451742172241, | |
| "learning_rate": 6.108893840348995e-06, | |
| "loss": 1.0823, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5921787709497206, | |
| "grad_norm": 1.1314361095428467, | |
| "learning_rate": 5.797886892324694e-06, | |
| "loss": 1.2311, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.6033519553072626, | |
| "grad_norm": 1.0532621145248413, | |
| "learning_rate": 5.494517259623477e-06, | |
| "loss": 1.1181, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.6145251396648046, | |
| "grad_norm": 1.0458093881607056, | |
| "learning_rate": 5.198837358959901e-06, | |
| "loss": 1.1474, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.6256983240223466, | |
| "grad_norm": 1.1469138860702515, | |
| "learning_rate": 4.910898278403669e-06, | |
| "loss": 1.1677, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.636871508379888, | |
| "grad_norm": 1.135291337966919, | |
| "learning_rate": 4.630749768552589e-06, | |
| "loss": 1.1351, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.64804469273743, | |
| "grad_norm": 1.09344482421875, | |
| "learning_rate": 4.358440233936617e-06, | |
| "loss": 1.1221, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.659217877094972, | |
| "grad_norm": 1.190507411956787, | |
| "learning_rate": 4.094016724654359e-06, | |
| "loss": 1.0893, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.6703910614525137, | |
| "grad_norm": 1.1962109804153442, | |
| "learning_rate": 3.837524928243774e-06, | |
| "loss": 1.0827, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.6815642458100557, | |
| "grad_norm": 1.3094158172607422, | |
| "learning_rate": 3.589009161788104e-06, | |
| "loss": 1.0457, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.6927374301675977, | |
| "grad_norm": 1.3989758491516113, | |
| "learning_rate": 3.3485123642587658e-06, | |
| "loss": 1.0812, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.7039106145251397, | |
| "grad_norm": 1.4350848197937012, | |
| "learning_rate": 3.116076089096265e-06, | |
| "loss": 1.0437, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.7150837988826817, | |
| "grad_norm": 1.4926385879516602, | |
| "learning_rate": 2.8917404970305097e-06, | |
| "loss": 1.0832, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.7262569832402237, | |
| "grad_norm": 1.7393081188201904, | |
| "learning_rate": 2.675544349141779e-06, | |
| "loss": 1.0668, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.7374301675977653, | |
| "grad_norm": 1.9382210969924927, | |
| "learning_rate": 2.4675250001635232e-06, | |
| "loss": 1.0107, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.7486033519553073, | |
| "grad_norm": 0.7568252086639404, | |
| "learning_rate": 2.2677183920281343e-06, | |
| "loss": 1.0928, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.7597765363128492, | |
| "grad_norm": 0.9040629267692566, | |
| "learning_rate": 2.076159047656889e-06, | |
| "loss": 1.3023, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.770949720670391, | |
| "grad_norm": 0.9287111163139343, | |
| "learning_rate": 1.892880064994934e-06, | |
| "loss": 1.1539, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.782122905027933, | |
| "grad_norm": 0.9362537860870361, | |
| "learning_rate": 1.7179131112926627e-06, | |
| "loss": 1.1163, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.793296089385475, | |
| "grad_norm": 0.9989909529685974, | |
| "learning_rate": 1.551288417634106e-06, | |
| "loss": 1.1343, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.793296089385475, | |
| "eval_loss": 1.2524751424789429, | |
| "eval_runtime": 3.5444, | |
| "eval_samples_per_second": 42.603, | |
| "eval_steps_per_second": 21.442, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.804469273743017, | |
| "grad_norm": 0.999933660030365, | |
| "learning_rate": 1.3930347737136196e-06, | |
| "loss": 1.1728, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.815642458100559, | |
| "grad_norm": 1.084299921989441, | |
| "learning_rate": 1.2431795228615372e-06, | |
| "loss": 1.1947, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.826815642458101, | |
| "grad_norm": 1.0853486061096191, | |
| "learning_rate": 1.101748557319715e-06, | |
| "loss": 1.1514, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.8379888268156424, | |
| "grad_norm": 1.0919440984725952, | |
| "learning_rate": 9.687663137678604e-07, | |
| "loss": 1.1536, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.8491620111731844, | |
| "grad_norm": 1.0306580066680908, | |
| "learning_rate": 8.442557691013043e-07, | |
| "loss": 1.1416, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.8603351955307263, | |
| "grad_norm": 1.1208820343017578, | |
| "learning_rate": 7.282384364610206e-07, | |
| "loss": 1.1873, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.871508379888268, | |
| "grad_norm": 1.0364559888839722, | |
| "learning_rate": 6.207343615165561e-07, | |
| "loss": 1.0184, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.88268156424581, | |
| "grad_norm": 0.9816321730613708, | |
| "learning_rate": 5.217621190024779e-07, | |
| "loss": 1.017, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.893854748603352, | |
| "grad_norm": 1.1331318616867065, | |
| "learning_rate": 4.3133880950905205e-07, | |
| "loss": 1.1131, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.905027932960894, | |
| "grad_norm": 1.209434151649475, | |
| "learning_rate": 3.494800565275125e-07, | |
| "loss": 1.18, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.916201117318436, | |
| "grad_norm": 1.2905470132827759, | |
| "learning_rate": 2.762000037506485e-07, | |
| "loss": 1.1171, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.927374301675978, | |
| "grad_norm": 1.216124176979065, | |
| "learning_rate": 2.115113126290258e-07, | |
| "loss": 1.0804, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.9385474860335195, | |
| "grad_norm": 1.300574541091919, | |
| "learning_rate": 1.554251601833201e-07, | |
| "loss": 1.0849, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.9497206703910615, | |
| "grad_norm": 1.376492977142334, | |
| "learning_rate": 1.0795123707312283e-07, | |
| "loss": 1.0931, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.9608938547486034, | |
| "grad_norm": 1.507287859916687, | |
| "learning_rate": 6.909774592258056e-08, | |
| "loss": 1.1502, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.972067039106145, | |
| "grad_norm": 1.6715766191482544, | |
| "learning_rate": 3.8871399903134265e-08, | |
| "loss": 1.0179, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.983240223463687, | |
| "grad_norm": 2.2043421268463135, | |
| "learning_rate": 1.7277421573608232e-08, | |
| "loss": 0.9667, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.994413407821229, | |
| "grad_norm": 1.0078221559524536, | |
| "learning_rate": 4.319541977831909e-09, | |
| "loss": 1.1912, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.005586592178771, | |
| "grad_norm": 2.012565851211548, | |
| "learning_rate": 0.0, | |
| "loss": 1.8972, | |
| "step": 269 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 269, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.738167723150541e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |