| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.8938906752411575, | |
| "eval_steps": 500, | |
| "global_step": 75, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03858520900321544, | |
| "grad_norm": 6.827114582061768, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.2682, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07717041800643087, | |
| "grad_norm": 6.703758239746094, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.2274, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.1157556270096463, | |
| "grad_norm": 6.74079704284668, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.2669, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15434083601286175, | |
| "grad_norm": 5.965632438659668, | |
| "learning_rate": 5e-06, | |
| "loss": 1.2216, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19292604501607716, | |
| "grad_norm": 4.468871593475342, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.1837, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.2315112540192926, | |
| "grad_norm": 2.849684715270996, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.1514, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.27009646302250806, | |
| "grad_norm": 5.281802177429199, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.1846, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3086816720257235, | |
| "grad_norm": 5.24518346786499, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1349, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34726688102893893, | |
| "grad_norm": 4.561065673828125, | |
| "learning_rate": 9.994504457428557e-06, | |
| "loss": 1.0946, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3858520900321543, | |
| "grad_norm": 4.529706001281738, | |
| "learning_rate": 9.978029910109491e-06, | |
| "loss": 1.0995, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.42443729903536975, | |
| "grad_norm": 3.045126438140869, | |
| "learning_rate": 9.950612572673255e-06, | |
| "loss": 1.0659, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4630225080385852, | |
| "grad_norm": 2.239044666290283, | |
| "learning_rate": 9.91231271437788e-06, | |
| "loss": 1.018, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5016077170418006, | |
| "grad_norm": 2.176582098007202, | |
| "learning_rate": 9.863214526624065e-06, | |
| "loss": 1.0193, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5401929260450161, | |
| "grad_norm": 1.5574040412902832, | |
| "learning_rate": 9.803425937884202e-06, | |
| "loss": 0.9824, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5787781350482315, | |
| "grad_norm": 1.3071755170822144, | |
| "learning_rate": 9.733078376452172e-06, | |
| "loss": 0.9843, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.617363344051447, | |
| "grad_norm": 1.1668739318847656, | |
| "learning_rate": 9.652326481535434e-06, | |
| "loss": 0.9856, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6559485530546624, | |
| "grad_norm": 1.3676371574401855, | |
| "learning_rate": 9.561347763324484e-06, | |
| "loss": 0.9434, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6945337620578779, | |
| "grad_norm": 1.3451117277145386, | |
| "learning_rate": 9.460342212786933e-06, | |
| "loss": 0.9642, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7331189710610932, | |
| "grad_norm": 0.9681344032287598, | |
| "learning_rate": 9.349531862043952e-06, | |
| "loss": 0.9553, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7717041800643086, | |
| "grad_norm": 1.0520753860473633, | |
| "learning_rate": 9.229160296295488e-06, | |
| "loss": 0.9128, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8102893890675241, | |
| "grad_norm": 0.9027125239372253, | |
| "learning_rate": 9.099492118367123e-06, | |
| "loss": 0.9291, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8488745980707395, | |
| "grad_norm": 0.8002101182937622, | |
| "learning_rate": 8.960812367055646e-06, | |
| "loss": 0.9195, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.887459807073955, | |
| "grad_norm": 0.787247896194458, | |
| "learning_rate": 8.81342589055191e-06, | |
| "loss": 0.9461, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9260450160771704, | |
| "grad_norm": 0.7884970903396606, | |
| "learning_rate": 8.657656676318346e-06, | |
| "loss": 0.9381, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9646302250803859, | |
| "grad_norm": 0.6913259625434875, | |
| "learning_rate": 8.49384713889421e-06, | |
| "loss": 0.8714, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0032154340836013, | |
| "grad_norm": 0.7232531309127808, | |
| "learning_rate": 8.32235736719411e-06, | |
| "loss": 0.9519, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0418006430868167, | |
| "grad_norm": 0.7492582201957703, | |
| "learning_rate": 8.143564332954426e-06, | |
| "loss": 0.8788, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0803858520900322, | |
| "grad_norm": 0.6388415694236755, | |
| "learning_rate": 7.957861062067614e-06, | |
| "loss": 0.9166, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1189710610932475, | |
| "grad_norm": 0.6874092817306519, | |
| "learning_rate": 7.765655770625997e-06, | |
| "loss": 0.8655, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.157556270096463, | |
| "grad_norm": 0.6590151190757751, | |
| "learning_rate": 7.56737096757421e-06, | |
| "loss": 0.8655, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1961414790996785, | |
| "grad_norm": 0.6995782852172852, | |
| "learning_rate": 7.363442525942827e-06, | |
| "loss": 0.8524, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.234726688102894, | |
| "grad_norm": 0.8694888949394226, | |
| "learning_rate": 7.1543187247048525e-06, | |
| "loss": 0.9178, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2733118971061093, | |
| "grad_norm": 0.72708660364151, | |
| "learning_rate": 6.9404592633612486e-06, | |
| "loss": 0.8379, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.3118971061093248, | |
| "grad_norm": 0.6599162220954895, | |
| "learning_rate": 6.722334251421665e-06, | |
| "loss": 0.9048, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3504823151125402, | |
| "grad_norm": 0.7191936373710632, | |
| "learning_rate": 6.500423175001705e-06, | |
| "loss": 0.8702, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3890675241157555, | |
| "grad_norm": 0.7617672085762024, | |
| "learning_rate": 6.275213842808383e-06, | |
| "loss": 0.875, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.427652733118971, | |
| "grad_norm": 0.5834156274795532, | |
| "learning_rate": 6.047201313830724e-06, | |
| "loss": 0.855, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4662379421221865, | |
| "grad_norm": 0.6282949447631836, | |
| "learning_rate": 5.816886809092651e-06, | |
| "loss": 0.8703, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.504823151125402, | |
| "grad_norm": 0.6048595905303955, | |
| "learning_rate": 5.584776609860414e-06, | |
| "loss": 0.8615, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5434083601286175, | |
| "grad_norm": 0.5948866605758667, | |
| "learning_rate": 5.351380944726465e-06, | |
| "loss": 0.8744, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5819935691318328, | |
| "grad_norm": 0.5789873600006104, | |
| "learning_rate": 5.117212868016303e-06, | |
| "loss": 0.8056, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6205787781350482, | |
| "grad_norm": 0.5074349045753479, | |
| "learning_rate": 4.882787131983698e-06, | |
| "loss": 0.8465, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6591639871382635, | |
| "grad_norm": 0.5655978322029114, | |
| "learning_rate": 4.6486190552735375e-06, | |
| "loss": 0.9188, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.697749196141479, | |
| "grad_norm": 0.5455473065376282, | |
| "learning_rate": 4.415223390139588e-06, | |
| "loss": 0.8093, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7363344051446945, | |
| "grad_norm": 0.6158671379089355, | |
| "learning_rate": 4.183113190907349e-06, | |
| "loss": 0.9393, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.77491961414791, | |
| "grad_norm": 0.5077454447746277, | |
| "learning_rate": 3.952798686169279e-06, | |
| "loss": 0.8114, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8135048231511255, | |
| "grad_norm": 0.47759971022605896, | |
| "learning_rate": 3.7247861571916183e-06, | |
| "loss": 0.8585, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.852090032154341, | |
| "grad_norm": 0.4739689230918884, | |
| "learning_rate": 3.4995768249982975e-06, | |
| "loss": 0.8259, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8906752411575563, | |
| "grad_norm": 0.5395683646202087, | |
| "learning_rate": 3.2776657485783357e-06, | |
| "loss": 0.8673, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9292604501607717, | |
| "grad_norm": 0.5374310612678528, | |
| "learning_rate": 3.059540736638751e-06, | |
| "loss": 0.7906, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.967845659163987, | |
| "grad_norm": 0.4914711117744446, | |
| "learning_rate": 2.8456812752951483e-06, | |
| "loss": 0.9054, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.0064308681672025, | |
| "grad_norm": 0.521967887878418, | |
| "learning_rate": 2.636557474057173e-06, | |
| "loss": 0.8763, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.045016077170418, | |
| "grad_norm": 0.4540090262889862, | |
| "learning_rate": 2.4326290324257896e-06, | |
| "loss": 0.8473, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.0836012861736335, | |
| "grad_norm": 0.4512139856815338, | |
| "learning_rate": 2.234344229374003e-06, | |
| "loss": 0.7919, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.122186495176849, | |
| "grad_norm": 0.5352345705032349, | |
| "learning_rate": 2.042138937932388e-06, | |
| "loss": 0.8196, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1607717041800645, | |
| "grad_norm": 0.44427061080932617, | |
| "learning_rate": 1.856435667045577e-06, | |
| "loss": 0.7881, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.19935691318328, | |
| "grad_norm": 0.5240706205368042, | |
| "learning_rate": 1.677642632805892e-06, | |
| "loss": 0.8631, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.237942122186495, | |
| "grad_norm": 0.5195144414901733, | |
| "learning_rate": 1.5061528611057917e-06, | |
| "loss": 0.7746, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.2765273311897105, | |
| "grad_norm": 0.43310004472732544, | |
| "learning_rate": 1.3423433236816563e-06, | |
| "loss": 0.8119, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.315112540192926, | |
| "grad_norm": 0.4922521710395813, | |
| "learning_rate": 1.186574109448091e-06, | |
| "loss": 0.8375, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3536977491961415, | |
| "grad_norm": 0.4353758692741394, | |
| "learning_rate": 1.0391876329443534e-06, | |
| "loss": 0.7619, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.392282958199357, | |
| "grad_norm": 0.4733580946922302, | |
| "learning_rate": 9.005078816328772e-07, | |
| "loss": 0.8132, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.4308681672025725, | |
| "grad_norm": 0.43942683935165405, | |
| "learning_rate": 7.708397037045129e-07, | |
| "loss": 0.7784, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.469453376205788, | |
| "grad_norm": 0.46390509605407715, | |
| "learning_rate": 6.50468137956049e-07, | |
| "loss": 0.8832, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.508038585209003, | |
| "grad_norm": 0.44283559918403625, | |
| "learning_rate": 5.396577872130676e-07, | |
| "loss": 0.8102, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5466237942122185, | |
| "grad_norm": 0.4312247037887573, | |
| "learning_rate": 4.386522366755169e-07, | |
| "loss": 0.7881, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.585209003215434, | |
| "grad_norm": 0.46290892362594604, | |
| "learning_rate": 3.4767351846456744e-07, | |
| "loss": 0.8807, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.6237942122186495, | |
| "grad_norm": 0.44196000695228577, | |
| "learning_rate": 2.669216235478295e-07, | |
| "loss": 0.7756, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.662379421221865, | |
| "grad_norm": 0.47514083981513977, | |
| "learning_rate": 1.9657406211579966e-07, | |
| "loss": 0.8685, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.7009646302250805, | |
| "grad_norm": 0.4315619468688965, | |
| "learning_rate": 1.3678547337593494e-07, | |
| "loss": 0.7501, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.739549839228296, | |
| "grad_norm": 0.4595395624637604, | |
| "learning_rate": 8.768728562211948e-08, | |
| "loss": 0.8121, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.778135048231511, | |
| "grad_norm": 0.4295967221260071, | |
| "learning_rate": 4.9387427326745287e-08, | |
| "loss": 0.7776, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.816720257234727, | |
| "grad_norm": 0.44801920652389526, | |
| "learning_rate": 2.1970089890509527e-08, | |
| "loss": 0.8149, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.855305466237942, | |
| "grad_norm": 0.4446917474269867, | |
| "learning_rate": 5.495542571443135e-09, | |
| "loss": 0.8291, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.8938906752411575, | |
| "grad_norm": 0.4660709500312805, | |
| "learning_rate": 0.0, | |
| "loss": 0.8268, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.8938906752411575, | |
| "step": 75, | |
| "total_flos": 66441907232768.0, | |
| "train_loss": 0.9123331260681152, | |
| "train_runtime": 3973.9028, | |
| "train_samples_per_second": 1.875, | |
| "train_steps_per_second": 0.019 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "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": 66441907232768.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |