| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 23.369036027263874, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.48931118845939636, | |
| "learning_rate": 1.98125e-05, | |
| "loss": 1.7417, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.56327223777771, | |
| "learning_rate": 1.9604166666666668e-05, | |
| "loss": 1.7626, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.6054126620292664, | |
| "learning_rate": 1.9395833333333335e-05, | |
| "loss": 1.5774, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.8035449385643005, | |
| "learning_rate": 1.9187500000000002e-05, | |
| "loss": 1.4772, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.9972027540206909, | |
| "learning_rate": 1.897916666666667e-05, | |
| "loss": 1.409, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.7526600360870361, | |
| "learning_rate": 1.8781250000000003e-05, | |
| "loss": 1.3088, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.8370587825775146, | |
| "learning_rate": 1.8572916666666666e-05, | |
| "loss": 1.2086, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.8337924480438232, | |
| "learning_rate": 1.8364583333333334e-05, | |
| "loss": 1.1321, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.0028570890426636, | |
| "learning_rate": 1.815625e-05, | |
| "loss": 1.0439, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.7663702368736267, | |
| "learning_rate": 1.7947916666666668e-05, | |
| "loss": 1.026, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.8835870623588562, | |
| "learning_rate": 1.7739583333333335e-05, | |
| "loss": 0.9994, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 1.010624885559082, | |
| "learning_rate": 1.7531250000000003e-05, | |
| "loss": 0.9381, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 0.9645354151725769, | |
| "learning_rate": 1.7322916666666666e-05, | |
| "loss": 0.8667, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 0.9926008582115173, | |
| "learning_rate": 1.7114583333333334e-05, | |
| "loss": 0.8729, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 1.0364482402801514, | |
| "learning_rate": 1.690625e-05, | |
| "loss": 0.836, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 1.3519001007080078, | |
| "learning_rate": 1.6697916666666668e-05, | |
| "loss": 0.8126, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 1.2280508279800415, | |
| "learning_rate": 1.6489583333333335e-05, | |
| "loss": 0.7953, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "grad_norm": 0.8931779265403748, | |
| "learning_rate": 1.6281250000000003e-05, | |
| "loss": 0.7813, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 2.7053027153015137, | |
| "learning_rate": 1.6072916666666667e-05, | |
| "loss": 0.7256, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 1.411024808883667, | |
| "learning_rate": 1.5864583333333334e-05, | |
| "loss": 0.7336, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "grad_norm": 1.0111807584762573, | |
| "learning_rate": 1.565625e-05, | |
| "loss": 0.7207, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "grad_norm": 2.213623523712158, | |
| "learning_rate": 1.544791666666667e-05, | |
| "loss": 0.664, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.17, | |
| "grad_norm": 1.3642323017120361, | |
| "learning_rate": 1.5239583333333334e-05, | |
| "loss": 0.6848, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "grad_norm": 1.3692028522491455, | |
| "learning_rate": 1.5031250000000001e-05, | |
| "loss": 0.6663, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "grad_norm": 1.9850131273269653, | |
| "learning_rate": 1.4822916666666667e-05, | |
| "loss": 0.6199, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 1.6070563793182373, | |
| "learning_rate": 1.4614583333333334e-05, | |
| "loss": 0.6295, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "grad_norm": 3.226116418838501, | |
| "learning_rate": 1.4406250000000001e-05, | |
| "loss": 0.6227, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "grad_norm": 1.5464348793029785, | |
| "learning_rate": 1.4197916666666667e-05, | |
| "loss": 0.589, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "grad_norm": 1.4290672540664673, | |
| "learning_rate": 1.3989583333333334e-05, | |
| "loss": 0.6263, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "grad_norm": 1.682243824005127, | |
| "learning_rate": 1.3781250000000001e-05, | |
| "loss": 0.5747, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.66, | |
| "grad_norm": 1.6785274744033813, | |
| "learning_rate": 1.3572916666666667e-05, | |
| "loss": 0.5809, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 9.97, | |
| "grad_norm": 2.156558036804199, | |
| "learning_rate": 1.3364583333333334e-05, | |
| "loss": 0.5725, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "grad_norm": 1.739134430885315, | |
| "learning_rate": 1.3156250000000001e-05, | |
| "loss": 0.549, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 10.59, | |
| "grad_norm": 1.2140896320343018, | |
| "learning_rate": 1.2947916666666667e-05, | |
| "loss": 0.5493, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 10.91, | |
| "grad_norm": 1.517707109451294, | |
| "learning_rate": 1.2739583333333334e-05, | |
| "loss": 0.5589, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.22, | |
| "grad_norm": 1.8752249479293823, | |
| "learning_rate": 1.2531250000000001e-05, | |
| "loss": 0.5001, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 11.53, | |
| "grad_norm": 1.8684614896774292, | |
| "learning_rate": 1.2322916666666667e-05, | |
| "loss": 0.52, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 11.84, | |
| "grad_norm": 2.3552052974700928, | |
| "learning_rate": 1.2114583333333334e-05, | |
| "loss": 0.5479, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 12.15, | |
| "grad_norm": 1.2074153423309326, | |
| "learning_rate": 1.1906250000000001e-05, | |
| "loss": 0.4977, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 12.46, | |
| "grad_norm": 3.0007381439208984, | |
| "learning_rate": 1.1697916666666667e-05, | |
| "loss": 0.5252, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.78, | |
| "grad_norm": 1.8956348896026611, | |
| "learning_rate": 1.1489583333333334e-05, | |
| "loss": 0.503, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 13.09, | |
| "grad_norm": 1.3779544830322266, | |
| "learning_rate": 1.1281250000000001e-05, | |
| "loss": 0.4991, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "grad_norm": 1.5404250621795654, | |
| "learning_rate": 1.1072916666666667e-05, | |
| "loss": 0.5026, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 13.71, | |
| "grad_norm": 2.149167060852051, | |
| "learning_rate": 1.0864583333333334e-05, | |
| "loss": 0.4788, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 1.4978464841842651, | |
| "learning_rate": 1.0656250000000002e-05, | |
| "loss": 0.4605, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 14.33, | |
| "grad_norm": 1.5203664302825928, | |
| "learning_rate": 1.0447916666666667e-05, | |
| "loss": 0.4858, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 14.64, | |
| "grad_norm": 1.850074291229248, | |
| "learning_rate": 1.0239583333333334e-05, | |
| "loss": 0.4797, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 14.96, | |
| "grad_norm": 1.7591063976287842, | |
| "learning_rate": 1.0031250000000002e-05, | |
| "loss": 0.4619, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 15.27, | |
| "grad_norm": 1.4994142055511475, | |
| "learning_rate": 9.822916666666667e-06, | |
| "loss": 0.4627, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 15.58, | |
| "grad_norm": 1.888311743736267, | |
| "learning_rate": 9.614583333333334e-06, | |
| "loss": 0.4469, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.89, | |
| "grad_norm": 1.7547377347946167, | |
| "learning_rate": 9.406250000000002e-06, | |
| "loss": 0.4379, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 16.2, | |
| "grad_norm": 1.353371262550354, | |
| "learning_rate": 9.197916666666667e-06, | |
| "loss": 0.4464, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 16.51, | |
| "grad_norm": 1.8960446119308472, | |
| "learning_rate": 8.989583333333334e-06, | |
| "loss": 0.4181, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 16.83, | |
| "grad_norm": 1.644089937210083, | |
| "learning_rate": 8.781250000000002e-06, | |
| "loss": 0.481, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 17.14, | |
| "grad_norm": 2.3077402114868164, | |
| "learning_rate": 8.572916666666667e-06, | |
| "loss": 0.4199, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 17.45, | |
| "grad_norm": 2.0317394733428955, | |
| "learning_rate": 8.364583333333334e-06, | |
| "loss": 0.4178, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 17.76, | |
| "grad_norm": 1.8424252271652222, | |
| "learning_rate": 8.156250000000002e-06, | |
| "loss": 0.4355, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 18.07, | |
| "grad_norm": 2.730163335800171, | |
| "learning_rate": 7.947916666666667e-06, | |
| "loss": 0.4449, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 18.38, | |
| "grad_norm": 1.7422763109207153, | |
| "learning_rate": 7.739583333333333e-06, | |
| "loss": 0.437, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 18.7, | |
| "grad_norm": 1.8822110891342163, | |
| "learning_rate": 7.531250000000001e-06, | |
| "loss": 0.4313, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 1.8389759063720703, | |
| "learning_rate": 7.322916666666667e-06, | |
| "loss": 0.4112, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 19.32, | |
| "grad_norm": 1.2449133396148682, | |
| "learning_rate": 7.114583333333334e-06, | |
| "loss": 0.3969, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "grad_norm": 1.4220046997070312, | |
| "learning_rate": 6.906250000000001e-06, | |
| "loss": 0.4306, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 19.94, | |
| "grad_norm": 1.3844712972640991, | |
| "learning_rate": 6.697916666666667e-06, | |
| "loss": 0.4085, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "grad_norm": 1.5047410726547241, | |
| "learning_rate": 6.489583333333334e-06, | |
| "loss": 0.4421, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 20.56, | |
| "grad_norm": 1.6129776239395142, | |
| "learning_rate": 6.281250000000001e-06, | |
| "loss": 0.42, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 20.88, | |
| "grad_norm": 3.006173849105835, | |
| "learning_rate": 6.0729166666666675e-06, | |
| "loss": 0.3902, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 21.19, | |
| "grad_norm": 1.5414382219314575, | |
| "learning_rate": 5.864583333333334e-06, | |
| "loss": 0.3927, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 21.5, | |
| "grad_norm": 2.1584370136260986, | |
| "learning_rate": 5.656250000000001e-06, | |
| "loss": 0.4052, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 21.81, | |
| "grad_norm": 1.9206900596618652, | |
| "learning_rate": 5.4479166666666675e-06, | |
| "loss": 0.3872, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 22.12, | |
| "grad_norm": 1.6601423025131226, | |
| "learning_rate": 5.239583333333333e-06, | |
| "loss": 0.4221, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 22.43, | |
| "grad_norm": 1.8656572103500366, | |
| "learning_rate": 5.031250000000001e-06, | |
| "loss": 0.4031, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 22.75, | |
| "grad_norm": 1.9469555616378784, | |
| "learning_rate": 4.822916666666667e-06, | |
| "loss": 0.395, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 23.06, | |
| "grad_norm": 1.801340103149414, | |
| "learning_rate": 4.614583333333334e-06, | |
| "loss": 0.3584, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 23.37, | |
| "grad_norm": 2.101327657699585, | |
| "learning_rate": 4.40625e-06, | |
| "loss": 0.4129, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 1920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 3.8986916806656e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |