| { | |
| "best_metric": 2.326810598373413, | |
| "best_model_checkpoint": "./output/training_results/C015_llama3-8b-base_pretrain_20240428_005832/checkpoint-155", | |
| "epoch": 1.0064935064935066, | |
| "eval_steps": 5, | |
| "global_step": 155, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006493506493506494, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.6141, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006493506493506494, | |
| "eval_loss": 2.635435104370117, | |
| "eval_runtime": 13.3872, | |
| "eval_samples_per_second": 81.421, | |
| "eval_steps_per_second": 0.672, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.032467532467532464, | |
| "grad_norm": 3.736102227472634, | |
| "learning_rate": 2.25e-06, | |
| "loss": 2.657, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.032467532467532464, | |
| "eval_loss": 2.6205813884735107, | |
| "eval_runtime": 13.3001, | |
| "eval_samples_per_second": 81.955, | |
| "eval_steps_per_second": 0.677, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 2.9360680843951172, | |
| "learning_rate": 5.25e-06, | |
| "loss": 2.6337, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "eval_loss": 2.5846035480499268, | |
| "eval_runtime": 13.3634, | |
| "eval_samples_per_second": 81.566, | |
| "eval_steps_per_second": 0.673, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09740259740259741, | |
| "grad_norm": 2.4441290480949354, | |
| "learning_rate": 9e-06, | |
| "loss": 2.5268, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09740259740259741, | |
| "eval_loss": 2.5515594482421875, | |
| "eval_runtime": 13.3452, | |
| "eval_samples_per_second": 81.677, | |
| "eval_steps_per_second": 0.674, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 2.5052749297132455, | |
| "learning_rate": 1.275e-05, | |
| "loss": 2.5275, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "eval_loss": 2.5321404933929443, | |
| "eval_runtime": 13.3665, | |
| "eval_samples_per_second": 81.547, | |
| "eval_steps_per_second": 0.673, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16233766233766234, | |
| "grad_norm": 2.570439112365278, | |
| "learning_rate": 1.4457320927100615e-05, | |
| "loss": 2.5005, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16233766233766234, | |
| "eval_loss": 2.513052225112915, | |
| "eval_runtime": 13.3936, | |
| "eval_samples_per_second": 81.382, | |
| "eval_steps_per_second": 0.672, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 2.1653661003420974, | |
| "learning_rate": 1.3178060763055965e-05, | |
| "loss": 2.5339, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "eval_loss": 2.496060609817505, | |
| "eval_runtime": 13.4225, | |
| "eval_samples_per_second": 81.207, | |
| "eval_steps_per_second": 0.671, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 2.4298156508230955, | |
| "learning_rate": 1.200291011775234e-05, | |
| "loss": 2.5335, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "eval_loss": 2.4807701110839844, | |
| "eval_runtime": 13.3964, | |
| "eval_samples_per_second": 81.365, | |
| "eval_steps_per_second": 0.672, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 2.371474812035467, | |
| "learning_rate": 1.092418047398154e-05, | |
| "loss": 2.4252, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "eval_loss": 2.464339256286621, | |
| "eval_runtime": 13.3751, | |
| "eval_samples_per_second": 81.495, | |
| "eval_steps_per_second": 0.673, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2922077922077922, | |
| "grad_norm": 2.233872363768768, | |
| "learning_rate": 9.934692235419926e-06, | |
| "loss": 2.4445, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2922077922077922, | |
| "eval_loss": 2.4518375396728516, | |
| "eval_runtime": 13.4165, | |
| "eval_samples_per_second": 81.243, | |
| "eval_steps_per_second": 0.671, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 2.228351075418733, | |
| "learning_rate": 9.02774500281382e-06, | |
| "loss": 2.4594, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "eval_loss": 2.4393582344055176, | |
| "eval_runtime": 13.3552, | |
| "eval_samples_per_second": 81.616, | |
| "eval_steps_per_second": 0.674, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 2.2182357621988693, | |
| "learning_rate": 8.197089350822288e-06, | |
| "loss": 2.4498, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "eval_loss": 2.4287211894989014, | |
| "eval_runtime": 13.3236, | |
| "eval_samples_per_second": 81.809, | |
| "eval_steps_per_second": 0.675, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 2.3864673038770854, | |
| "learning_rate": 7.436900041840997e-06, | |
| "loss": 2.3821, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "eval_loss": 2.4184141159057617, | |
| "eval_runtime": 13.3882, | |
| "eval_samples_per_second": 81.415, | |
| "eval_steps_per_second": 0.672, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.42207792207792205, | |
| "grad_norm": 1.9510161344368442, | |
| "learning_rate": 6.741750615310939e-06, | |
| "loss": 2.4317, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42207792207792205, | |
| "eval_loss": 2.4091267585754395, | |
| "eval_runtime": 13.358, | |
| "eval_samples_per_second": 81.599, | |
| "eval_steps_per_second": 0.674, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.9482390057148204, | |
| "learning_rate": 6.106589293139538e-06, | |
| "loss": 2.3931, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "eval_loss": 2.40012788772583, | |
| "eval_runtime": 13.4261, | |
| "eval_samples_per_second": 81.185, | |
| "eval_steps_per_second": 0.67, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.487012987012987, | |
| "grad_norm": 2.112671713085814, | |
| "learning_rate": 5.526716143930102e-06, | |
| "loss": 2.3695, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.487012987012987, | |
| "eval_loss": 2.3934359550476074, | |
| "eval_runtime": 13.3826, | |
| "eval_samples_per_second": 81.449, | |
| "eval_steps_per_second": 0.673, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 2.0831282196798315, | |
| "learning_rate": 4.997761450728939e-06, | |
| "loss": 2.3981, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "eval_loss": 2.3855459690093994, | |
| "eval_runtime": 13.4481, | |
| "eval_samples_per_second": 81.052, | |
| "eval_steps_per_second": 0.669, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.551948051948052, | |
| "grad_norm": 1.850619691727126, | |
| "learning_rate": 4.515665228960038e-06, | |
| "loss": 2.3952, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.551948051948052, | |
| "eval_loss": 2.37890887260437, | |
| "eval_runtime": 13.4148, | |
| "eval_samples_per_second": 81.254, | |
| "eval_steps_per_second": 0.671, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 1.92650957243675, | |
| "learning_rate": 4.0766578431245434e-06, | |
| "loss": 2.4137, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "eval_loss": 2.3720638751983643, | |
| "eval_runtime": 13.3735, | |
| "eval_samples_per_second": 81.504, | |
| "eval_steps_per_second": 0.673, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6168831168831169, | |
| "grad_norm": 1.9166696037048836, | |
| "learning_rate": 3.6772416726983343e-06, | |
| "loss": 2.3614, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6168831168831169, | |
| "eval_loss": 2.366936683654785, | |
| "eval_runtime": 13.4784, | |
| "eval_samples_per_second": 80.87, | |
| "eval_steps_per_second": 0.668, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 1.9337017607143168, | |
| "learning_rate": 3.3141737794662055e-06, | |
| "loss": 2.3467, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "eval_loss": 2.361203193664551, | |
| "eval_runtime": 13.4232, | |
| "eval_samples_per_second": 81.203, | |
| "eval_steps_per_second": 0.67, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 1.800114241259619, | |
| "learning_rate": 2.984449530286649e-06, | |
| "loss": 2.4012, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "eval_loss": 2.3568994998931885, | |
| "eval_runtime": 13.3387, | |
| "eval_samples_per_second": 81.717, | |
| "eval_steps_per_second": 0.675, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.8774879479561137, | |
| "learning_rate": 2.685287130987944e-06, | |
| "loss": 2.3224, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "eval_loss": 2.352806329727173, | |
| "eval_runtime": 13.3565, | |
| "eval_samples_per_second": 81.608, | |
| "eval_steps_per_second": 0.674, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7467532467532467, | |
| "grad_norm": 1.8730632952307495, | |
| "learning_rate": 2.4141130287548048e-06, | |
| "loss": 2.3348, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7467532467532467, | |
| "eval_loss": 2.348268985748291, | |
| "eval_runtime": 13.3814, | |
| "eval_samples_per_second": 81.456, | |
| "eval_steps_per_second": 0.673, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 1.9700141360167842, | |
| "learning_rate": 2.168548141976706e-06, | |
| "loss": 2.3573, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "eval_loss": 2.3447518348693848, | |
| "eval_runtime": 13.3872, | |
| "eval_samples_per_second": 81.421, | |
| "eval_steps_per_second": 0.672, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8116883116883117, | |
| "grad_norm": 1.8580210267921047, | |
| "learning_rate": 1.946394878094437e-06, | |
| "loss": 2.306, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8116883116883117, | |
| "eval_loss": 2.3411996364593506, | |
| "eval_runtime": 13.4315, | |
| "eval_samples_per_second": 81.153, | |
| "eval_steps_per_second": 0.67, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 1.8818877549057513, | |
| "learning_rate": 1.745624901501792e-06, | |
| "loss": 2.342, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "eval_loss": 2.338190793991089, | |
| "eval_runtime": 13.3467, | |
| "eval_samples_per_second": 81.668, | |
| "eval_steps_per_second": 0.674, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8766233766233766, | |
| "grad_norm": 1.9087045844270383, | |
| "learning_rate": 1.564367615035273e-06, | |
| "loss": 2.3045, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8766233766233766, | |
| "eval_loss": 2.3356211185455322, | |
| "eval_runtime": 13.3981, | |
| "eval_samples_per_second": 81.355, | |
| "eval_steps_per_second": 0.672, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.8238462929669437, | |
| "learning_rate": 1.4008993200171148e-06, | |
| "loss": 2.2959, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 2.3329813480377197, | |
| "eval_runtime": 13.4034, | |
| "eval_samples_per_second": 81.322, | |
| "eval_steps_per_second": 0.671, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9415584415584416, | |
| "grad_norm": 1.824603087231258, | |
| "learning_rate": 1.253633021206854e-06, | |
| "loss": 2.3545, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9415584415584416, | |
| "eval_loss": 2.330482006072998, | |
| "eval_runtime": 13.3871, | |
| "eval_samples_per_second": 81.422, | |
| "eval_steps_per_second": 0.672, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 1.7913666925001428, | |
| "learning_rate": 1.1211088443646446e-06, | |
| "loss": 2.3446, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "eval_loss": 2.3284924030303955, | |
| "eval_runtime": 13.3985, | |
| "eval_samples_per_second": 81.353, | |
| "eval_steps_per_second": 0.672, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0064935064935066, | |
| "grad_norm": 3.0555751125596067, | |
| "learning_rate": 1.0019850354367667e-06, | |
| "loss": 2.2502, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0064935064935066, | |
| "eval_loss": 2.326810598373413, | |
| "eval_runtime": 13.3767, | |
| "eval_samples_per_second": 81.485, | |
| "eval_steps_per_second": 0.673, | |
| "step": 155 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 616, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 5, | |
| "total_flos": 15912853831680.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |