| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "global_step": 342, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.917808219178082e-06, | |
| "loss": 0.0322, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_oasst_export_accuracy": 0.49051699074786653, | |
| "eval_oasst_export_loss": 3.484375, | |
| "eval_oasst_export_runtime": 6.0372, | |
| "eval_oasst_export_samples_per_second": 181.21, | |
| "eval_oasst_export_steps_per_second": 2.982, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5.800391389432485e-06, | |
| "loss": 0.0246, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_oasst_export_accuracy": 0.489678997143981, | |
| "eval_oasst_export_loss": 3.779296875, | |
| "eval_oasst_export_runtime": 6.056, | |
| "eval_oasst_export_samples_per_second": 180.647, | |
| "eval_oasst_export_steps_per_second": 2.972, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5.682974559686888e-06, | |
| "loss": 0.0203, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_oasst_export_accuracy": 0.4914062900826022, | |
| "eval_oasst_export_loss": 3.734375, | |
| "eval_oasst_export_runtime": 6.0349, | |
| "eval_oasst_export_samples_per_second": 181.279, | |
| "eval_oasst_export_steps_per_second": 2.983, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5.565557729941292e-06, | |
| "loss": 0.0187, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_oasst_export_accuracy": 0.4879859080259265, | |
| "eval_oasst_export_loss": 3.7109375, | |
| "eval_oasst_export_runtime": 6.0607, | |
| "eval_oasst_export_samples_per_second": 180.508, | |
| "eval_oasst_export_steps_per_second": 2.97, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 5.448140900195694e-06, | |
| "loss": 0.0162, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_oasst_export_accuracy": 0.4880030099362099, | |
| "eval_oasst_export_loss": 3.783203125, | |
| "eval_oasst_export_runtime": 6.0304, | |
| "eval_oasst_export_samples_per_second": 181.415, | |
| "eval_oasst_export_steps_per_second": 2.985, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 5.330724070450098e-06, | |
| "loss": 0.0159, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_oasst_export_accuracy": 0.486292818907872, | |
| "eval_oasst_export_loss": 3.7890625, | |
| "eval_oasst_export_runtime": 6.0367, | |
| "eval_oasst_export_samples_per_second": 181.224, | |
| "eval_oasst_export_steps_per_second": 2.982, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5.213307240704501e-06, | |
| "loss": 0.0172, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_oasst_export_accuracy": 0.48716501633232434, | |
| "eval_oasst_export_loss": 3.806640625, | |
| "eval_oasst_export_runtime": 6.0405, | |
| "eval_oasst_export_samples_per_second": 181.11, | |
| "eval_oasst_export_steps_per_second": 2.98, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5.095890410958904e-06, | |
| "loss": 0.0157, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_oasst_export_accuracy": 0.48454842405896736, | |
| "eval_oasst_export_loss": 3.791015625, | |
| "eval_oasst_export_runtime": 6.0411, | |
| "eval_oasst_export_samples_per_second": 181.092, | |
| "eval_oasst_export_steps_per_second": 2.98, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.978473581213307e-06, | |
| "loss": 0.0163, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_oasst_export_accuracy": 0.49138918817231886, | |
| "eval_oasst_export_loss": 3.822265625, | |
| "eval_oasst_export_runtime": 5.6229, | |
| "eval_oasst_export_samples_per_second": 194.561, | |
| "eval_oasst_export_steps_per_second": 3.201, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.86105675146771e-06, | |
| "loss": 0.0139, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_oasst_export_accuracy": 0.4956817676534469, | |
| "eval_oasst_export_loss": 3.826171875, | |
| "eval_oasst_export_runtime": 6.0223, | |
| "eval_oasst_export_samples_per_second": 181.659, | |
| "eval_oasst_export_steps_per_second": 2.989, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.7436399217221134e-06, | |
| "loss": 0.0151, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_oasst_export_accuracy": 0.4845655259692508, | |
| "eval_oasst_export_loss": 3.904296875, | |
| "eval_oasst_export_runtime": 6.0257, | |
| "eval_oasst_export_samples_per_second": 181.554, | |
| "eval_oasst_export_steps_per_second": 2.987, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.626223091976516e-06, | |
| "loss": 0.0139, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_oasst_export_accuracy": 0.49049988883758316, | |
| "eval_oasst_export_loss": 3.958984375, | |
| "eval_oasst_export_runtime": 6.0136, | |
| "eval_oasst_export_samples_per_second": 181.92, | |
| "eval_oasst_export_steps_per_second": 2.993, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 4.50880626223092e-06, | |
| "loss": 0.0133, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_oasst_export_accuracy": 0.4803413541292562, | |
| "eval_oasst_export_loss": 3.845703125, | |
| "eval_oasst_export_runtime": 6.0221, | |
| "eval_oasst_export_samples_per_second": 181.663, | |
| "eval_oasst_export_steps_per_second": 2.989, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.3913894324853226e-06, | |
| "loss": 0.0142, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_oasst_export_accuracy": 0.4777076599456159, | |
| "eval_oasst_export_loss": 3.828125, | |
| "eval_oasst_export_runtime": 6.0368, | |
| "eval_oasst_export_samples_per_second": 181.223, | |
| "eval_oasst_export_steps_per_second": 2.982, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 4.273972602739726e-06, | |
| "loss": 0.0141, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_oasst_export_accuracy": 0.47088399774254786, | |
| "eval_oasst_export_loss": 3.998046875, | |
| "eval_oasst_export_runtime": 6.0502, | |
| "eval_oasst_export_samples_per_second": 180.82, | |
| "eval_oasst_export_steps_per_second": 2.975, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.156555772994129e-06, | |
| "loss": 0.0157, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_oasst_export_accuracy": 0.4726796983223026, | |
| "eval_oasst_export_loss": 3.8828125, | |
| "eval_oasst_export_runtime": 5.995, | |
| "eval_oasst_export_samples_per_second": 182.487, | |
| "eval_oasst_export_steps_per_second": 3.003, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.0391389432485325e-06, | |
| "loss": 0.0139, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_oasst_export_accuracy": 0.4744240931712072, | |
| "eval_oasst_export_loss": 3.9609375, | |
| "eval_oasst_export_runtime": 6.0696, | |
| "eval_oasst_export_samples_per_second": 180.244, | |
| "eval_oasst_export_steps_per_second": 2.966, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 3.921722113502935e-06, | |
| "loss": 0.0125, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_oasst_export_accuracy": 0.48210285088844423, | |
| "eval_oasst_export_loss": 4.03125, | |
| "eval_oasst_export_runtime": 5.616, | |
| "eval_oasst_export_samples_per_second": 194.8, | |
| "eval_oasst_export_steps_per_second": 3.205, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 3.8043052837573385e-06, | |
| "loss": 0.017, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_oasst_export_accuracy": 0.4897303028748311, | |
| "eval_oasst_export_loss": 3.970703125, | |
| "eval_oasst_export_runtime": 6.03, | |
| "eval_oasst_export_samples_per_second": 181.426, | |
| "eval_oasst_export_steps_per_second": 2.985, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 3.6868884540117416e-06, | |
| "loss": 0.0137, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_oasst_export_accuracy": 0.4794862586150873, | |
| "eval_oasst_export_loss": 3.986328125, | |
| "eval_oasst_export_runtime": 6.0192, | |
| "eval_oasst_export_samples_per_second": 181.751, | |
| "eval_oasst_export_steps_per_second": 2.99, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 3.569471624266145e-06, | |
| "loss": 0.0133, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_oasst_export_accuracy": 0.47693807398286386, | |
| "eval_oasst_export_loss": 4.140625, | |
| "eval_oasst_export_runtime": 6.0148, | |
| "eval_oasst_export_samples_per_second": 181.886, | |
| "eval_oasst_export_steps_per_second": 2.993, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 3.452054794520548e-06, | |
| "loss": 0.0133, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_oasst_export_accuracy": 0.48713081251175755, | |
| "eval_oasst_export_loss": 4.1484375, | |
| "eval_oasst_export_runtime": 6.0347, | |
| "eval_oasst_export_samples_per_second": 181.285, | |
| "eval_oasst_export_steps_per_second": 2.983, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 3.334637964774951e-06, | |
| "loss": 0.0154, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "eval_oasst_export_accuracy": 0.4752620867750928, | |
| "eval_oasst_export_loss": 4.09375, | |
| "eval_oasst_export_runtime": 5.6373, | |
| "eval_oasst_export_samples_per_second": 194.065, | |
| "eval_oasst_export_steps_per_second": 3.193, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 3.217221135029354e-06, | |
| "loss": 0.0139, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_oasst_export_accuracy": 0.4760658765584116, | |
| "eval_oasst_export_loss": 4.09765625, | |
| "eval_oasst_export_runtime": 6.055, | |
| "eval_oasst_export_samples_per_second": 180.676, | |
| "eval_oasst_export_steps_per_second": 2.973, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 3.099804305283757e-06, | |
| "loss": 0.0136, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "eval_oasst_export_accuracy": 0.47102081302481486, | |
| "eval_oasst_export_loss": 4.20703125, | |
| "eval_oasst_export_runtime": 6.0408, | |
| "eval_oasst_export_samples_per_second": 181.103, | |
| "eval_oasst_export_steps_per_second": 2.98, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2.9823874755381603e-06, | |
| "loss": 0.0129, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_oasst_export_accuracy": 0.4641458450908967, | |
| "eval_oasst_export_loss": 4.2578125, | |
| "eval_oasst_export_runtime": 6.0349, | |
| "eval_oasst_export_samples_per_second": 181.279, | |
| "eval_oasst_export_steps_per_second": 2.983, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2.8649706457925635e-06, | |
| "loss": 0.0128, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "eval_oasst_export_accuracy": 0.4649838386947822, | |
| "eval_oasst_export_loss": 4.24609375, | |
| "eval_oasst_export_runtime": 6.0355, | |
| "eval_oasst_export_samples_per_second": 181.26, | |
| "eval_oasst_export_steps_per_second": 2.982, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2.7475538160469666e-06, | |
| "loss": 0.0133, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_oasst_export_accuracy": 0.46412874318061326, | |
| "eval_oasst_export_loss": 4.2421875, | |
| "eval_oasst_export_runtime": 6.0367, | |
| "eval_oasst_export_samples_per_second": 181.224, | |
| "eval_oasst_export_steps_per_second": 2.982, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2.63013698630137e-06, | |
| "loss": 0.013, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_oasst_export_accuracy": 0.47182460280813365, | |
| "eval_oasst_export_loss": 4.28125, | |
| "eval_oasst_export_runtime": 6.0187, | |
| "eval_oasst_export_samples_per_second": 181.768, | |
| "eval_oasst_export_steps_per_second": 2.991, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 2.512720156555773e-06, | |
| "loss": 0.0133, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_oasst_export_accuracy": 0.47093530347339796, | |
| "eval_oasst_export_loss": 4.30078125, | |
| "eval_oasst_export_runtime": 6.0452, | |
| "eval_oasst_export_samples_per_second": 180.969, | |
| "eval_oasst_export_steps_per_second": 2.978, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2.395303326810176e-06, | |
| "loss": 0.0117, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_oasst_export_accuracy": 0.4726454945017358, | |
| "eval_oasst_export_loss": 4.41796875, | |
| "eval_oasst_export_runtime": 6.0234, | |
| "eval_oasst_export_samples_per_second": 181.624, | |
| "eval_oasst_export_steps_per_second": 2.988, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2.2778864970645793e-06, | |
| "loss": 0.0121, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_oasst_export_accuracy": 0.47096950729396475, | |
| "eval_oasst_export_loss": 4.4375, | |
| "eval_oasst_export_runtime": 6.0349, | |
| "eval_oasst_export_samples_per_second": 181.279, | |
| "eval_oasst_export_steps_per_second": 2.983, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2.1604696673189825e-06, | |
| "loss": 0.0121, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "eval_oasst_export_accuracy": 0.480375557949823, | |
| "eval_oasst_export_loss": 4.421875, | |
| "eval_oasst_export_runtime": 6.0242, | |
| "eval_oasst_export_samples_per_second": 181.602, | |
| "eval_oasst_export_steps_per_second": 2.988, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 2.0430528375733853e-06, | |
| "loss": 0.0128, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_oasst_export_accuracy": 0.4811964496434252, | |
| "eval_oasst_export_loss": 4.40625, | |
| "eval_oasst_export_runtime": 6.0198, | |
| "eval_oasst_export_samples_per_second": 181.734, | |
| "eval_oasst_export_steps_per_second": 2.99, | |
| "step": 340 | |
| } | |
| ], | |
| "max_steps": 513, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.4125538591477596e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |