| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.072805643081665, | |
| "learning_rate": 4.995066821070679e-05, | |
| "loss": 1.1181, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.6626899242401123, | |
| "learning_rate": 4.980286753286195e-05, | |
| "loss": 1.048, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 3.8898932933807373, | |
| "learning_rate": 4.9557181268217227e-05, | |
| "loss": 1.0172, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.715027093887329, | |
| "learning_rate": 4.9214579028215776e-05, | |
| "loss": 1.0621, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.233584403991699, | |
| "learning_rate": 4.877641290737884e-05, | |
| "loss": 0.994, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.1543967723846436, | |
| "learning_rate": 4.8244412147206284e-05, | |
| "loss": 1.0483, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.3235280513763428, | |
| "learning_rate": 4.762067631165049e-05, | |
| "loss": 1.0077, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.6280996799468994, | |
| "learning_rate": 4.690766700109659e-05, | |
| "loss": 0.9627, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.4375860691070557, | |
| "learning_rate": 4.610819813755038e-05, | |
| "loss": 0.9753, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.634675979614258, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 1.0424, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.759512424468994, | |
| "learning_rate": 4.426283106939474e-05, | |
| "loss": 0.9185, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.3730366230010986, | |
| "learning_rate": 4.3224215685535294e-05, | |
| "loss": 1.0032, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 3.0645976066589355, | |
| "learning_rate": 4.211367764821722e-05, | |
| "loss": 0.995, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.892836093902588, | |
| "learning_rate": 4.093559974371725e-05, | |
| "loss": 1.0167, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.1179070472717285, | |
| "learning_rate": 3.969463130731183e-05, | |
| "loss": 0.9908, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.6164073944091797, | |
| "learning_rate": 3.8395669874474915e-05, | |
| "loss": 0.9903, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 2.0717203617095947, | |
| "learning_rate": 3.704384185254288e-05, | |
| "loss": 1.0031, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.2452661991119385, | |
| "learning_rate": 3.564448228912682e-05, | |
| "loss": 0.9469, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.4002890586853027, | |
| "learning_rate": 3.4203113817116957e-05, | |
| "loss": 0.9221, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.4350810050964355, | |
| "learning_rate": 3.287466299138262e-05, | |
| "loss": 0.9588, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.338216781616211, | |
| "learning_rate": 3.1369268142084556e-05, | |
| "loss": 0.9762, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.716254472732544, | |
| "learning_rate": 2.9838736701271514e-05, | |
| "loss": 0.9577, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.5938873291015625, | |
| "learning_rate": 2.8289108977307067e-05, | |
| "loss": 1.0061, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.575774669647217, | |
| "learning_rate": 2.6726500642860154e-05, | |
| "loss": 1.0507, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.5186479091644287, | |
| "learning_rate": 2.5157078599138977e-05, | |
| "loss": 0.951, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.128981590270996, | |
| "learning_rate": 2.3587036637949388e-05, | |
| "loss": 0.8208, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 2.1169350147247314, | |
| "learning_rate": 2.2022570997628256e-05, | |
| "loss": 0.7953, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.3190295696258545, | |
| "learning_rate": 2.0469855909321564e-05, | |
| "loss": 0.7427, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.6951375007629395, | |
| "learning_rate": 1.908752507440689e-05, | |
| "loss": 0.763, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.6969361305236816, | |
| "learning_rate": 1.7573960460574133e-05, | |
| "loss": 0.7914, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 2.5656094551086426, | |
| "learning_rate": 1.6089703032168733e-05, | |
| "loss": 0.7915, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 3.075153112411499, | |
| "learning_rate": 1.4640610475167898e-05, | |
| "loss": 0.785, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 3.3679215908050537, | |
| "learning_rate": 1.3232401695866687e-05, | |
| "loss": 0.7762, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 3.499119281768799, | |
| "learning_rate": 1.1870634250967605e-05, | |
| "loss": 0.8098, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.9737751483917236, | |
| "learning_rate": 1.0560682414443315e-05, | |
| "loss": 0.7372, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 3.3601698875427246, | |
| "learning_rate": 9.307715967732491e-06, | |
| "loss": 0.8166, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 3.3548429012298584, | |
| "learning_rate": 8.116679796974388e-06, | |
| "loss": 0.7472, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.8706016540527344, | |
| "learning_rate": 6.992274377802327e-06, | |
| "loss": 0.7355, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.9783272743225098, | |
| "learning_rate": 5.9389372247138e-06, | |
| "loss": 0.7639, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.8318257331848145, | |
| "learning_rate": 4.960825378228082e-06, | |
| "loss": 0.8115, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 3.5943961143493652, | |
| "learning_rate": 4.061798998946459e-06, | |
| "loss": 0.7761, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 3.3045473098754883, | |
| "learning_rate": 3.245406133261858e-06, | |
| "loss": 0.805, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.879966974258423, | |
| "learning_rate": 2.514868710840723e-06, | |
| "loss": 0.7729, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 4.045947551727295, | |
| "learning_rate": 1.8730698291385518e-06, | |
| "loss": 0.7135, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 3.1562774181365967, | |
| "learning_rate": 1.3225423751313942e-06, | |
| "loss": 0.827, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 3.1504297256469727, | |
| "learning_rate": 8.65459029168153e-07, | |
| "loss": 0.7455, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 3.496607780456543, | |
| "learning_rate": 5.036236903938285e-07, | |
| "loss": 0.8043, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 3.2006499767303467, | |
| "learning_rate": 2.384643575837203e-07, | |
| "loss": 0.8145, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 3.356722116470337, | |
| "learning_rate": 7.102749348465165e-08, | |
| "loss": 0.7569, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.4808757305145264, | |
| "learning_rate": 1.973894904597207e-09, | |
| "loss": 0.7338, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 500, | |
| "total_flos": 3.216214367225774e+17, | |
| "train_loss": 0.888002516746521, | |
| "train_runtime": 10329.3048, | |
| "train_samples_per_second": 0.387, | |
| "train_steps_per_second": 0.048 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000, | |
| "total_flos": 3.216214367225774e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |